[359]
Constitutional End Games: Making Presidential
Term Limits Stick
ROSALIND DIXON
†
& DAVID LANDAU
†
Presidential term limits are an important and common protection of constitutional democracy
around the world. But they are often evaded because they raise particularly difficult compliance
problems that we call “end game” problems. Because presidents have overwhelming incentives
to remain in power, they may seek extraordinary means to evade term limits. Comparative
experience shows that presidents rely on a wide range of devices, such as formal constitutional
change, wholesale constitutional replacement, and manipulation of the judiciary to get around
permanent bans on reelection. In this Article, we draw on this experience to show that, in many
contexts, weaker bans on reelection for consecutive terms, rather than permanent bans on any
reelection, are the best response to the end game problem. Would-be authoritarian presidents are
more likely to comply with term limits that force a temporary exit from the presidency because
they hold open the prospect of an eventual return to power. Furthermore, a ban on consecutive
reelection will allow alternative political forces to strengthen and make substantial democratic
erosion less likely. In this sense, the United States’ oft-cited presidential term limit, which allows
two consecutive terms in office, but prohibits all future reelection, may not be the best model for
preserving democracy.
†
Professor of Law, University of New South Wales (UNSW) (Australia).
†
Mason Ladd Professor and Associate Dean for International Programs, Florida State University
College of Law. For comments on prior drafts or ideas, we thank Mark Tushnet, Mila Versteeg, Eric Posner,
and Brian Sheppard. We also thank Melissa Vogt for outstanding research assistance.
360 HASTINGS LAW JOURNAL [Vol. 71:359
TABLE OF CONTENTS
INTRODUCTION ............................................................................................... 361
I. WHY TERM LIMITS? TERM LIMITS AND DEMOCRACY ........................... 365
A. DEMOCRACY AND THE ADVANTAGES OF TERM LIMITS .............. 365
B. OBJECTIONS AND DISADVANTAGES ............................................. 369
II. MODELS OF PRESIDENTIAL TERM LIMITS .............................................. 371
III. THE PROBLEM OF COMPLIANCE ............................................................ 376
IV. THE CAUSES AND CONSEQUENCES OF NON-COMPLIANCE .................... 385
A. THE CAUSES OF EVASION ............................................................ 385
B. THE RISKS POSED BY TERM LIMIT EVASIONS ............................. 389
V. SOLVING THE END-GAME PROBLEM: THE ROLE OF CONSECUTIVE
T
ERM LIMITS ....................................................................................... 393
A. CHANGING PRESIDENTIAL INCENTIVES ....................................... 393
B. ARE CONSECUTIVE BANS STRONG ENOUGH? .............................. 398
C. DESIGNING CONSECUTIVE TERM LIMITS: ONE TERM OR TWO .... 402
VI. AVOIDING THE PROBLEM OF SHADOW PRESIDENTS .............................. 403
A. CONSTITUTIONAL DESIGN AND SHADOW RULE .......................... 405
B. CREATING INCENTIVES TO AVOID PARTISAN BEHAVIOR ............ 408
VII. ALTERNATIVE PROPOSALS .................................................................... 409
A. SCRAPPING TERM LIMITS ............................................................. 410
B. POPULAR ENFORCEMENT OF TERM LIMITS .................................. 412
C. PROSPECTIVE-ONLY RULES OF CHANGE ..................................... 414
CONCLUSION ................................................................................................... 416
February 2020] CONSTITUTIONAL END GAMES 361
INTRODUCTION
Presidential term limits are a common feature in democratic constitutions
worldwide. By our own calculations, over 80% of presidential and semi-
presidential constitutions in force today have presidential term limits.
1
Presidential term limits are common in constitutional democracies because they
are often seen as fundamental for the preservation of constitutional democracy.
Where presidents are able to remain in office indefinitely, comparative
experience shows that they can consolidate enormous amounts of power that
vitiate checks and balances by institutions such as legislatures and courts.
2
While
elections may continue to be held, they often become increasingly non-
competitive, as presidents amass formal and informal resources and use
institutions like the judiciary to undermine the opposition. In the United States,
although formal presidential term limits date only from the Twenty-Second
Amendment, which was passed in 1947, many modern commentators now see
the limits as a core protection of the democratic order.
3
As we will show, despite
its continued reputation as an international gold standard, the U.S. presidential
term limit is vulnerable to term limit evasion in key respects.
Presidents have very strong incentives to circumvent constitutional term
limits in order to remain in power. A study by Mila Versteeg and her co-authors
has recently shown that presidents, since 2000, have sought to evade term limits
in roughly 25% of cases.
4
Where presidents try to evade their term limit, they
succeed about two-thirds of the time.
5
Presidents evade term limits through a variety of routes. Most commonly,
many presidents seek a formal amendment to the term limit; in other cases, they
induce courts to reinterpret the term limits or even to excise them entirely from
1. See infra Table 1. Other studies have reached similar results. For example, Ginsburg, Melton, and
Elkins found that, between 1789 and 2006, about 60% of presidential and semi-presidential constitutions have
had presidential term limits; this rises to over 75% for constitutions in force in 2006. See Tom Ginsburg et al.,
On the Evasion of Executive Term Limits, 52 WM. & MARY L. REV. 1807, 1835–36, 1839 fig.1 (2011). In a study
of 92 presidential and semi-presidential systems between 1992 and 2006, Gideon Maltz found that 87 had some
form of presidential term limit. The relevant 92 countries included almost all key presidential democracies—
those with a population of more than two million people which had minimal norms of political openness. The
99 “regimes” in those 92 countries included 47 democracies and 52 competitive or electoral authoritarian
regimes. Gideon Maltz, The Case for Presidential Term Limits, 18 J. DEMOCRACY 128, 128–29 (2007).
2. See generally David Landau, Abusive Constitutionalism, 47 U.C. DAVIS L. REV. 189 (2013) (pointing
out that there is a degree of abuse of power if the president may remain in office indefinitely); Maltz, supra note
1 (explaining the adoption of presidential term limits after presidential power and the ongoing presidential power
abuse in countries that have yet to adopt a term limit).
3. See, e.g., Aziz Huq & Tom Ginsberg, How to Lose a Constitutional Democracy, 65 UCLA L. REV. 78,
143–44 (2018) (arguing that the Twenty-Second Amendment and Article V cut off one key route towards
authoritarianism in the United States).
4. See Mila Versteeg et al., The Law and Politics of Presidential Term Limits Evasion, 120 COLUM. L.
REV. (forthcoming 2020) (documenting cases and techniques of evasion since 2000).
5. See id.
362 HASTINGS LAW JOURNAL [Vol. 71:359
the constitutional order.
6
Because of the frequency of evasion attempts,
presidential term limits raise special challenges to democratic constitutionalism.
More than any other feature of a democratic constitution, presidential term
limits create limited incentives for compliance. One of the important insights in
constitutional scholarship in recent years is the degree to which many
constitutional norms effectively become self-enforcing, or self-stabilizing, over
time because they often serve as a basis for valuable forms of coordination
between different political parties or government officials.
7
Life-long bans on
reelection limits, however, remove almost all incentives for presidents to engage
in co-operation of this kind.
Faced with such limits, incumbent presidents face a form of “end period”
or “end game” problem. Compliance with term limits means that, in the short-
run, incumbent presidents are certain to lose the power and privileges associated
with high electoral office, and, in the long-run, gain only limited or uncertain
reputational benefits.
8
In many fragile democracies, political parties will also be
insufficiently strong and independent to exert pressure on a president to leave
office. Instead, they may actively encourage the president to extend their term
in office.
What is the response to this problem? One possibility, as some recent work
has suggested, is to give up the game entirely, and to remove term limits from
constitutions in contexts where they are likely to prove ineffective as constraints
on presidents.
9
But this solution throws the baby out with the bath water. It gives
up on a tool that is important for the preservation of democratic governance
simply because compliance is problematic.
Another response, which we have discussed extensively in recent work, is
to entrench term limits by requiring especially demanding procedures to change
them.
10
In the extreme, term limits are sometimes made completely
unamendable; less dramatically, constitutions can require special procedures
like heightened supermajorities or referenda before term limits can be altered.
These design solutions are sometimes helpful, but do not make the compliance
problem go away. Indeed, in some circumstances, these demanding or special
procedures to amend the term limits may worsen the end game problem because
these procedures heighten the pressure on presidential leaders to seek other
routes to eliminate term limits.
6. See id.
7. See, e.g., Tonja Jacobi et al., Judicial Review as a Self-Stabilizing Constitutional Mechanism, in
COMPARATIVE JUDICIAL REVIEW 185 (Erin F. Delaney & Rosalind Dixon eds., 2018); Daryl J. Levinson,
Parchment and Politics: The Positive Puzzle of Constitutional Commitment, 124 HARV. L. REV. 657 (2011).
8. See Bruce Baker, Outstaying One’s Welcome: The Presidential Third-Term Debate in Africa, 8
CONTEMP. POL. 285, 287 (2002).
9. See Ginsburg et al., supra note 1, at 1855–66.
10. See Rosalind Dixon & David Landau, Tiered Constitutional Design, 86 GEO. WASH. L. REV. 438
(2018).
February 2020] CONSTITUTIONAL END GAMES 363
The optimal solution, we assert here, is to focus on temporary or
consecutive, rather than permanent, bans on reelection. Consecutive bans, which
prohibit consecutive reelection after one or more terms, require leaders to leave
power periodically, but allow a return after one or more terms out of power. In
other words, consecutive bans are “soft,” “flexible,” or “weak” term limits that
limit the scope of consecutive presidential reelection but allow non-consecutive
reelection.
11
Consecutive bans, as compared to permanent bans, have a key advantage.
By giving presidents greater incentive to comply with democratic constitutional
requirements, consecutive bans ameliorate the end game problem. Presidents
who know they may be able to return to power later are more likely to leave
power in the first place.
At the same time, pushing powerful incumbents out of power, even
temporarily, will be crucial in preventing the erosion of democracy. When
presidents re-contest an election, they no longer enjoy the benefits of
incumbency, and social and political conditions will often have changed. Voters
may no longer see the president as necessary or indispensable to their well-
being. Other members of a president’s party may also have gained strength and
an independent reputation, such that the party itself has a greater incentive and
ability to support a broader range of candidates.
Following this introduction, this Article is divided into seven parts. Part I
outlines the existing scholarly literature on the relationship between term limits
and democracy. Part II explains the prevalence and design of different kinds of
term limits. Part III outlines the special problems of compliance posed by
presidential term limits, and the empirical evidence of term limit evasion drawn
from around the world over recent decades, particularly in Latin America and
Africa. Part IV explores the causes and consequences of evasions of term limits.
Parts V, VI, and VII deal with solutions. Part V explains and defends our
proposal for weaker, consecutive bans on reelection as a solution to the end game
problem. The proposal is rooted in the successful use of non-consecutive term
limits in several countries.
Part VI deals with an important caveat to our proposal: the problem of
shadow presidents, or, in other words, circumstances where non-consecutive
bans induce presidents to leave power formally but maintain power informally.
For example, consider Russia, where Vladimir Putin temporarily left the
presidency between 2008 and 2012 but continued to exercise considerable power
11. Branko Milanovic et al., Political Alternation as a Restraint on Investing in Influence: Evidence from
the Post-Communist Transition (World Bank Dev. Research Grp., Working Paper No. 4747, 2008). This is in
line with the work of political researchers, such as Cain and Lopez, who suggest that the effectiveness of term
limits is deeply dependent on questions of design and institutional setting. See Bruce E. Cain, The Varying
Impact of Legislative Term Limits, in LEGISLATIVE TERM LIMITS: PUBLIC CHOICE PERSPECTIVES 21, 22 (Bernard
Grofman ed., 1996); Edward J. López, Term Limits: Causes and Consequences, 114 PUB. CHOICE 1, 29 (2003).
364 HASTINGS LAW JOURNAL [Vol. 71:359
as both prime minister and party leader throughout his “absence.”
12
To be
effective, non-consecutive term limits require the creation of incentives for
presidents to move onto other non-partisan roles and the development of limits
that prevent proxy rule by a president’s family members or close associates.
Part VII briefly discusses three other proposals raised by recent
constitutional design scholarship: the proposal to scrap term limits completely
and rely instead on substitutes; the proposal to focus on popular enforcement as
a way to protect term limits; and the proposal to allow changes to term limits
only on a prospective-only basis such that it does not benefit the incumbent.
While some of these proposals are complementary to our own and have
considerable promise, they all raise problems from the standpoint of
constitutional design.
Finally, this Article concludes by considering the U.S. term limit provision
in light of the arguments developed in this Article. The United States’
presidential two-term limit is highly entrenched because Article V makes the
entire U.S. Constitution extremely difficult to change, and generally gives the
minority party the ability to block that change.
13
However, comparative
experience shows that this very rigidity may increase the incentives of presidents
to find other routes, such as manipulation of the judiciary, to achieve their goals.
Thus, the United States’ presidential term limit is more vulnerable to democratic
erosion than is commonly assumed.
The U.S. Constitution does not deal with the end game problem as
effectively as a weaker, or non-consecutive term limit. This may become a
matter of immediate concern in the United States, given that a myriad of
commentators have noted how the country currently appears to be particularly
vulnerable to democratic erosion.
14
It also suggests that the United States’ two-
term presidential term limit may be a poor model for presidential systems
abroad, despite its popularity (it is one of the most commonly used presidential
term limit found in the world today)
15
and the United States’ continued
reputation for democratic stability.
16
12. See J. L. BLACK, THE RUSSIAN PRESIDENCY OF DMITRY MEDVEDEV, 2008–12: THE NEXT STEP
FORWARD OR MERELY A TIME OUT? 12 (2015); Christian Need & Matthias Schepp, The Puppet President:
Medvedev’s Betrayal of Russian Democracy, SPIEGEL ONLINE (Oct. 4, 2011, 3:51 PM),
https://www.spiegel.de/international/world/the-puppet-president-medvedev-s-betrayal-of-russian-democracy-
a-789767.html.
13. See Huq & Ginsburg, supra note 3, at 143–44.
14. See id. at 165 (arguing that “there is a present danger of constitutional retrogression” in the United
States). See generally CAN IT HAPPEN HERE? AUTHORITARIANISM IN AMERICA (Cass R. Sunstein ed., 2018)
(containing a series of essays on the United States’ vulnerabilities to authoritarianism); STEVEN LEVITSKY &
DANIEL ZIBLATT, HOW DEMOCRACIES DIE (2018) (pointing out various ways in which the United States may be
vulnerable to global pathways of democratic erosion).
15. See Ginsburg et al., supra note 1, at 1836.
16. See infra Table 1.
February 2020] CONSTITUTIONAL END GAMES 365
I. WHY TERM LIMITS? TERM LIMITS AND DEMOCRACY
Presidential term limits have a wide range of defenders and critics.
Defenders of term limits often point to several arguments in favor of such
limits.
17
A. D
EMOCRACY AND THE ADVANTAGES OF TERM LIMITS
One argument in favor of term limits is that their existence may help draw
more people into office. This argument has more force in some contexts than
others (for example, in local and state elections, where there is a greater chance
of participation by ordinary citizens, than say in national elections) and intersects
with arguments for increasing the representation of women and racial minorities
in political elections.
18
Generally, the argument reflects deeper philosophical
commitments to participatory forms of government and decision-making by all
citizens. As such, the argument is sometimes labelled populist in nature.
19
In the
current climate of illiberal populism, it is perhaps better understood as an
argument for more citizen participation in democratic self-government.
A second argument focuses on the behavior of existing representatives and
their tendency to vote in their own narrow self-interest, or that of their
constituents, as opposed to the broader public interest.
20
Term limits can help
reduce pressure on legislators to vote with reelection in mind because it “frees”
them of “career considerations” or eliminates the need to win reelection. Thus,
legislators are given the scope to engage in reasoned deliberation and decision-
making, or “republican” forms of debate and representation.
21
Some political
scientists further suggest that term limits will reduce overall government
expenditures, especially inefficient forms of expenditure designed to ensure the
reelection of specific representatives.
22
17. See Bruce E. Cain & Marc A. Levin, Term Limits, 2 ANN. REV. POL. SCI. 163, 167–72 (1999).
18. See, e.g., Mark P. Petracca, A Legislature in Transition: The California Experience with Term Limits
19 (Inst. Governmental Studies, Working Paper No. 96-19, 1996).
19. See Cain & Levin, supra note 17, at 168–69; Robert Kurfirst, Term-limit Logic: Paradigms and
Paradoxes, 29 POLITY 119, 123–24 (1996).
20. This, of course, reflects more general tensions within democratic theory about the nature of democracy
and the broader role of legislators. See generally RICHARD A. POSNER, LAW, PRAGMATISM, AND DEMOCRACY
(2003); JEREMY WALDRON, LAW AND DISAGREEMENT (1999).
21. See HARVEY C. MANSFIELD, JR., AMERICA’S CONSTITUTIONAL SOUL (1991); GEORGE F. WILL,
RESTORATION: CONGRESS, TERM LIMITS, AND THE RECOVERY OF DELIBERATIVE DEMOCRACY (1992); Cain &
Levin, supra note 17, at 170; Ginsburg et al., supra note 1, at 1822; Kurfirst, supra note 19, at 125–26. For
empirical support for this idea, see, for example, Holger Sieg & Chamna Yoon, Estimating Dynamic Games of
Electoral Competition to Evaluate Term Limits in U.S. Gubernational Elections, 107 AM. ECON. REV. 1824
(2017); Daniel J. Smith et al., Long Live the King? Death as a Term Limit on Executives 8–9 (Feb. 22, 2018)
(unpublished manuscript) (on file with authors).
22. See Timothy Besley & Anne Case, Does Electoral Accountability Affect Economic Policy Choices?
Evidence from Gubernational Term Limits, 110 Q.J. ECONOMICS 769 (1995); Timothy Besley & Anne Case,
Incumbent Behavior: Vote-Seeking, Tax-Setting, and Yardstick Competition, 85 AM. ECON. REV. 25 (1995);
Timothy Besley & Anne Case, Political Institutions and Policy Choices: Evidence from the United States, 41 J.
ECON. LITERATURE 7 (2003). But see López, supra note 11.
366 HASTINGS LAW JOURNAL [Vol. 71:359
The most important modern argument in favor of presidential term limits–
and the one emphasized in this Article—is that term limits have the capacity to
protect democracy and democratic competition by reducing the advantages that
an incumbent possesses in democratic elections and the strength of an
incumbent’s individual personal rule.
23
Incumbents enjoy a range of advantages in democratic elections. For
example, they generally enjoy a stronger reputation or name-recognition among
voters, compared to their competitors. John Lott describes these advantages as
the product of a prior “investment in brand name capital” that is both “sunk” and
“non-transferable,” and that can create an effective barrier to entry by political
challengers.
24
Voters may also be inherently biased toward incumbents because
they may perceive less risk associated with incumbents than with challengers.
25
In addition, incumbents may have greater access to state resources, the
support of the media and interest groups, and the ability to rely on forms of
clientelist or patronage politics to ensure reelection.
26
In hybrid or electoral
authoritarian regimes, incumbents may benefit from the active support of state
media outlets and the ability to use both the civil and criminal law to intimidate
and harass the political opposition and its supporters. The advantages
incumbents enjoy are not the same across all political systems. As Nic
Cheeseman notes, U.S. incumbents often benefit from strong name recognition,
whereas African incumbents rely more heavily on patronage networks.
27
However, there is strong empirical evidence that incumbent legislators and
executive actors enjoy advantages as a result of their incumbency worldwide.
In the United States, the incumbent reelection rate was approximately 97%
in the House of Representatives and 93% in the Senate in 2016.
28
At the state
23. Some political scientists further suggest that this gives incumbents opportunities for “rent extraction.”
Term limits can also potentially limit this kind of rent extraction both directly and indirectly—by reducing
incumbent reelection, and by limiting the timeframe over which incumbents can reach agreements to engage in
log-rolling or rent-sharing behavior. See Barry R. Weingast & William J. Marshall, The Industrial Organization
of Congress; or, Why Legislatures, Like Firms, Are Not Organized as Markets, 96 J. POL. ECON. 132, 138 (1988).
24. John R. Lott, Jr., The Effect of Nontransferable Property Rights on the Efficiency of Political Markets:
Some Evidence, 32 J. PUB. ECON. 231 (1987); see also John R. Lott, Jr., Brand Names and Barriers to Entry in
Political Markets, 51 PUB. CHOICE 87, 89–90 (1986).
25. See M. Daniel Bernhardt & Daniel E. Ingberman, Candidate Reputations and the “Incumbency Effect,”
27 J. PUB. ECON. 47, 49 (1985). The counter-argument is of course that in some elections, voters are looking for
change and will thus, tend to be biased toward challengers.
26. See Nic Cheeseman, African Elections as Vehicles for Change, 21 J. DEMOCRACY 139, 145–46 (2010);
John N. Friedman & Richard T. Holden, The Rising Incumbent Reelection Rate: What’s Gerrymandering Got
to Do with It?, 71 J. POLITICS 593, 596 (2009); Ginsburg et al., supra note 1, at 1820; Maltz, supra note 1, at
131–35.
27. See Cheeseman, supra note 26, at 140, 145–46.
28. Kyle Kondik & Geoffrey Skelley, Incumbent Reelection Rates Higher than Average in 2016,
RASMUSSEN REP. (Dec. 15, 2016), http://www.rasmussenreports.com/public_content/
political_commentary/commentary_by_kyle_kondik/incumbent_reelection_rates_higher_than_average_in_20
16; see also Andrew Gelman & Gary King, Estimating Incumbency Advantage Without Bias, 34 AM. J. POL.
SCI. 1142 (1990).
February 2020] CONSTITUTIONAL END GAMES 367
level, sitting governors enjoyed an 80% reelection rate.
29
For presidents, the
reelection rate was lower: of twenty-six presidents who ran for reelection in a
general election, only sixteen, or 62%, won a second consecutive term.
30
But
this pattern is not replicated in other presidential systems, where incumbents
seem to have larger advantages.
In Latin America, Javier Corrales and Michael Penfold found that, between
1998 and 2006, sitting presidents enjoyed a 90% chance of winning a second
consecutive term, and an 83% chance of subsequent or indefinite reelection,
such that there was effectively a 62.8% increased chance of reelection for
presidential incumbents.
31
Incumbency also affected the margin of victory in
presidential elections; it increased the margin of a president’s victory over their
nearest rival by approximately 11.2%,
32
and was the strongest factor in
predicting both the probability and margin of victory for presidential elections.
33
In Africa, Gideon Maltz likewise found that, in elections between 1992 and
2006, incumbent presidents were re-elected at a rate of 93%.
34
The incumbent
reelection rate is so high in Africa that, when combined with patterns of ongoing
authoritarian rule, only twelve sub-Saharan countries between 1989 and 2010
experienced a change in presidential leadership through democratic elections.
35
Since then, there have been only three notable instances of a change in
presidential leadership as a result of democratic elections in which incumbents
were eligible to run: Nigeria and Zambia in 2015,
36
and Ghana in 2016.
37
As
Cheeseman notes, this is not simply due to the significant number of electoral or
competitive authoritarian systems in Africa.
38
Even when these countries are
excluded, incumbents won reelection in 64% of elections.
39
29. Kondik & Skelley, supra note 28.
30. See Michael Medved, For U.S. Presidents, Odds for a Second Term Are Surprisingly Long, DAILY
BEAST, https://www.thedailybeast.com/for-us-presidents-odds-for-a-second-term-are-surprisingly-long (last
updated July 13, 2017, 1:20 PM) (reporting 15). President Obama is the 16th. See Michael E. Purdy, Only 30%
of U.S. Presidents Served 2 Full Terms, PRESIDENTIAL HISTORY (Jan. 16, 2013),
https://presidentialhistory.com/2013/01/only-30-of-u-s-presidents-served-2-full-terms.html. Of the other 17
previous presidents, 5 died while in office, 7 declined to run for reelection, and 5 failed to gain their party’s
nomination. PRESIDENTS, https://www.whitehouse.gov/about-the-white-house/presidents/ (last visited Jan. 24,
2020).
31. Javier Corrales & Michael Penfold, Manipulating Term Limits in Latin America, 25 J. DEMOCRACY
157, 163 (2014).
32. Id. at 164.
33. See id. at 162–64.
34. Cheeseman, supra note 26, at 139–40; Maltz, supra note 1, at 134.
35. Cheeseman, supra note 26, at 139.
36. See Africa’s 2015 Election Experiences Present Dilemmas for 2016 Polls, CONVERSATION (Jan. 26,
2016, 11:02 PM), https://theconversation.com/africas-2015-election-experiences-present-dilemmas-for-2016-
polls-53312.
37. Yomi Kazeem, Ghana Has Elected Nana Akufo-Addo as Its New President, QUARTZ AFR. (Dec. 9,
2016), https://qz.com/858481/ghana-decides-nana-akufo-addo-has-been-elected-as-ghanas-new-president/.
38. See Cheeseman, supra note 26, at 142.
39. Id.
368 HASTINGS LAW JOURNAL [Vol. 71:359
Term limits can significantly reduce the advantage of incumbency in
democratic elections. In the United States, there is strong evidence that term
limits tend to increase the competitiveness of legislative elections. Kermit
Daniel and John Lott, for example, found that term limits had a strong and
significant effect on a range of measures of competitiveness in state legislative
elections, including who won, who was defeated, the margin between the top
two candidates, and the number of unopposed races.
40
In a global context, there is likewise evidence that term limits promote an
increase in political competition, as well as alternation in individual rule. A
transition in political leadership often weakens the dominant political party in a
competitive authoritarian regime such that it is less able to engage in tactics
designed to undermine true political competition, such as electoral intimidation,
voter registration fraud, and vote tampering.
41
A new leader of a party also
generally has less electoral name recognition and respect than the outgoing
president.
42
New party leaders may even be selected in part because they are not
seen to pose a serious threat to the ongoing power and prestige of the outgoing
president. This gives opposition parties a significantly greater chance of success
in elections against successor candidates than against incumbents.
Two leading examples, highlighted by Maltz, are the changes in political
control of the presidency in Croatia, in 2000, and Kenya, in 2002.
43
In Croatia,
President Franjo Tudjman and his Croatian Democratic Union had been in
power since 1990, but when Tudjman died, his political successor failed to reach
the runoff stage at subsequent presidential elections.
44
In Kenya, President
Daniel Arp Moi ruled from 1978 to 2002.
45
When Moi left office in 2002, at the
end of the formal term limits he agreed to in the 1990s, his party lost control of
the presidency, and Mwa Kibaki was elected to office.
46
Moi’s successor, Uhuru
Kenyatta, did not enjoy the same popular support or appeal as Moi, and many
commentators believe that Moi in fact chose Kenyatta because of this weakness,
understanding that it would make him dependent upon Moi.
47
These patterns are also borne out by quantitative studies of presidential
reelection rates. In Africa, for example, Maltz found that successor candidates
(candidates from the same party as an outgoing president) have a 52% chance of
40. See Kermit Daniel & John R. Lott, Jr., Term Limits and Electoral Competitiveness: Evidence from
California’s State Legislative Races, 90 PUB. CHOICE 165, 181 tbl.7 (1997).
41. Maltz, supra note 1, at 133–34.
42. See Joel Lieske, The Political Dynamics of Urban Voting Behavior, 33 AM. J. POL. SCI. 150, 168
(1989).
43. Maltz, supra note 1, at 131–32.
44. Id. at 131.
45. Id.
46. See id.
47. See id. at 132.
February 2020] CONSTITUTIONAL END GAMES 369
a successful election, compared to a 93% reelection rate for incumbents.
48
In
competitive democracies, the figure is 50%, compared to 64% for incumbents.
49
The mere alternation in individual presidential rule can help protect
democracy, even where the alternation occurs within an existing party.
Libertarian arguments for term limits, for example, focus on the capacity of term
limits to weaken the power of the legislative or executive branch, and thus,
promote commitments to limited government and individual liberty.
50
These
arguments overlap with democratic arguments for term limits, or at least
executive term limits.
A common hallmark of authoritarian government is an extremely strong
executive branch that has a tradition of personalist presidential rule. Guarding
against the danger of authoritarianism, therefore, will generally require limiting
the powers of the executive branch, especially if there are individual executive
leaders.
51
Term limits are an obvious way to do this because they force a
dominant president to leave office in ways that reduce the informal power of the
presidency. Term limits also undermine the network of clientelist and patronage
relationships that help sustain electoral authoritarian systems.
52
The spread of
presidential term limits in both Africa and Latin America in recent decades
reflects this logic.
53
As John Carey notes, these prohibitions have been
“motivated both by theory and by experiences of individual politicians who
endeavored to entrench themselves in power.”
54
B. O
BJECTIONS AND DISADVANTAGES
Critics of term limits, on the other hand, suggest that they tend to
undermine democracy because they deprive institutions of the professional
expertise and experience needed for success, undermine the incentives and
accountability of elected officials, and deny voters the opportunity to re-elect
their preferred representative.
55
These arguments also have a long lineage—
48. Id. at 134.
49. Cheeseman, supra note 26, at 142.
50. See Cain & Levin, supra note 17, at 171; Kurfirst, supra note 19, at 126–27.
51. See Baker, supra note 8, at 288–89; Maltz, supra note 1, at 130.
52. See Baker, supra note 8, at 288–89, 297–98; Cheeseman, supra note 26, at 150–51; Maltz, supra note
1, at 136–37; Milanovic et al., supra note 11.
53. See John M. Carey, The Reelection Debate in Latin America, 45 LATIN AM. POL. & SOC’Y 119, 122,
127 (2003); see also Ginsburg et al., supra note 1; Maltz, supra note 1, at 131; Denis M. Tull & Claudia Simons,
The Institutionalization of Power Revisited: Presidential Term Limits in Africa, 52 AFR. SPECTRUM 79, 82
(2017).
54. Carey, supra note 53, at 122. Simon Bolivar made an earlier argument for term limits in Latin America
on this basis. See Ginsburg et al., supra note 1, at 1819–20. Note, however, that Bolivar ultimately reversed his
position. See Carey, supra note 53, at 121–22; Ginsburg et al., supra note 1, at 1819.
55. See Cain & Levin, supra note 17, at 182–84; Ginsburg et al., supra note 1, at 1824; Mark P. Petracca,
Why Political Scientists Oppose Term Limits (Feb. 18, 1992) (unpublished briefing paper) (on file with the Cato
Institute) (addressing professionalism).
370 HASTINGS LAW JOURNAL [Vol. 71:359
many of them were made by Alexander Hamilton in Federalist 72.
56
Some
scholarship credit Hamilton’s arguments for defeating proposals to include
presidential term limits in the original U.S. Constitution.
57
Concerns about expertise and electoral incentives, discussed in Part V, can
be addressed by appropriately generous and flexible forms of term limits. The
concern about democracy seems to overlook two key arguments. First, term
limits can play a role in preventing a slide toward authoritarianism or democratic
regression. Second, there are democratic procedural arguments in favor of such
limits (for example, the strong degree of popular support for the enactment of
term limits, and the fact that many term limit provisions are introduced by way
of an amendment proposed or ratified by voters at referenda).
In Africa, for example, public opinion polling by Afrobarometer between
2011 and 2013 found that 75% of voters across thirty-four countries favored a
two-term limit for presidents.
58
Ed Glaeser suggests that this is consistent with
risk aversion among democratic voters due to a preference for “cycling of
ideologies rather than locking into a single ideology,” or a preference for a
degree of ongoing political or ideological alternation.
59
Some scholars further suggest that, at least under certain conditions, term
limits may help promote democratic choice. Term limits solve a coordination
problem among voters in different parties who wish to see, in addition to a norm
of alternation in office, candidates from their own party succeed in elections.
60
We do not suggest that democratic concerns necessarily favor term limits
in all contexts and for all institutions. The strength of the executive in fragile
democracies, for example, is itself often the product of a history of weak
legislatures incapable of imposing any meaningful checks on the executive
branch. A key means of checking the potential for abuse by the legislature will
thus be to increase the power and independence of the legislature. A stronger
legislature will often impose term limits on the executive.
61
Term limits, as
Robert Kurfirst notes, have an important impact not only on the absolute power
of institutions, but also on their relative power and standing.
62
56. See THE FEDERALIST NO. 72 (Alexander Hamilton) (Yale Univ. Press ed., 2009).
57. See Carey, supra note 53, at 120–21.
58. See BONIFACE DULANI, AFRICAN PUBLICS STRONGLY SUPPORT TERM LIMITS, RESIST LEADERS’
EFFORTS TO EXTEND THEIR TENURE 3 fig.1 (2015); see also Adrienne LeBas, Term Limits and Beyond: Africa’s
Democratic Hurdles, 115 CURRENT HIST. 169, 170 (2016) (explaining that Africa is paying heightened attention
to the idea of presidential term limits).
59. See Edward L. Glaeser, Self-Imposed Term Limits, 93 PUB. CHOICE 389, 390 (1997).
60. See, e.g., Andrew R. Dick & John R. Lott, Jr., Reconciling Voters’ Behavior with Legislative Term
Limits, 50 J. PUB. ECON. 1, 8 (1993); Einer Elhauge, Are Term Limits Undemocratic?, 64 U. CHI. L. REV. 83,
85–86 (1997); Glaeser, supra note 59, at 389–90.
61. Some economists likewise suggest that it is important to weaken personal rule so as to promote the rule
of law—this encourages private actors to invest in general rule of law protections, rather than in developing
“clientilistic” relationships with individual political leaders. See Milanovic et al., supra note 11, at 3–4.
62. See Kurfirst, supra note 19, at 129–34. For the argument that term limits on executive and legislative
officials may in fact be symbiotic or complementary in this context, see Michael J. Malbin & Gerald Benjamin,
February 2020] CONSTITUTIONAL END GAMES 371
Thus, we focus here on presidential term limits. Some countries may
choose to adopt legislative term limits as an additional tool for promoting good
government. However, the case for legislative term limits is more contingent,
especially in cases where they weaken the power of the legislature and, thus,
create an imbalance in the separation of powers.
63
II.
MODELS OF PRESIDENTIAL TERM LIMITS
In constitutional design, arguments in favor of presidential term limits have
won the day in the modern period. However, historically, this tradeoff was not
always resolved in favor of presidential term limits, and term limits have not
always been a standard part of constitutional design. Many early national
constitutions, such as the United States Constitution of 1787 and the French
Constitution of 1791, did not include them.
64
The U.S. Constitution instead
adopted an informal two-term limit, which lasted until President Franklin
Delano Roosevelt ran for, and won, four consecutive terms in the 1930s and
1940s.
65
Following his death, the United States formalized a presidential two-
term limit in the 22nd Amendment, which was adopted in 1951.
66
In recent times, the adoption of presidential term limits has become the
overwhelming design choice in presidential and semi-presidential systems.
67
Public opinion data shows that presidential term limits are generally very
popular.
68
Even where individual presidents are popular, voters do not want
them to remain in office forever. In contrast, term limits for other kinds of actors
are less common. For example, a much smaller number of countries have term
limits for members of the legislature.
69
Similarly, term limits for subnational
officials, such as governors, appear to be less common than for presidents.
70
Table 1, based on our own calculations, presents term limits in all presidential
and semi-presidential systems as of 2019. The most immediate point is that most
systems have presidential term limits—only 16% lack them. Moreover, the
characteristics of these countries support an argument that the absence of
presidential term limits tends to erode constitutional democracy. Most of these
countries, such as Azerbaijan, Belarus, Kazakhstan, Cameroon, Nicaragua, and
Legislatures After Term Limits, in LIMITING LEGISLATIVE TERMS 209, 220 (Gerald Benjamin & Michael J.
Malbin eds., 1992).
63. In this sense, we take the opposite approach to much of the literature, which tends to be more heavily
focused on legislative as opposed to executive term limits. See López, supra note 11, at 2.
64. As noted above, Hamilton, in the Federalist Papers, argued against them. See THE FEDERALIST NO. 72,
supra note 56.
65. See Peter Feuerherd, How FDR’s Presidency Inspired Term Limits, DAILY JSTOR (Apr. 12, 2018),
https://daily.jstor.org/how-fdrs-presidency-inspired-term-limits/.
66. Id.
67. See ALEXANDER BATURO, DEMOCRACY, DICTATORSHIP, AND TERM LIMITS 32 (2014).
68. See Dulani, supra note 58.
69. See VENICE COMMISSION, REPORT ON TERM LIMITS (2019) [hereinafter REPORT ON TERM LIMITS]
(exemplifying the ongoing debate about whether the legislature should be subject to term limits).
70. See id.
372 HASTINGS LAW JOURNAL [Vol. 71:359
Venezuela, are ones that observers have argued are hybrid or pure authoritarian
regimes.
71
Table 1: Presidential Term Limits in Presidential and Semi-Presidential
Systems, 2019
Type of Term Limit Percent of Countries
No reelection allowed 8%
Bar on consecutive term after one term 8%
Bar on consecutive term after two terms 7%
Absolute bar after two terms 57%
Ambiguous whether absolute or consecutive bar
after two terms
4%
No term limit 16%
*NOTE: SOURCE: AUTHORS’ CALCULATIONS FROM CONSTITUTIONAL TEXTS.
Initially, constitutions that now lack term limits often included them.
72
But
in order to remain in power, incumbents used constitutional amendments,
through the use of referenda, or other devices such as judicial decisions, to
remove the term limits.
73
The argument in favor of the removal of term limits
generally stressed the importance of continuity in exceptional circumstances.
74
For example, in Venezuela, a 1999 Constituent Assembly dominated by
President Chavez stretched the presidential term limit from one term to two
terms.
75
Later, during Chavez’s second full term, he called for two referenda,
one in 2007 and another in 2009, to remove presidential term limits.
76
The first
effort narrowly failed, but the second succeeded.
77
The argument from Chavez
and his allies was essentially that reelection was a regrettable necessity to keep
71. See Nicolas Cherry, The Abolition of Presidential Term Limits in Nicaragua: The Rise of Nicaragua’s
Next Dictator?, 2 CORNELL INT’L L.J. ONLINE 31 (2014) (discussing Nicaragua); Maltz, supra note 1, at 130,
128, 135 (discussing Azerbaijan, Belarus, Kazakhstan and Venezuela); Cheryl Hendricks & Gabriel Ngah
Kiven, Cameroon Presidential Poll Underscores the Need for Term Limits, CONVERSATION (Oct. 8, 2018, 11:19
AM), https://theconversation.com/cameroon-presidential-poll-underscores-the-need-for-term-limits-104583
(discussing Cameroon).
72. See Ginsburg et al., supra note 1, 1835–36; Cheryl Hendricks & Gabriel Ngah Kiven, Presidential
Term Limits: Slippery Slope Back to Authoritarianism in Africa, CONVERSATION (May 17, 2018, 8:44 AM),
https://theconversation.com/presidential-term-limits-slippery-slope-back-to-authoritarianism-in-africa-96796;
Boniface Madalitso Dulani, Personal Rule and Presidential Term Limits in Africa (2011) (unpublished Ph.D
dissertation, Michigan State University) (on file with Michigan State University Library).
73. See, e.g., REPORT ON TERM LIMITS, supra note 69, at 14 (prohibiting use of referenda to override a
constitutional amendment to instate term limits, if adopted); Ginsburg et al., supra note 1, at 1810, 1812, 1847–
48.
74. See, e.g., Ginsburg et al., supra note 1, 1823–27.
75. See Venezuela’s Chavez Era 1958–2013, COUNCIL ON FOREIGN RELS.,
https://www.cfr.org/timeline/venezuelas-chavez-era (last visited Jan. 24, 2020).
76. Id.
77. Id.
February 2020] CONSTITUTIONAL END GAMES 373
his project on track, and the sweeping transformations that his regime was
carrying out could not be entrusted to anyone else.
78
Some countries, about 8% of presidential and semi-presidential systems,
take the strictest possible stance towards presidential term limits and prohibit
any presidential reelection once an incumbent has served a single term in
office.
79
The position taken by these countries obviously stresses the risks of
reelection to democracy, even at the cost of foregoing the expertise and
incentives that reelection may promote.
Of course, the effect of a strict one-term limit will also depend on other
formal and informal aspects of the constitution. The most obvious interaction
here is the length of presidential terms. Globally, presidential terms appear to
vary between four and seven years, and term lengths closer to the latter end of
the spectrum will give presidents more space to pursue their agendas than the
former.
In Mexico, for example, presidents can serve only one six-year term in their
lives.
80
The principle of no reelection is a defining principle of Mexican politics,
even during its lengthy one-party dictatorship throughout most of the 20th
century.
81
In some ways, a lengthy term counterbalances the lack of reelection.
As a result, Mexico has had a series of highly consequential administrations
during the dictatorship, and during and after its transition to democracy.
82
Colombia, in contrast, historically allowed only one four-year term in a
president’s lifetime.
83
In 2005, President Alvaro Uribe amended the constitution
to allow two consecutive terms and won reelection to a second term.
84
However,
his subsequent attempt to amend the constitution again to allow three straight
terms was blocked by the courts.
85
As a result, both Uribe and his successor,
Juan Manuel Santos, were highly consequential presidents who served two terms
each.
86
The former pursued a strategy of “democratic security” to repress the
78. See David Landau, Constitution-Making and Authoritarianism in Venezuela: The First Time as
Tragedy, the Second as Farce, in CONSTITUTIONAL DEMOCRACY IN CRISIS? 161–63 (Mark A. Graber et al. eds.,
2018).
79. See supra Table 1.
80. Constitución Política de los Estados Unidos Mexicanos, CP, Arts. 82–83, Diario Oficial de la
Federacíon [DOF] 05-02-1917, últimas reformas DOF 09-08-2019 (Mex.).
81. See Jeffrey Weldon, The Political Sources of Presidencialismo in Mexico, in PRESIDENTIALISM AND
DEMOCRACY IN LATIN AMERICA 225 (Scott Mainwaring & Matthew Soberg Shugart eds., 1997).
82. See EMILY EDMONDS-POLI & DAVID A. SHIRK, CONTEMPORARY MEXICAN POLITICS 49–96 (3d ed.
2016).
83. See Rosalind Dixon & David Landau, Transnational Constitutionalism and a Limited Doctrine of
Unconstitutional Constitutional Amendment, 13 INT’L J. CONST. L. 606, 615 (2015).
84. See id. at 615–16.
85. Id. at 616.
86. See, e.g., Alia M. Matanock & Miguel García-Sánchez, The Colombian Paradox: Peace Processes,
Elite Divisions and Popular Plebiscites, DAEDALUS J. AM. ACAD. ARTS & SCI. 152 (2017); Jeremy McDermott,
How President Alvaro Uribe Changed Colombia, BBC NEWS (Aug. 4, 2010),
https://www.bbc.com/news/world-latin-america-10841425; Maria Alejandra Silva, Alvaro Uribe: The Most
Dangerous Man in Colombian Politics, COUNCIL ON HEMISPHERIC AFF. (Oct. 20, 2017),
374 HASTINGS LAW JOURNAL [Vol. 71:359
FARC guerilla movement, and the latter sought a peace agreement that took
many years to pursue.
87
Santos, however, re-imposed a one-term, four-year limit
during his second term in office, and it remains to be seen how this will impact
the strength of future administrations.
88
Most presidential and semi-presidential systems around the world take an
intermediate position, balancing the benefits and costs of presidential term
limits. The most common design, implemented by over half of all presidential
and semi-presidential systems, prohibits any reelection after two consecutive
terms in office.
89
This absolute bar on reelection after two terms appears to be
aimed at balancing different aspects of the tradeoff. An absolute two-term limit
allows voters to reward good performance and punish bad performance.
Moreover, successful programs can continue for eight or more years, depending
on the length of presidential term, giving presidents considerable time to develop
their policies. Nonetheless, the term limit forces turnover in office and prevents
incumbent presidents from amassing too much power.
An alternative design choice, one we argue for in this Article, bars
consecutive reelection.
90
A system that bars consecutive reelection allows
presidents to serve one or two terms in office consecutively, but then require that
they sit out for one or more terms before running for election again.
91
Chile, for
example, bars consecutive reelection from presidents after serving one four-year
term, but allows them to return to power after sitting out one term.
92
Former
president Michele Bachelet, for example, served between 2006 and 2010, and
again from 2014 until 2018.
93
http://www.coha.org/alvaro-uribe-the-most-dangerous-man-in-colombian-politics/; Lally Weymouth, An
Interview with Colombian President Juan Manuel Santos, WASH. POST (Aug. 7, 2014),
https://www.washingtonpost.com/opinions/an-interview-with-colombian-presidentjuan-manuel-
santos/2014/08/07/3eaff428-1ced-11e4-82f9-2cd6fa8da5c4_story.html.
87. See Matanock & García-Sánchez, supra note 86, at 155; McDermott, supra note 86.
88. See Elizabeth Reyes L., Colombian Lawmakers Approve a One-Term Limit for Presidents, EL PAÍS
(June 4, 2015, 3:59 PM), https://elpais.com/elpais/2015/06/04/inenglish/1433416990_898964.html.
89. See supra Table 1.
90. Ginsburg, Melton, and Elkins find that the most common design of all presidential term limits
historically (as opposed to just those in force today) allowed consecutive reelection after one term in office—
27% of all constitutions used this model, according to their data. See Ginsburg et al., supra note 1, at 1836.
91. As noted in Table 1, there is a small group of countries where it is clear from the constitutional text
that presidents can serve only two consecutive terms in office, but it is unclear whether they are only temporarily
or permanently barred after serving two terms. Taiwan, for example, recently had some controversy about
whether ex two-term President Ma Ying-Jeou could seek a third term after sitting out one term. See Brian Hioe,
Claims by SCMP that Ma Can Run for a Third Presidential Term Are Ludicrous, NEW BLOOM (May 10, 2018),
https://newbloommag.net/2018/05/10/scmp-ma-third-term/. In Malawi, the Supreme Court, in 2009, held that
an ambiguous term limit should be interpreted to bar a potential nonconsecutive third term by former President
Bakili Muluzi. See Muluzi Denied Slot in Malawi Election, UPI (May 16, 2009, 12:50 PM),
https://www.upi.com/Top_News/2009/05/16/Muluzi-denied-slot-in-Malawi-election/88971242492659/?ur3=1.
92. CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE CHILE [C.P.] art. 25.
93. See Ernesto Londoño, President Bachelet of Chile Is the Last Woman Standing in the Americas, N.Y.
TIMES (July 24, 2017), https://www.nytimes.com/2017/07/24/world/americas/michelle-bachelet-president-of-
chile.html.
February 2020] CONSTITUTIONAL END GAMES 375
Until this year, the island country of Comoros allowed presidents to serve
unlimited, non-consecutive five-year terms.
94
Current President Azali
Assoumani previously served as president between 1999 and 2002, as well as
between 2002 and 2006.
95
Brazil requires that presidents sit out for at least one
term after serving two consecutive terms in office.
96
President Luiz Inacio Lula
da Silva (Lula), for example, served as president for two terms between 2003
and 2010, and was planning to run again for a second term in 2019, before being
jailed on corruption charges.
97
The bar against consecutive reelection balances the tradeoff involved in
presidential term limits in a somewhat different way than an absolute bar.
Forcing presidents to leave power may help to break excessive consolidation of
power—even if presidents are able to return later—but still gives presidents
incentives towards electoral accountability, since they may seek a new term in
the future. A bar against consecutive reelection also gives presidents more than
one term to pursue their projects, especially when they are allowed to pursue
two consecutive terms before leaving power.
As we emphasize in Part V, a design that stresses consecutive, instead of
absolute, bans on reelection may reduce the inclination of incumbents to evade
term limits by lessening the costs of compliance, thus reducing what we call the
“end game” problem.
Notably, only one system with any presidential term limits appears to allow
more than two consecutive terms in office. In the Republic of the Congo, a 2015
referendum amended the constitution to allow three consecutive four-year terms
in office.
98
The apparent lack of countries that follow a similar system reflects a
fairly broad consensus, at least at the level of constitutional design, that allowing
more than two consecutive terms in office poses threats to constitutional
democracy that outweigh any gains in electoral accountability and continuity in
policy. Despite this consensus, a number of presidents have managed to use
devices, such as temporary constitutional provisions or judicial interpretation, to
circumvent their term limits and remain in power.
99
94. CONSTITUTION DE L’UNION DES COMOROS [CONSTITUTION] Mar. 24, 2001, art. 13. The new system
allows two consecutive terms. See Comoros Islanders Vote in Presidential Election, BBC NEWS (Mar. 24, 2019),
https://www.bbc.com/news/world-africa-47685991?intlink_from_url=https://www.bbc.com/news/topics/cx1m
7zg0gnlt/comoros&link_location=live-reporting-story.
95. See Former Coup Leader Wins Presidential Race in Comoros, VOA NEWS (Apr. 15, 2016, 8:43 PM),
https://www.voanews.com/africa/former-coup-leader-wins-presidential-race-comoros.
96. CONSTITUIҪÃO FEDERAL [C.F.] [CONSTITUTION] Dec. 2017, art. 14 (Braz.).
97. See Dom Phillips, Brazil’s Lula Launches Presidential Bid from Jail as Thousands March in Support,
GUARDIAN (Aug. 15, 2018, 6:48 PM), https://www.theguardian.com/world/2018/aug/15/brazil-lula-
presidential-election-campaign-jail.
98. See Philon Bondenga, Congo Votes by Landslide to Allow Third Presidential Term, REUTERS (Oct. 27,
2015, 12:13 AM), https://www.reuters.com/article/us-congo-politics/congo-votes-by-landslide-to-allow-third-
presidential-term-idUSKCN0SL0JW20151027.
99. A very small number of systems combine the logic of consecutive and absolute prohibitions on
reelection. These systems generally require presidents to sit out after serving one term in office, but also limit
376 HASTINGS LAW JOURNAL [Vol. 71:359
In summary, while the vast majority of presidential and semi-presidential
systems have some kind of term limit to prevent the erosion of democracy,
almost all systems allow presidents to serve more than one term. This reflects
the value that constitutional designers place significant value on other goals,
such as accountability, efficiency, and continuity in public policy.
III.
THE PROBLEM OF COMPLIANCE
There is a strong positive correlation worldwide between the increasing
constitutional entrenchment of presidential term limits and the tendency of
presidents to leave office “voluntarily” as part of a peaceful democratic
transition, rather than through a military coup.
In Africa especially, Daniel Posner and Daniel Young note that many
countries introduced new, entrenched constitutional term limits from the 1990s
onward: thirty-two out of thirty-eight constitutions between 1990 and 2005
adopted or entrenched such limits.
100
During this period, there was a marked
increase in the number of leaders who left “voluntarily,” rather than through
coups, assassination, or other involuntary means. Between 2000 and 2005, only
19% of leaders left power involuntarily, whereas 70–75% did so in the 1960s,
1970s, and 1980s.
101
Of those seventeen leaders that left office “voluntarily”
between 2000 and 2005, nine also departed when they reached the end of their
presidential term limits.
102
The overall pattern globally, however, is not quite so positive. Rather, it
involves a significant amount of non-compliance with presidential term
limits.
103
Tom Ginsburg, James Melton, and Zachary Elkins note that, of the 352
cases in their dataset where presidents had the opportunity to over-stay (in other
words, did not depart early from office), 89 involved an attempt by presidents to
stay beyond their constitutionally permitted term, and 71 of those attempts were
successful.
104
Of these 71 cases, 56 involved formal constitutional amendment
or replacement, 29 and 27 respectively, five involved the suspension of the
constitution, and 10 involved the circumvention or disregard of constitutional
the total number of terms they can serve in their lives. For example, Haiti limits presidents to serving one
consecutive five-year term; they can return to power after sitting out one term, but are limited to serving only
two terms in their lives. LA CONSTITUTION DE LA RÉPUBLIQUE D’HAITI [CONSTITUTION] 1987, art. 134–3 (“The
President of the Republic may not be re-elected. He may serve an additional term only after an interval of five
(5) years. He may in no case run for a third term.”).
100. See Daniel N. Posner & Daniel J. Young, The Institutionalization of Political Power in Africa, 18 J.
DEMOCRACY 126, 132 fig. 3 (2007).
101. See id. at 128 fig.1, 128–29.
102. Id. at 129.
103. For skepticism about compliance, see Daron Acemoglu et al., A Political Theory of Populism, 18 Q.J.
ECONOMICS 771, 792 (2013); Baker, supra note 8, at 285; Javier Corrales & Michael Penfold, Manipulating
Term Limits in Latin America, 25 J. DEMOCRACY 157, 158 (2014); Ginsburg et al., supra note 1, 1847–50; Maltz,
supra note 1, at 129–30; Smith et al., supra note 21, at 6–7.
104. Ginsburg et al., supra note 1, at 1848–49.
February 2020] CONSTITUTIONAL END GAMES 377
limitations.
105
Moreover, 15 of these “constitutional overstaying” attempts
occurred in countries that were constitutional democracies at the time.
106
Some
of these occurred in the 19th
and early 20th centuries: Costa Rica in 1876,
107
as
well as in 1898;
108
Colombia in 1886;
109
Uruguay in 1935;
110
and Honduras in
1936.
111
Others were more recent, including the Philippines in 1973, Peru in
1995, Argentina in 1995, Brazil in 1998, Venezuela in 2004, Colombia in 2006,
Niger in 2009, and Belarus in 2006.
112
Gideon Maltz likewise notes twenty-six instances in which presidents
chose not to comply with relevant term limits between 1992 and 2006: fourteen
of these instances involved the amendment or repeal of relevant limits, while
twelve involved some form of constitutional “over-staying” or breach of the
relevant limit.
113
Six of these countries were constitutional democracies at the
time of the relevant constitutional non-compliance.
114
In Africa, Posner and Young note the amendment or repeal of
constitutional term limits between 1990 and 2005 in Chad, Gabon, Guinea,
Namibia, Togo and Uganda.
115
The incumbent president in each case ultimately
won the relevant election, such that these changes led to a third consecutive
presidential term for Presidents Idriss Débby of Chad, Omar Bongo of Gabon,
Lansana Conté of Guinea, Samuel Nujoma of Namibia, Gnassingbé Eyadéma of
Togo, and Yoweri Museveni of Uganda.
116
Denis Tull and Claudia Simons extended this analysis to 2016 and found
that, of the thirty-nine presidents in Africa who reached the end of their
constitutionally permitted time in office, eighteen chose not to comply with
relevant term limits, but instead chose to circumvent or to formally amend those
limits.
117
Attempts at formal constitutional change were met with an extremely
105. Id. at 1849.
106. Id.
107. Tomas Guardia Gutierrez, where there was a single term limit with non-immediate reelection not
permitted. Ginsburg et al., supra note 1, at 1851–52.
108. Rafael Yglesias Castro, where there was a single term limit, with non-immediate reelection not
permitted. Id. at 1852.
109. Rafael Nunez, where there was a single term limit, with non-immediate reelection permitted after a full
two-year term elapsed. Id. at 1850 n.213.
110. Gabriel Terra, where there was a single term limit, with non-immediate reelection permitted: two terms
must elapse before the executive can be re-elected. See id. at 1869 tbl.A1.
111. Carias Andino, where there was a single term limit, with non-immediate reelection permitted. See id.
112. See id. at 1869–72.
113. Maltz, supra note 1, at 128.
114. See id. at 130 fig.1 (listing Argentina, Brazil, the Dominican Republic, Namibia, Romania and
Venezuela).
115. See Posner & Young, supra note 100, at 132–34. In Guinea, the relevant amendment also repealed all
term limits, thereby allowing President Lansana Conté the possibility of indefinite reelection. See Baker, supra
note 8, at 291–92.
116. Posner & Young, supra note 100, at 133–34. On how changes in Guinea and Burkina Faso have led to
the scope for indefinite reelection in those countries, see Baker, supra note 8, at 291–92.
117. See Tull & Simons, supra note 53, at 83–85.
378 HASTINGS LAW JOURNAL [Vol. 71:359
high rate of success. For instance, out of the eighteen presidential attempts to
formally amend term limits, fifteen succeeded, compared to only three cases of
failure.
118
Perhaps the most comprehensive study of term limits evasion to date has
been carried out by Mila Versteeg and several coauthors, in which they studied
106 countries.
119
Since 2000, of 234 constitutionally-required presidential exits
from office, there were 60 cases of attempted evasion of existing term limits.
120
Thus, like Ginsburg and Elkins, they find that attempted term limits evasion
occurs about 25% of the time.
121
By far the most common tool to evade term limits is formal amendment to
the constitution, which removes or loosens the term limit.
122
But this is not the
only tool that incumbent presidents wishing to stay in power possess. The study
also outlines other techniques that are used fairly commonly. One is to replace
the entire constitution, which often has the effect of resetting the clock on term
limits.
123
Another is to go to the judiciary and to convince it to reinterpret the
term limit or to throw it out entirely.
124
These latter techniques may actually be
more damaging to the rule of law than constitutional amendment, since they may
have collateral costs to stability or to judicial independence.
Consider the problem through a number of recent cases, drawn first from
Africa. In 2008, the parliament of Cameroon voted to amend its constitution to
remove all presidential term limits, thereby allowing President Paul Biya to
extend his term beyond the twenty-five years he had already served.
125
Similarly,
in 2012 in Senegal, supporters of President Abdoulaye Wade successfully
proposed amending the constitution to allow him to run for a third term, despite
significant public protests.
126
The only silver lining from a democratic
perspective was that the amendment process galvanized a new grass-roots
political opposition movement that defeated Wade’s actual bid for reelection.
127
In 2010, the Djibouti parliament voted to remove presidential term limits
from its constitution, to shorten the presidential term to five years, and to impose
118. See id. at 85.
119. Versteeg et al., supra note 4.
120. See id.
121. See supra text accompanying note 90.
122. See Versteeg et al., supra note 4.
123. See id.
124. See id.
125. LeBas, supra note 58, at 171; Cameroon Parliament Extends Biya’s Term Limit, FR.24,
http://www.france24.com/en/20080411-cameroon-parliament-paul-biya-term-limit-extension (last updated
Nov. 4, 2008, 4:06 PM); Will Ross, Cameroon Makes Way for a King, BBC NEWS,
http://news.bbc.co.uk/2/hi/africa/7341358.stm (last updated Apr. 11, 2008, 7:08 AM).
126. See Lamin Jahateh, Controversy of Abdoulaye Wade’s Presidential Bid, AL JAZEERA (Jan. 28, 2012),
https://www.aljazeera.com/indepth/opinion/2012/01/201212712295177724.html.
127. See LeBas, supra note 58, at 171.
February 2020] CONSTITUTIONAL END GAMES 379
a mandatory retirement age of seventy-five for the president.
128
These
amendments paved the way for President Ismael Omar Guelleh to stand for
reelection for a third term in 2011 and a fourth term in 2016.
129
Guelleh was
ultimately re-elected in 2016 with 87% of the vote, against a backdrop of
significant alleged political repression and electoral irregularities.
130
In 2015, the Rwandan parliament passed a constitutional amendment,
which was then approved at a national referendum, to reduce presidential term
limits from seven to five years.
131
In addition, the Rwandan parliament also
created a set of “transitional” arrangements that allowed the winner of the 2017
presidential election to serve an initial transitional seven-year term, and
subsequently be eligible for two additional five-year terms.
132
In aggregate, these
changes created the possibility for President Paul Kagame to stay in office for a
further seventeen years, until 2034.
133
In in the lead up to the 2020 Burundi presidential elections, President
Nkurunziza, who had already served for three terms, proposed changes to the
Burundi constitution to allow him to seek reelection for two more consecutive
terms.
134
Following his reelection to a third term in 2015, he appointed a
commission to consider the possibility of further constitutional amendments.
135
After a process of public consultation, the commission announced its
recommendation to extend presidential term limits from five to seven years.
136
The commission also recommended various parallel changes to the power of the
presidency.
137
Despite criticisms of the commission, and its processes, the
128. See MPs in Djibouti Scrap Term Limits, BBC News, http://news.bbc.co.uk/2/hi/africa/8630616.stm
(last updated Apr. 19, 2010, 5:57 PM).
129. See Djibouti President Ismail Omar Guelleh Wins Fourth Term, BBC NEWS (Apr. 9, 2016),
http://www.bbc.com/news/world-africa-35995628.
130. Id.
131. See Rwanda Changes Constitution to Allow President to Extend His Rule Until 2034, ABC NEWS,
http://www.abc.net.au/news/2015-12-20/rwanda-changes-constitution-to-allow-president-to-stay-in-
power/7043698 (last updated Dec. 19, 2015, 5:18 PM); Rwandans Vote on Allowing Third Kagame Presidential
Term, BBC NEWS (Dec. 18, 2015), http://www.bbc.com/news/world-africa-35125690.
132. See Rwanda Changes Constitution to Allow President to Extend His Rule Until 2034, supra note 131;
Rwandans Vote on Allowing Third Kagame Presidential Term, supra note 131.
133. Rwanda Changes Constitution to Allow President to Extend His Rule Until 2034, supra note 131;
Claudine Vidal, Rwanda: Paul Kagame Is in Line to Stay in Office Until 2034, CONVERSATION (Jan. 18, 2016,
7:17 AM), https://theconversation.com/rwanda-paul-kagame-is-in-line-to-stay-in-office-until-2034-53257.
134. See Will the 2020 Elections in Burundi Be Bloody?, AFR. NEWS (July 29, 2019, 10:03 AM),
https://www.africanews.com/2019/07/29/will-the-2020-elections-in-burundi-be-bloody/.
135. See Mohammed Yusuf, Burundi Parliament to Review Plan on Scrapping Term Limits, VOA NEWS
(Aug. 29, 2016, 2:12 PM), https://www.voanews.com/a/burundi-parliament-term-limits/3485090.html.
136. Burundi Backs New Constitution Extending Presidential Term Limits, AL JAZEERA (May 22, 2018),
https://www.aljazeera.com/news/africa/2018/05/burundi-backs-constitution-extending-presidential-term-
limits-180521134736408.html.
137. Yolande Bouka & Sarah Jackson, Burundi Votes Tomorrow on Controversial Constitutional
Amendments. A Lot Is at Stake, WASH. POST (May 16, 2018, 6:00 AM),
https://www.washingtonpost.com/news/monkey-cage/wp/2018/05/16/burundi-votes-tomorrow-on-
controversial-constitutional-amendments-a-lot-is-at-stake/.
380 HASTINGS LAW JOURNAL [Vol. 71:359
proposed changes were subsequently approved by voters at a national
referendum in 2018.
138
Another method used to circumvent presidential term limits in Africa is
through the use of the judiciary. For example, in Burundi, President Pierre
Nkurunziza’s party, in 2015, asked the Constitutional Court of Burundi to find
that the existing term limits did not apply to President Nkurunziza because he
was elected under transitional provisions that provided for indirect,
parliamentary election rather than direct elections.
139
The court upheld the
argument, finding that the transitional provisions operated separately from the
provisions imposing presidential term limits.
140
The court’s decision ultimately
paved the way for President Nkurunziza to be re-elected for a third term.
141
A similar pattern of evasion applies in Latin America.
142
Between 1993 and
2009, Argentina, Brazil, and Colombia all passed formal constitutional changes
that relaxed constitutional term limits and allowed some scope for presidential
reelection.
143
Ecuador and Venezuela formally repealed term limits altogether
so presidents could seek indefinite reelection.
144
In both countries, the changes
were passed despite a tiered constitutional design that required a more
demanding standard when attempting to alter the fundamental structure of the
constitution.
145
In Ecuador, for example, changes to the “fundamental structure,
or the nature and constituent elements of the State” require a referendum,
whereas most other changes can be carried out by Parliament alone.
146
Despite
compelling arguments that the elimination of all term limits would be the type
of fundamental change that would require a more demanding procedure,
147
138. See id. (noting criticisms of the process, and its selectivity, by the Burundi Forum for Strengthening
Civil Society). The referendum itself was also critiqued as tainted by voter intimidation and a lack of
transparency. See Abdi Latif Dahir, Burundi Has Backed Constitutional Changes that Could See Its President
Rule till 2034, QUARTZ AFR. (May 21, 2018), https://qz.com/1284514/burundi-backs-new-constitution-
extending-president-term-limit/.
139. See Stef Vandeginste, Legal Loopholes and the Politics of Executive Term Limits: Insights from
Burundi, 51 AFR. SPECTRUM, 39, 45 (2016).
140. Id. at 52.
141. Clement Manirabarusha, Burundi President’s Commission Says People Want Term Limits Removed,
REUTERS (Aug. 25, 2016, 1:26 AM), https://www.reuters.com/article/us-burundi-politics/burundi-presidents-
commission-says-people-want-term-limits-removed-idUSKCN1100O1.
142. For an overview of judicial decisions in this area, see David Landau et al., Term Limits and the
Unconstitutional Constitutional Amendment Doctrine: Lessons from Latin America, in THE POLITICS OF
PRESIDENTIAL TERM LIMITS 53 (Alexander Baturo & Robert Elgie eds., 2019).
143. See Corrales & Penfold, supra note 31, at 160.
144. See id.
145. See Dixon & Landau, supra note 10, at 448–49.
146. REFORMA DE LA CONSTITUCIÓN, 2008, art. 441 (Ecuador).
147. See, e.g., Carlos Bernal Pulido, There Are Still Judges in Berlin: On the Proposal to Amend the
Ecuadorian Constitution to Allow Indefinite Presidential Reelection, INT’L J. CONST. L. BLOG (Sept. 10,
2014), http://www.iconnectblog.com/2014/09/there-are-still-judges-in-berlin-on-the-proposal-to-amend-the-
ecuadorian-constitution-to-allow-indefinite-presidential-reelection (arguing that the abolition of term limits in
Ecuador should require either a more demanding procedure or a constituent assembly).
February 2020] CONSTITUTIONAL END GAMES 381
presidents in both countries used the less demanding default standard, and the
high courts allowed them to do so.
148
In another class of Latin American countries, courts played a more direct
role and actively removed presidential term limits from their constitution by
holding that the term limits themselves were unconstitutional. This route was
generally taken in cases where presidents lacked the ability to make formal
changes to the constitution that eradicated term limits. In 2009, for example, the
Nicaraguan Supreme Court held that a constitutional term limit that would have
prevented President Ortega from running for a third term in office was
unconstitutional.
149
The court set the term limit aside, allowing Ortega to run for
and subsequently win a new term.
150
Once Ortega had sufficient parliamentary
support, his allies passed an amendment formally removing the term limit, and
he has remained in power ever since.
151
In 2015, the Honduran court used a similar maneuver to remove a
supposedly unamendable term limit,
152
allowing President Juan Orlando
Hernandez to run for, and subsequently win, reelection.
153
Similarly, in Bolivia,
after losing a referendum to extend presidential term limits that would allow
Bolivian President Evo Morales to run for a fourth consecutive term,
154
Morales
turned to the judiciary.
155
The Bolivian Constitutional Court obliged by holding
that the constitutional term limit was unconstitutional,
156
which paved the way
for Morales to run for reelection indefinitely.
157
There have been formal changes to term limits in other regions. For
example, term limits have recently been relaxed or repealed in Belarus,
148. See Corte Constitucional del Equador [Constitutional Court of Ecuador] Oct. 31, 2014, Dictamen
[Decision] 0001-14-DRC-CC; Republica Bolivariana de Venezuela Tribunal Supremo de Justicia Sala
Constitucional [Supreme Tribunal of Justice, Bolivarian Republic of Venezuela] Dec. 11, 2008, Expediente
[Case] 1610/08.
149. See Dante Figueroa, Nicaragua Supreme Court Decision Permitting President, Others to Seek
Reelection, LIBR. CONG.: GLOBAL LEGAL MONITOR (Dec. 18, 2009) https://www.loc.gov/law/foreign-
news/article/nicaragua-supreme-court-decision-permitting-president-others-to-seek-reelection/.
150. See Nicaragua: Ortega Allowed to Run for Third Successive Term, BBC NEWS (Jan. 29, 2014),
https://www.bbc.com/news/world-latin-america-25937292.
151. See id.
152. See Corte Suprema de Justicia, Sala de lo Constitucional [Supreme Court of Justice, Constitutional
Court] Apr. 22, 2015, http://www.poderjudicial.gob.hn/Documents/FalloSCONS23042015.pdf.
153. The term limit in Honduras, however, was part of the original 1982 Constitution, and not a later
amendment, forcing the court to use something like an “unconstitutional constitution” doctrine. For discussion,
see David Landau et al, From an Unconstitutional Constitutional Amendment to an Unconstitutional
Constitution? Lessons from Honduras, 8 GLOBAL CONST. 40, 52–53 (2019).
154. See Daniel Ramos & Monica Machicao, Bolivia’s Morales Accepts Defeat in Term Limit Referendum,
REUTERS (Feb. 23, 2016, 8:02 PM), https://www.reuters.com/article/us-bolivia-referendum-idUSKCN0VX09E.
155. See Bolivia Court Allows President Morales to Run for Fourth Term, AL JAZEERA (Dec. 5, 2018),
https://www.aljazeera.com/news/2018/12/bolivia-court-morales-run-fourth-term-181205052250163.html.
156. See Tribunal Constitucional Plurinacional, Sentencia Constitucional [Plurinational Constitutional
Court] Nov. 28, 2017, Expediente [Case] 20960-2017-42-AIA (BOL.).
157. In Bolivia, as in Honduras, the relevant term limit was part of the original 2009 Constitution, and not
that later amendment. See Landau et al., supra note 142, at 69–70.
382 HASTINGS LAW JOURNAL [Vol. 71:359
Kazakhstan, Tajikistan, and Uzbekistan.
158
In 2016, voters in Tajikistan voted to
approve a referendum to exempt the sitting President, Emomali Rahmon, from
all term limits under the 1994 Constitution, on the basis that he brought the
country “peace” and unity.
159
The relevant changes were approved by the
Constitutional Court of Tajikistan in 2016 and aligned with the 1999
amendments that lengthened the presidential term limit from five to seven
years.
160
Presidential attempts to lengthen or remove term limits have a high success
rate. Versteeg et al., for example, found forty successes out of the sixty attempts
since 2000.
161
This suggests that constitutional entrenchment of term limits has
not been especially successful, a problem that we explore in more detail in the
next Part. That said, there have been some prominent cases where presidents
tried and failed to change or evade their term limits, and we survey those cases
here.
In Africa, for example, between 1990 and 2016, attempts to evade term
limits failed in Zambia in 2001, Malawi in 2002, Nigeria in 2006, Senegal in
2011, and Burkina Faso in 2014.
162
In Zambia, President Frederick Chiluba tried
but failed to amend the constitution to allow a run for a third term in office.
163
Chiluba’s attempts to extend his term faced widespread popular opposition, as
well as opposition from within his own party.
164
Fifteen senior members of the
ruling Movement for Multiparty Democracy (MMD) publicly opposed the
amendment, and fifty-nine MMD legislators, including the Vice President,
signed a document vowing to oppose the President gaining a third term.
165
Ultimately, Chiluba resigned.
166
In Malawi, President Bakili Mulizi failed to amend the constitution to
allow a run for a third term because the amendment fell three votes short of
158. Ginsburg et al., supra note 1, at 1812–13; see also Christopher Walker, East: The “Leader for Life”
Governance Model, RADIO FREE EUR./RADIO LIBERTY (Feb. 29, 2008, 12:15 PM),
https://www.rferl.org/a/1079557.html (discussing changes in Belarus, Kazakhstan, Tajikistan and Uzbekistan).
159. Agence France-Presse, Tajikistan Votes to Allow President to Rule Indefinitely, GUARDIAN (May 23,
2016, 4:03 PM), https://www.theguardian.com/world/2016/may/23/tajikistan-votes-to-allow-president-
emomali-rahmon-to-rule-indefinitely.
160. See Alexis Wheeler, Tajikstan Court Approves Constitutional Amendments Removing Limit on
Presidential Term, JURIST (Feb. 5, 2016, 2:19 PM), https://www.jurist.org/news/2016/02/tajikistan-no-limit-on-
presidential-term.
161. See Versteeg et al., supra note 4.
162. See Tull & Simons, supra note 53, at 84–87 (coding this as a case of voluntary departure).
163. See Mwangi S. Kimenyi & Nelipher Moyo, The Late Zambian President Frederick Chiluba: A Legacy
of Failed Democratic Transition, BROOKINGS (June 24, 2011), https://www.brookings.edu/opinions/the-late-
zambian-president-fredrick-chiluba-a-legacy-of-failed-democratic-transition.
164. See Henri E. Cauvin, Zambia’s President Abandons Re-Election Bid, N.Y. TIMES (May 5, 2001),
https://www.nytimes.com/2001/05/05/world/zambia-s-president-abandons-re-election-bid.html.
165. See Posner & Young, supra note 100, at 133; see also Baker, supra note 8, at 293–94.
166. See Posner & Young, supra note 100, at 133; Tull & Simons, supra note 53, at 85; see also Baker,
supra note 8, at 294 (describing the influence of donor pressure in this context especially).
February 2020] CONSTITUTIONAL END GAMES 383
achieving a two-thirds super-majority in the legislature needed for successful
passage.
167
In Nigeria, President Olusegun Obasango’s attempt to run for a third term
was blocked by the Senate after a parliamentary minority voted against the
proposal from Obasango’s supporters to remove the existing two-presidential
term limit from the Nigerian constitution.
168
This failed attempt occurred against
the backdrop of two previous military coups in Nigeria (in 1963–66 and 1979–
83), and thus represents a notable instance of successful democratic
entrenchment.
169
In Senegal, President Wade was met with broad opposition when he
proposed to amend the constitution in 2011 to allow a third term in office.
170
The
civil society organization “M23” (named after protests held on May 23), or
“Y’en a Marre,” led to mass demonstrations against the proposed changes;
ultimately, Wade withdrew the proposed changes from parliament.
171
While he
still chose to run for a third term in the 2011 presidential elections, despite
contravening the constitution’s two-term limit, he was defeated in the polls by
the opposition candidate.
172
In Burkina Faso, plans by President Compaoré to amend the constitution
to seek a third presidential term in 2014 were met with mass public protest.
Inspired by their counterparts in Senegal, citizens in Burkina Faso formed a
protest movement entitled “Le Balai Citoyen” (Citizen Broom) calling for
Compaoré to leave office and the parliament to reject the proposed constitutional
changes.
173
Their rallying cry was “Hands off my Constitution.”
174
In the short-
term, these protests triggered, or at least provided a pretext for, a major
regression in commitments to constitutional democracy in the country.
175
They
led Compaoré to dissolve the parliament and the military to announce a take-
over of the government.
176
However, continued protests led the military to re-
instate civilian rule and call for fresh democratic elections in 2015, in which the
opposition candidate Roch Marc Christian Kaboré won a majority.
177
167. See Baker, supra note 8, at 294–95; Posner & Young, supra note 100, at 133.
168. See Posner & Young, supra note 100, at 126; Ross, supra note 125.
169. See Posner & Young, supra note 100, at 126.
170. See Janette Yarwood, The Struggle over Term Limits in Africa: The Power of Protest, 27 J.
DEMOCRACY 51, 53–55 (2016).
171. See id. at 53–54.
172. See LeBas, supra note 58, at 171; Yarwood, supra note 170, at 54.
173. See Yarwood, supra note 170, at 52–53.
174. See id. at 55.
175. See id. at 54–55.
176. See id. at 55.
177. See Matthew Bonkoungou & Nadoun Coulibaly, Kabore Wins Burkina Faso Presidential Election,
REUTERS (Nov. 30, 2015, 4:37 PM), https://www.reuters.com/article/us-burkina-election-
idUSKBN0TJ0QT20151201; Roch Marc Christian Kabore Elected Burkina Faso President, BBC NEWS (Dec.
1, 2015), https://www.bbc.com/news/world-africa-34971505.
384 HASTINGS LAW JOURNAL [Vol. 71:359
In Latin America, there have also been several failed attempts to change
term limits in recent years.
178
For example, in 2010, an attempt by President
Alvaro Uribe was blocked by a decision of the Colombian Constitutional Court,
which held that allowing a third consecutive term in office would be
unconstitutional because it would clash with core constitutional principles, such
as the separation of powers.
179
Versteeg et al. found that this was the only time
since 2000 that a court successfully blocked an attempted change to presidential
term limits.
180
In Panama in 1998 and Bolivia in 2016, attempts to change presidential
term limits through referendum were defeated by voters.
181
However in Bolivia,
as noted above, Morales eventually succeeded by successfully petitioning the
judiciary to remove the term limit.
182
In 2009, a possible attempt by Manuel
Zelaya of Honduras to change a one-term limit was blocked by the military,
which ousted the incumbent in a coup.
183
Finally, in 2017, an attempt by Horacio
Cartes of Paraguay to loosen presidential term limits via amendment failed in
the House of Representatives in the face of massive popular protests.
184
The
attempt at change via amendment was undertaken despite a constitutional text
that appeared to lay out with clarity that changes to presidential term limits
require a Constituent Assembly and cannot be undertaken using the normal
amendment mechanism.
185
A number of civil society groups, with the support
of the powerful Paraguayan Catholic Church, mobilized in opposition to the
measure.
186
At one stage the Congress was even set on fire.
187
Thus, attempted changes or evasions of presidential term limits to benefit
incumbents are quite common. While there have been cases where presidents
have failed to achieve their goals of loosening or eliminating term limits, often
in the face of popular mobilization, evasion attempts are usually successful.
178. See Corrales & Penfold, supra note 31, at 160; Frida Ghitis, Bolivia’s Term Limit Referendum a
Warning to Latin American Autocrats, WORLD POL. REV. (Feb. 25, 2016),
https://www.worldpoliticsreview.com/articles/18049/bolivia-s-term-limit-referendum-a-warning-to-latin-
american-autocrats; Daniel Ramos & Monica Machicao, Bolivia’s Morales Accepts Defeat in Term Limit
Referendum, REUTERS (Feb. 23, 2016, 8:02 PM), https://www.reuters.com/article/us-bolivia-referendum-
idUSKCN0VX09E.
179. See Corte Constitucional [C.C.] [Constitutional Court], febrero [Feb.] 2, 2010, Expediente [Decision]
C-141/10 (Colom.).
180. See Versteeg et al., supra note 44.
181. See Marco A. Gandásegui, Jr., The 1998 Referendum in Panama: A Popular Vote Against
Neoliberalism, 26 LATIN AM. PERSP. 159, 164 (1999); Ramos & Machicao, supra note 154.
182. See Bolivia Court Allows President Morales to Run for Fourth Term, supra note 155.
183. See Corrales & Penfold, supra note 31, at 160.
184. See Laurence Blair, Paraguay’s Re-election Crisis Is Over—for Now, WORLD POL. REV. (May 2,
2017), https://www.worldpoliticsreview.com/articles/22024/paraguay-s-re-election-crisis-is-over-for-now.
185. See CONSTITUCIÓN DE LA REPÚBLICA DEL PARAGUAY [CONSTITUTION] June 20, 1992, art. 229.
186. See Simon Romero, Protests Erupt in Paraguay over Efforts to Extend President’s Term, N.Y. TIMES
(Mar. 31, 2017), https://www.nytimes.com/2017/03/31/world/americas/paraguay-protests-horacio-cartes.html.
187. See id.
February 2020] CONSTITUTIONAL END GAMES 385
IV. THE CAUSES AND CONSEQUENCES OF NON-COMPLIANCE
The prior two Parts showed that, while presidential term limits are
extremely common, compliance with those provisions is problematic.
Incumbents use a range of tools, including constitutional amendment,
constitutional replacement, or the judiciary, to extend their tenure in office. This
Part asks two questions using the data of the prior section.
First, why are presidents often so determined to evade presidential term
limits? We root the issue in incentives. Second, what are the consequences of
presidential evasion of term limits? While the consequences obviously vary by
case, we think they are often very negative for constitutional democracy. Thus,
the term limits problem is one in urgent need of solutions.
A. T
HE CAUSES OF EVASION
What explains this pattern of weak compliance with relevant constitutional
limits? A range of factors are likely at play, including a limited history of respect
for constitutional limits in certain countries, the difficulty of constraining the
power of already extremely long-serving and dominant presidents, and the
insufficient entrenchment of relevant limits. However, we assert that the core
problem is one of presidential incentives.
Where a president has already served substantially more than two terms,
they may gain so much power that it becomes extremely difficult to dislodge
them. They may enjoy strong public support and tight control over the military,
and thus have a high capacity to withstand the threat of forcible removal from
office.
188
Another problem is that some countries, in which term limits have been
removed, have imposed weak entrenchment of their term limits, and require only
ordinary legislative majority requirements for the amendment or repeal of such
provisions. They have not increased entrenchment to these provisions or placed
them on a higher constitutional “tier” that requires some form of heightened
super-majority or double-majority support.
189
Countries that have made term
limits especially easy to change are likely to be vulnerable to attacks on those
limits.
In Africa, for example, Tull and Simons note that, of the fifteen cases of
successful amendment to term limits, five of these occurred by way of
amendments approved by parliaments using procedures similar to those for
ordinary statutes.
190
This was the case in Namibia in 1999, Togo in 2002, Gabon
in 2003, Cameroon in 2008, and Djibouti in 2010.
191
188. See Nic Cheeseman, Presidents for Life Spell Danger, MAIL & GUARDIAN (Apr. 12, 2019), at
https://mg.co.za/article/2019-04-12-00-presidents-for-life-spell-danger.
189. See Maltz, supra note 1, at 141 (“[E]ntrenchment would not ensure compliance with presidential term
limits, but it would increase the costs of abrogation.”).
190. See Tull & Simons, supra note 53, at 87.
191. Id. at 53 fig.2.
386 HASTINGS LAW JOURNAL [Vol. 71:359
In contrast, heightened super-majority requirements might be more likely
to lead to the defeat of proposed amendments to presidential term limits, even in
cases where a dominant party controls the legislature. In Burundi, for example,
before turning to the court to ratify his bid for a third term, President Nkurunziza
had first sought formally to amend relevant term limits through the legislature.
192
The proposed amendments failed to gain the necessary super-majority of the
Burundi Parliament by only a single vote.
193
This failure was largely because
Article 96 of the Burundi Constitution required amendments to receive the
support of at least 80% of the legislature,
194
an especially high super-majority
requirement in global terms.
195
A lack of constitutional entrenchment does not fully explain why
presidents often choose not to comply with term limits, and it is far from a
complete protection against evasion attempts. As noted above in Part III, in
many cases, incumbents have used tools other than formal amendment to
achieve their goals, such as constitutional replacement and judicial review.
Merely creating a high level of entrenchment against constitutional change will
not prevent the use of these other evasion tactics, and may indeed provide
additional incentive for their use, because they provide a substitute for formal
change.
Presidents facing a binding term-limit face a range of incentives not to
leave office. One incentive for a president to evade a term limit is to prevent or
delay the almost immediate reduction in a president’s power and privileges once
they leave office. Some democracies include provisions maintaining some of a
president’s privileges once they leave office, for example by continuing to
provide them with a salary, generous allowances, security, and staff.
196
As we
explore further below, the continued possession of presidential privileges is
extremely important to providing presidents the incentive to leave at the
constitutionally appointed time.
Even where they retain these privileges, presidents will inevitably face
some loss in their power and influence when they leave office. In addition,
192. See Vandeginste, supra note 139, at 43 n.1.
193. See id.
194. See id.
195. See Donald S. Lutz, Toward a Theory of Constitutional Amendment, 88 AM. POL. SCI. REV. 355, 361
tbl.4 (1994) (showing that most required legislative majorities are around 67%).
196. See Former Presidents Act, 3 U.S.C. § 102 (2018); Members of Parliament Retiring Allowances Act,
R.S.C. 1985, c. M-5.; see also WENDY GINSBERG & DANIEL J. RICHARDSON, CONG. RES. SERV., RL34631,
FORMER PRESIDENTS: PENSIONS, OFFICE ALLOWANCES, AND OTHER FEDERAL BENEFITS (2016); George
Bowden, Theresa May’s Salary Dwarfed By Her £2m Downing Street Perks, HUFFINGTON POST (July 20, 2017),
https://www.huffingtonpost.co.uk/entry/theresa-may-pay-salary-benefits-
perks_uk_5970be18e4b0aa14ea7810c4; Shikha Goyal, Benefits Given to Former Prime Minister and President
of India, JAGRAN JOSH (June 21, 2018), https://www.jagranjosh.com/general-knowledge/benefits-given-to-
former-prime-minister-and-president-of-india-1529585888-1; Paul Karp & Nick Evershed, Tony Abbott Gets
$90,000 Raise on Backbencher’s Salary Now He’s a Retired PM, GUARDIAN (May 30, 2019, 2:18 PM),
https://www.theguardian.com/australia-news/2019/may/30/tony-abbott-gets-100000-raise-on-backbenchers-
salary-now-hes-a-retired-pm.
February 2020] CONSTITUTIONAL END GAMES 387
leaving office inevitably means losing the ability to protect themselves from the
threat of future prosecution, imprisonment, or confiscation of their assets. There
is an increasing trend globally toward holding former leaders accountable for
international crimes, a risk which constitutions cannot “insure” former
presidents against.
197
This is one reason, Tom Ginsburg has suggested, that
democratic elites may agree to bind themselves to a minimum set of
constitutional guarantees.
198
Guarantees of this kind, however, will always be
subject to the danger of downstream renegotiation or nullification.
199
In new or fragile democracies, presidents may face few countervailing
pressures to abide by term limits. In more consolidated constitutional systems,
political parties may exert pressure on presidents to abide by the terms of a
democratic constitution. Because parties have a long-term interest in preserving
their reputation for respecting constitutional limitations, they may encourage the
president to step aside in favor of another member of the party. This reflects a
broader pattern in which long-term political actors may contribute to what
political scientists have called “self-enforcing” or “self-stabilizing”
constitutional term limitations.
200
As scholars such as Tonja Jacobi, Sonia
Mittal, and Barry Weingast have argued, constitutional term limits can be self-
stabilizing in two key ways.
201
First, constitutional term limits “lower[] the stakes of [democratic]
politics,” and therefore give political actors an incentive to support constitutional
action over extra-constitutional action.
202
Second, constitutional term limits
create a focal point for co-ordination among political actors, which can help
solve problems of co-ordination in ordinary politics.
203
Furthermore,
constitutional term limits are a classic mechanism for lowering the stakes of
democratic elections of a political party because they reduce the prospect of a
197. See, e.g., Leaders Facing Justice 1945–2013, COUNCIL OF FOREIGN REL., https://
www.cfr.org/timeline/leaders-facing-justice. See generally Jean-Baptiste Jeangène Vilmer, The African Union
and the International Criminal Court: Counteracting the Crisis, 92 INT’L AFF. 1319 (2016) (explaining
countries’ participation in the Rome Statute of the International Criminal Court to hold those in power
accountable after their reign ends); Amy Gunia, The Philippines Has Officially Left the International Criminal
Court, TIME (Mar. 18, 2019), https://time.com/5553323/philippines-leaves-international-criminal-court
(explaining one country’s decision to leave the International Criminal Court).
198. See Tom Ginsburg, Constitutional Courts in East Asia, in COMPARATIVE CONSTITUTIONAL LAW IN
ASIA 47, 48–49 (Rosalind Dixon & Tom Ginsburg eds., 2014); Rosalind Dixon & Tom Ginsburg, The South
African Constitutional Court and Socio-Economic Rights as “Insurance Swaps,” 4 CONST. CT. REV. 1, 4 (2011)
Rosalind Dixon and Tom Ginsburg, The Forms and Limits of Constitutions as Political Insurance, 15 INT’L J.
CONST. L. 998 (2017) (discussing using constitutional judicial review as a form of personal insurance).
199. See Dixon & Ginsberg, supra note 198, at 1000.
200. Tonja Jacobi et al., Creating a Self-Stabilizing Constitution: The Role of the Takings Clause, 109 NW.
U. L. REV. 601 (2015).
201. Id. at 603–606.
202. Id. at 604–05.
203. See id. at 605–06. Jacobi, Mittal and Weingast also identify a third condition: a constitution’s capacity
for change, or “adaption.” Id. at 606–07.
388 HASTINGS LAW JOURNAL [Vol. 71:359
single party gaining long-term control of the state or state power by winning a
single election.
204
Many fragile democracies, however, lack strong, independent political
parties with this kind of long-term interest in the alternation of political power.
205
Parties are often a vehicle for a single candidate to run for office, and they define
their policies in terms identical to those of a president or leader. Rightly or
wrongly, a political party’s electoral fate is often perceived by party members as
tied to that leader. When it comes time for a president to leave office under a
term limit, the party thus has little incentive to encourage a president to comply
with the relevant constitutional limitation. Instead of encouraging a president to
leave office, parties in a fragile democracy may actively encourage a president
to stay beyond the constitutional term limit as a means of increasing the chances
that the party itself will continue to hold power.
206
Latin America is a sobering reminder of the extraordinary length to which
presidents will go to overcome even a well-drafted constitutional design in
pursuit of their goals to remain in office. For example, both Venezuela and
Ecuador attempted to prevent changes to the “fundamental structure” of its
constitution by requiring a more demanding procedure to make any
constitutional amendments that would extend a president’s tenure beyond the
constitutional term limit.
207
Nonetheless, presidents were still able to circumvent
constitutional term limits using a procedure with a lower standard, and then have
high courts they controlled validate those maneuvers.
208
An even more dramatic example is given by countries like Nicaragua,
Bolivia, and Honduras, where courts issued decisions directly removing the
presidential term limits. In Bolivia, the regime-dominated court did so after
President Morales had already lost a referendum that would allow him to run for
another term.
209
Honduras was even more dramatic. The Honduran Court
removed a strict one-term limit that was part of the original 1982 Constitution,
which included an “eternity” clause that supposedly made the Constitution
unamendable and punished any political attempt to change the limit.
210
While an
earlier president, Manuel Zelaya, had been removed in a military coup after
204. Id.
205. On the role of weak parties in undermining presidential term-limits, see, for example, Baker, supra
note 8, at 297.
206. See Ginsburg et al., supra note 1, at 1828; Vandeginste, supra note 139, at 45.
207. See Dixon & Landau, supra note 10, at 448–49.
208. See supra text accompanying note 99.
209. See supra text accompanying notes 154–157.
210. See ASAMBLEA NACIONAL CONSTITUYENTE [CONSTITUTION] Jan. 20, 1982, art. 239 (HOND.) (“A
citizen who has held the Office of President under any title may not be President or a Presidential Designate.
Any person who violates this provision or advocates its amendment as well as those that directly or indirectly
support him shall immediately cease to hold their respective offices and shall be disqualified for ten years from
holding any public office.”); Id. at art. 374 (making the reelection provision unamendable).
February 2020] CONSTITUTIONAL END GAMES 389
allegedly seeking to change the clause,
211
incumbent president Juan Orlando
Hernandez successfully evaded the limit by leaning on the judiciary to do his
dirty work for him.
212
B. T
HE RISKS POSED BY TERM LIMIT EVASIONS
This pattern of non-compliance with constitutional term limits has
significant costs, or risks, from a democratic perspective. Some scholars suggest
that the formal removal of term limits constitutes a form of “hard” contravention,
whereas constitutional over-staying is a “softer” form of contravention.
213
However, formal constitutional change and constitutional non-compliance both
pose distinct dangers to constitutional democracy. Formal amendments
permanently remove key democratic checks and balances, whereas non-
compliance undermines basic commitments to the rule of law.
Presidents who simply ignore constitutional term limits are likely to
undermine a culture of respect for the rule of law. A disregard for constitutional
term limits can encourage disrespect of other constitutional limitations by other
government officials, including limitations on their power and prohibitions
against corruption. It may even encourage popular disobedience of legal rules
and thus an increase in looting, violence, and other forms of disrespect for legal
norms.
Other “informal” approaches to circumventing term limits can also put
pressure on the rule of law. One of the hallmarks of the rule of law, most scholars
agree, is that it involves respect for the independence of the judiciary and a
willingness to follow judicial decisions.
214
Judicial independence is, of course,
a question of degree.
Some scholars argue in favor of a “dialogic” or “departmentalist” approach
to constitutional construction,
215
where legislative or executive officials are
entitled to give narrow effect to court decisions with which they disagree, or to
disregard certain aspects of courts’ reasonings. However, there is no suggestion
211. For an accounting of these events, see NOAH FELDMAN et al., REPORT TO THE COMMISSION ON TRUTH
AND RECONCILIATION OF HONDURAS: CONSTITUTIONAL ISSUES (2011), https://papers.ssrn.com/sol3/
papers.cfm?abstract_id=1915214.
212. See Landau et al., supra note 142, at 63–65.
213. See Maltz, supra note 1, at 128.
214. See generally RULE BY LAW: THE POLITICS OF COURTS IN AUTHORITARIAN REGIMES (Tom Ginsberg
& Tamir Moustafa eds., 2008) (offering a compilation of articles discussing authoritarian regimes); Guillermo
O’Donnell, Quality of Democracy: Why the Rule of Law Matters, 15 J. DEMOCRACY 32 (2004) (explaining the
importance of democracy).
215. See, e.g., Rosalind Dixon, Constitutional “Dialogue” and Deference, in CONSTITUTIONAL DIALOGUE:
RIGHTS, DEMOCRACY, INSTITUTIONS 161 (Geoffrey Sigalet et al. eds., 2019); Jacob T. Levy, Departmentalism
and Dialogue, in CONSTITUTIONAL DIALOGUE: RIGHTS, DEMOCRACY, INSTITUTIONS 68 (Geoffrey Sigalet et al.
eds., 2019); Geoffrey Sigalet et al., Introduction to CONSTITUTIONAL DIALOGUE: RIGHTS, DEMOCRACY,
INSTITUTIONS 1 (Geoffrey Sigalet et al. eds., 2019).
390 HASTINGS LAW JOURNAL [Vol. 71:359
among constitutional scholars that the executive should be free to influence the
outcome of specific judicial decisions.
216
In Burundi, for example, there were several reports of direct interference
by the President and his supporters regarding the independence of the Burundi
Constitutional Court in 2015 during its deliberations over the application of
presidential term limits.
217
The Vice President of the Court, Sylvère
Nimpagaritse, refused to sign his name to the opinion and immediately fled to
Rwanda upon it being handed down.
218
In a Radio France International
broadcast, Nimpagaritse reported that the Burundi Constitutional Court had been
subject to intense political pressure in the lead-up to its decision, and that several
judges had received death threats before changing their vote to uphold the
constitutionality of the President’s third term.
219
Later, in Belgium,
Nimpagaritse identified specific judges in the majority who changed their
position due to such threats and discussed the threats made against him and his
family on behalf of the President.
220
Formal processes of amending or repealing term limits can erode
democratic checks and balances. In a few cases, presidents have been forced to
commit to constitutional changes that limit their power in order to obtain
political support for proposed changes to constitutional term limits. John Carey
notes that concessions of this kind can make the removal of term limits
significantly less damaging to democracy.
221
Scholars, such as Javier Corrales
and Michael Penfold, suggest that in some cases, where cross-party political
negotiations are at play, changes to term limits “can enhance accountability and
strengthen checks and balances.”
222
Presidents may make more substantial concessions to the opposition and
trade broader changes that limit the power of the presidency in exchange for
support to extend the term limit. In Argentina, for example, to gain support for
proposed changes to existing term limits, President Carlos Menem was required
to work with opposition legislators.
223
To secure their agreement for a formal
216. See generally Adam M. Samaha, Dead Hand Arguments and Constitutional Interpretation, 108
COLUM. L. REV. 606 (2008) (arguing that this would create some sort of “dead hand” control over constitutional
interpretation).
217. See Vandeginste, supra note 139, at 53.
218. Busingye Kabumba, A Legal Expert’s View on Burundi Term Limits Saga, OBSERVER (May 13, 2015),
https://observer.ug/viewpoint/37809-a-legal-expert-s-view-on-burundi-term-limits-saga; Vandeginste, supra
note 139, at 53.
219. See Vandeginste, supra note 139; Kabumba, supra note 218.
220. See Jerome Delay, Exiled Burundi Judge Says Court Forced to Back President, SAN DIEGO UNION-
TRIB. (May 6, 2015, 10:44 PM), http://www.sandiegouniontribune.com/sdut-burundi-opposition-politician-
arrested-over-2015may06-story.html. Note that this account was also contested. See Burundi Court Backs
President Nkurunziza on Third-Term, BBC NEWS (May 5, 2015), https://www.bbc.com/news/world-africa-
32588658 (disputing the Vice President’s account of the court receiving these threats).
221. See Carey, supra note 53.
222. See Corrales & Penfold, supra note 31, at 162.
223. See Carey, supra note 53, at 128; Corrales & Penfold, supra note 31, at 160 (describing the so-called
“Olivos Pact”); Ginsburg et al., supra note 1, at 1831–32.
February 2020] CONSTITUTIONAL END GAMES 391
change to term limits, Menem also agreed to reverse previous changes limiting
the independence of the Argentinian Supreme Court—Menem agreed to restore
the previous size of the bench and to appoint new, more independent judges.
224
This meant that the court was able to act as a greater check on the President and
his constitutional ambitions.
225
This will not be the case where concessions are simply designed to co-opt
members of the political opposition without making real changes. In the
Democratic Republic of Congo, for example, President Kabila attempted to
diffuse criticism of his attempts to circumvent term limits by appointing the
leader of the opposition as Prime Minister.
226
It was unlikely that the
appointment was a genuine attempt to share power with the opposition or reduce
the powers of the president.
Furthermore, the process of amending constitutional term limits may more
commonly create a window of opportunity for would-be authoritarian leaders to
achieve even broader forms of anti-democratic change. This is especially true
where presidents choose to repeal term limits via a process of wholescale
constitutional replacement, but similar effects may also be achieved by
amendment.
In Venezuela, for example, formal changes to existing constitutional term
limits were made by President Hugo Chavez without any meaningful opposition
involvement or support.
227
The first attempt, a referendum that narrowly failed
in 2007, was paired with a series of measures that would have greatly expanded
and centralized the power of the presidency.
228
While most of the broader
measures were dropped when Chavez passed a narrower measure eliminating
term limits in 2009, he was able to ultimately pass these measures through a
series of ordinary laws.
229
Likewise, in Algeria, the parliament voted in 2008 to repeal presidential
term limits, which would allow President Abdelaziz Bouteflika to seek a third
224. Carey, supra note 53, at 128; Corrales & Penfold, supra note 31, at 160 (describing the so-called
“Olivos Pact”).
225. See John M. Carey, Presidentialism and Representative Institutions, in CONSTRUCTING DEMOCRATIC
GOVERNANCE IN LATIN AMERICA 11, 20 (Jorge I. Domínguez & Michael Shifter eds., 2d ed. 2003). This also
included a decision to reject bids to interpret the Constitution so as to allow Menem a third term in office. Carey,
supra note 53, at 128.
226. See Jason Burke, Joseph Kabila Makes Opposition Leader DRC Prime Minister, GUARDIAN (Nov. 17,
2016, 12:00 PM), https://www.theguardian.com/world/2016/nov/17/joseph-kabila-makes-opposition-leader-
samy-badibanga-drc-prime-minister.
227. See Corrales & Penfold, supra note 31, at 161; Ginsburg et al., supra note 1, at 1831–32.
228. See Simon Romero, Venezuela Hands Narrow Defeat to Chavez Plan, N.Y. TIMES (Dec. 3, 2007),
https://www.nytimes.com/2007/12/03/world/americas/03venezuela.html?mtrref=www.google.com&gwh=77C
A77CD8541974C312BCAB406D952CF&gwt=pay.
229. See, e.g., Joel D. Hirst, Chavez Alters Venezuela’s Constitutional Regime, COUNCIL ON FOREIGN REL.,
(Dec. 20, 2010), https://www.cfr.org/expert-brief/chavez-alters-venezuelas-constitutional-regime (“Using the
gradual passage of ‘organic laws’ by his overwhelming majority in the National Assembly, Chavez has enacted
almost all aspects of his proposed 2007 constitution.”).
392 HASTINGS LAW JOURNAL [Vol. 71:359
term in office.
230
In passing these amendments, the parliament also approved a
parallel set of constitutional amendments that increased the scope of presidential
power and reduced the power of the prime minister.
231
This kind of “bundling”
is quite common in processes of formal constitutional change, especially those
that involve voting by ordinary voters at a referendum.
In Burundi, voters were asked to vote on fundamental constitutional
changes as a single bundle.
232
Leading news reports suggested that the ballot
paper did not contain any specific proposal for constitutional change, but rather
asked voters to vote “yes” or “no” in favor of the “constitutional referendum of
May 2018.”
233
As a result, the previous two-term limit on presidents was
removed and the length of each term increased from five to seven years.
234
The
changes also resulted in the creation of new offices for a prime minister that was
appointed by the president and a ceremonial vice president, which would be of
a different ethnicity to the president, and the removal of provisions that required
two vice presidents of different ethnicities.
235
Moreover, the bundled vote led to
the removal of the right to representation in cabinet for small parties.
236
It also
led to the removal of other power-sharing principles, such as norms of parity in
various parts of government service, including the police,
237
that were part of the
“Arusha Accords” that helped end the previous civil war in Burundi.
238
The net
effect of these changes, most commentators agreed, undermined Burundi’s
status as a stable, multi-ethnic democracy, and impaired the capacity of the
political opposition from rival ethnic groups to effectively challenge the
President.
239
As a result, the President’s practice of issue bundling relaxed
presidential term limits, which effectively expanded presidential power and
removed commitments to a multi-party, multi-ethnic democracy.
240
In most democracies, the issue bundling was not as crude as in Burundi,
where voters were given no formal legal opportunity to support selected or
limited forms of constitutional change. But democratic leaders can still, and
often, draw political connections between distinct legal proposals for change, so
230. See Heba Saleh, Algeria Lifts Presidential Term Limits, FIN. TIMES (Nov. 12, 2008),
https://www.ft.com/content/2e34aa3c-b0b8-11dd-8915-0000779fd18c.
231. Id.
232. See Jason Burke & Ignatius Ssuuna, Burundi Votes in Referendum over President’s 2034 “Power
Grab,” GUARDIAN (May 17, 2018, 4:08 PM), https://www.theguardian.com/world/2018/may/17/burundi-
referendum-presidential-powers-tensions-high.
233. See id.
234. See Burundi Votes in Referendum That Could Prolong President’s Rule, ENCA (May 17, 2018, 6:03
PM), https://www.enca.com/africa/burund-votes-in-referendum-that-could-prolong-presidents-rule.
235. See Bouka & Jackson, supra note 137; Burke & Ssuuna, supra note 232.
236. See Bouka & Jackson, supra note 232.
237. See id.
238. See Burke & Ssuuna, supra note 232.
239. See Eyder Peralta, Tension Grows Around Referendum in Burundi, NPR (May 16, 2018, 6:11 PM),
https://www.npr.org/sections/thetwo-way/2018/05/16/611668253/tension-grows-around-referendum-in-
burundi.
240. See Bouka & Jackson, supra note 137.
February 2020] CONSTITUTIONAL END GAMES 393
that supporters of one change vote in support of other changes that they know
less about or have less strong views about.
Thus, the formal removal of constitutional term limits can pose a double
danger to democracy. It can not only lead to a reduction in the competitiveness
of presidential elections and greater risk of informal abuses by presidents, but
also a formal increase in the scope of presidential power.
241
The process of
overcoming term limits may also pose other dangers to democracy. It may
provoke a form of political crisis that leads to the death or injury of hundreds of
citizens and calls for military or martial rule.
242
Political crises almost always
have somewhat unpredictable results. And they often threaten the stability of the
existing constitutional order.
243
V.
SOLVING THE END-GAME PROBLEM: THE ROLE OF CONSECUTIVE TERM
LIMITS
Constitutional designers can ameliorate the risks of term limit evasion by
barring only consecutive terms. As noted above in Part II, a common model
globally is the American one, where presidents are prohibited from regaining the
presidency for their entire lives after serving two consecutive terms in office.
244
A few systems go even further and prohibit reelection forever after only one
term in office.
245
This kind of term limit, however, maximizes the incentive
problem. If presidents can never return to power, they may try to hold on to it at
all costs.
In contrast, a less popular but still fairly common design choice is to bar
only consecutive attempts at reelection. One of the first bans of this kind was in
the Argentinian Constitution of 1853.
246
As noted in Table 1, as of 2019, we find
that eight percent of systems bar presidents from holding an immediately
consecutive term after one term in office; while another seven percent bar them
from doing so after holding two terms in office.
247
In both designs, however,
presidents can return to power later. We assert that such a model gives presidents
greater incentives to leave office rather than seeking to bend term limits to their
will.
A. C
HANGING PRESIDENTIAL INCENTIVES
If presidents can one day stand again for election to the presidency, they
face a different calculus as to the attractiveness of evasion attempts. While
241. See Ginsburg et al., supra note 1, at 1826.
242. See id.
243. Id. at 1849–50.
244. See id. at 1836.
245. See id. at 1837.
246. See Jorge Streb, Reelection or Term Limits? The Short and the Long View of Economic Policy, 26
ESTUDIOS DE ECONOMÍA [ECON. STUD.] 187, 193 (1999).
247. See supra text accompanying note 79; see also supra Table 1. We note that some presidential and
semi-presidential systems (4% in our data) have rules that are ambiguous as to whether they are permanent or
only consecutive bars. Sometimes, courts have clarified the meaning of these provisions.
394 HASTINGS LAW JOURNAL [Vol. 71:359
compliance means foregoing the immediate powers and privileges of office,
attempted evasion may risk undermining a leader’s reputation in ways that
damage their long-term reelection prospects.
Whether this long-term risk is sufficient to outweigh the immediate
benefits of non-compliance will depend on a range of factors, including the age
of an incumbent, as well as prevailing social, economic, and political conditions.
The younger a president is, the lower the “discount rate” they will generally have
when trading off current versus future political gains. All else being equal,
younger presidents are more likely to respond to the incentives created by
temporary bans on reelection.
Other factors may also affect a president’s choice regarding whether to step
down. If the economy is strong, for example, a risk-averse president may risk
damaging their long-term chance of reelection, and instead pursue a high
possibility of immediate reelection by seeking to repeal or circumvent current
term limits.
248
The same is true for presidents with a high rate of voter approval.
Approval ratings indicate a high chance of reelection and strong support for
formal constitutional amendment to make reelection possible.
249
Conversely, if
economic conditions are declining, or if a government is faced with voter
dissatisfaction, a president may have greater reason to prioritize his or her
chances of long-term reelection.
250
There is strong empirical evidence to suggest that the prospect of later
reelection is a relevant factor in presidents’ actual decision-making. Corrales and
Penfold identify eleven countries in Latin America that barred consecutive terms
but allowed presidential reelection after one or two intervening terms.
251
In those
eleven countries, they found that former presidents ran in 50% of elections in
which they were eligible to run.
252
Some of these elections—Argentina in 2003,
Bolivia in 1997 and 2002, Haiti in 2006, and Suriname in 2000 and 2005—
actually featured more than one former president.
253
There have been several recent instances of presidents “voluntarily”
choosing to leave office at the end of their constitutionally appointed terms in
office because of the possibility of running for reelection in the future. In
Ecuador, as previously noted, Congress, in 2015, passed a constitutional
amendment that allowed indefinite presidential reelection.
254
This was done
248. See, e.g., Corrales & Penfold, supra note 31, at 164 (discussing the relevance of economic factors both
to presidential reelection, and the chances of successful removal of term limits); see also Baker, supra note 8, at
291.
249. See, e.g., Michael S. Lewis-Beck & Tom W. Rice, Presidential Popularity and Presidential Vote, 46
PUB. OPINION Q. 534, 534 (1982); Maltz, supra note 1, at 132.
250. Maltz, supra note 1, at 132.
251. See Carey, supra note 53, at 125.
252. See Corrales & Penfold, supra note 31, at 163.
253. See id. In Europe, 17.4% of presidential races involved a former president or prime minister as a
candidate. Id. at 163; see also Javier Corrales, Latin America’s Neocaudillismo: Ex-Presidents and Newcomers
Running for President . . . and Winning, 50 LATIN AM. POL. & SOC’Y 1, 6 (2008).
254. See supra text accompanying notes 144–148.
February 2020] CONSTITUTIONAL END GAMES 395
using the default procedure for amendment, rather than the more demanding
procedure for constitutional changes that affect the “fundamental structure.”
255
As a result, President Correa was able to avoid a popular referendum that the
opposition sought and he feared.
256
When faced with significant street protests, Correa agreed to bless a small
but significant last-minute change to the amendment: the elimination of
presidential term limits that would take place only after the next election in
2017.
257
Effectively, this change meant that Correa would need to leave power
in 2017, although he could potentially run for reelection and remain in the
presidency indefinitely. His hand-picked successor, Lenin Moreno, won a
narrow victory in the 2017 election.
258
However, as explained in more detail
below, Moreno quickly turned against Correa, and one of his first major acts was
to hold a successful referendum reinstating presidential term limits, blocking
Correa’s return to power.
259
The Ecuadorian case has been celebrated as one of “democracy’s near
misses”: a set of cases where democratic constitutionalism was close to suffering
major erosion towards authoritarianism, but escaped long-term damage in the
end.
260
Correa was widely viewed as a significant threat to democracy.
261
He
took a number of formal and informal steps to consolidate power, including
replacing the existing constitution with a new one unilaterally drafted by his
political allies, and used an expanding set of legal tools to harass opposing
political, civil society, and media groups.
262
Scholars have asserted that Ecuador
had already eroded into a competitive authoritarian regime during Correa’s
tenure.
263
The temporary term limit that emerged from the 2015 crisis and forced
Correa’s exit from power should thus be given considerable credit in salvaging
a democratic constitutional order.
We do not know exactly why Correa agreed to the provision preventing
him from running in the 2017 election, but it seems highly likely that the fairly
young incumbent president did so because he thought he could return to power
255. See REFORMA DE LA CONSTITUCIÓN, 2008, art. 441 (Ecuador).
256. See Megan Alpert, Correa’s Gamble, ABC (Dec. 14, 2015, 2:19 PM),
https://foreignpolicy.com/2015/12/14/correas-gamble-ecuador.
257. See, e.g., Thalía Flores, The Official Majority Approves the Indefinite Reelection in Ecuador, ABC
(Dec. 5, 2015, 9:40 PM), https://www.abc.es/internacional/abci-mayoria-oficialista-aprueba-reeleccion-
indefinida-ecuador-201512041908noticia.html.
258. See Ecuador Election: Socialist Lenin Moreno Declared Winner, BBC NEWS (Apr. 4, 2017),
https://www.bbc.com/news/world-latin-america-39498404.
259. See infra text accompanying notes 300–302.
260. See Tom Ginsburg & Aziz Huq, Democracy’s Near Misses, 29 J. DEMOCRACY 16, 26 (2018).
261. See generally Marc Becker, The Stormy Relations Between Rafael Correa and Social Movements in
Ecuador, 40 LATIN AM. PERSPECTIVES 43 (2013); Carlos de la Torre & Andrés Ortiz Lemos, Populist
Polarization and the Show Death of Democracy in Ecuador, 23 DEMOCRATIZATION 221 (2016).
262. See generally Becker, supra note 261; de la Torre & Lemos, supra note 261.
263. See Steven Levitsky & James Loxton, Populism and Competitive Authoritarianism in the Andes, 20
DEMOCRATIZATION 107, 121 (2013) (stating that the Ecuadorian constitution-making process by Correa in 2008
“ushered in a competitive authoritarian regime”).
396 HASTINGS LAW JOURNAL [Vol. 71:359
later. Indeed, some provisions of the design made it possible for Correa to return
without sitting out a full presidential term.
264
Many analysts predicted economic
and fiscal problems for Ecuador during the period in which Correa left power,
265
making it particularly attractive for him to take a break before making a
triumphant return. At the same time, the case shows that temporarily forcing
would-be autocrats out of power may have dramatic effects on a country’s
democratic trajectory, even when their own party retains the presidency.
In Panama, during the lead-up to the 2014 presidential election, President
Martinelli enjoyed a 60% approval rating but faced a binding term limit that
prevented him from seeking reelection.
266
He initially sought to propose
constitutional changes aimed at removing these limits from the Constitution.
267
First, Martinelli proposed a formal constitutional amendment to allow
presidents to serve two consecutive terms.
268
However, he faced significant
resistance from within his own party.
269
His Vice President, Juan Carlos Varela,
publicly opposed the measure, and effectively began his bid to challenge
Martinelli for the presidency.
270
Second, Martinelli proposed measures designed informally to reduce the
effect of term limits by seeking to appoint new justices to the Supreme Court of
Panama who would be willing to invalidate the existing term limits as an
unconstitutional provision.
271
Nicaragua provided an important model for
informal change of this kind.
272
In proposing to recreate a “Fifth Court” or
chamber in the Supreme Court that would be responsible for constitutional cases,
and that would staffed entirely by judges appointed by the President, Martinelli
264. A provision of the Ecuadorian Constitution called the “crossed-death” provision could potentially
allow the president to dissolve the legislature and call new elections for both the executive and legislative
branches under certain conditions. See REFORMA DE LA CONSTITUCIÓN, 2008, art. 148 (ECUADOR).
265. See, e.g., Andrés Oppenheimer, Correa’s “Economic Miracle” in Ecuador Was Actually a
Monumental Sham, MIAMI HERALD (Mar. 3, 2017, 7:00 AM), https://www.miamiherald.com/news/local/news-
columns-blogs/andres-oppenheimer/article136054198.html.
266. See Juan Zamorano, Panama’s Vice President Wins Presidential Election, USA TODAY (May 5, 2014,
12:06 AM), https://www.usatoday.com/story/news/world/2014/05/05/panama-presidential-election/8710369/;
see also Kevin Lees, Panamanian Presidential Race is All About Martinelli, SUFFRAGIO (May 2, 2014),
http://suffragio.org/2014/05/02/panamanian-presidential-race-is-all-about-martinelli/.
267. See Kevin Lees, Meet Juan Carlos Varela, Panama’s New President, SUFFRAGIO (May 4, 2014),
http://suffragio.org/2014/05/04/meet-juan-carlos-varela-panamas-new-president.
268. See id.
269. See id.
270. See id.; Christine Murray & Elida Moreno, Panama Leader’s Deputy-Turned-Rival Wins Presidency,
REUTERS (May 5, 2014, 5:50 PM), https://www.reuters.com/article/us-panama-election/panama-leaders-deputy-
turned-rival-wins-presidency-idUSBREA4308520140505.
271. See Juan Carlos Hidalgo, A Brewing Institutional Crisis in Panama, CATO INST. LIBERTY (Feb. 1, 2012,
10:57 AM), https://www.cato.org/blog/brewing-institutional-crisis-panama.
272. See id.
February 2020] CONSTITUTIONAL END GAMES 397
sought to use similar tactics.
273
He only ultimately backed down in the face of
widespread public protest.
274
When these attempts failed, Martinelli chose to step aside and declined to
contest the 2014 presidential election.
275
Instead, Martinelli chose to support
former Housing Minister, Jose Domingo Arias, as a presidential candidate, and
to promote Martinelli’s wife, Marta Linares, as a Vice Presidential candidate.
276
He broke with convention for an ex-president and actively campaigned for Arias
and his own wife.
277
Martinelli also repeatedly defended Arias’ achievements in
reducing unemployment and promoting economic growth, and warned that these
gains could be reversed if the opposition’s candidates were elected.
278
Martinelli
took on an active campaign role despite the risk of corruption and other criminal
charges that he was facing.
279
As a result of these investigations, in 2015
Martinelli left Panama and moved to the United States, where he faced
extradition proceedings,
280
before subsequently returning to Panama.
Martinelli’s decision to step down from the presidency likely reflected a
desire to maintain his long-term reputation and to preserve his capacity to run in
future presidential elections after the two-term “cooling off” period for which
the Constitution provides. He sought to run for mayor of Panama City in 2019
and would have been the front runner, but the Electoral Court rejected his
candidacy as he failed to comply with residency requirements since he fled to
the United States for several years before returning to Panama.
281
He has also
stated that he intends to be a candidate for president in 2024 following the two-
term cooling off period.
282
We add one additional piece of relevant evidence. Despite being a country
with considerable political turmoil, Brazil has nonetheless been devoid of
attempts to eliminate or loosen term limits. Brazil has had two presidential
273. See id.
274. See Lees, supra note 267 ("[V]oters weren’t prepared to back Martinelli’s ‘extraconstitutional’
workaround to retain power for another five years.”).
275. See Christine Murray & Elida Moreno, Panama Election to Test Martinelli’s Hold on Power, REUTERS
(May 1, 2014, 10:34 AM), https://af.reuters.com/article/worldNews/idAFKBN0DH37A20140501.
276. See Panama Election: Juan Carlos Varela Thwarts Rival by Winning Presidency, GUARDIAN (May 4,
2014, 10:08 PM), https://www.theguardian.com/world/2014/may/05/panama-election-juan-carlos-varela-
thwarts-rival-by-winning-presidency.
277. See Lees, supra note 267; Zamorano, supra note 266.
278. Lees, supra note 267; Zamorano, supra note 266.
279. See Randal C. Archibold, Incumbent’s Party Loses Presidency in Panama, N.Y. TIMES, May 5, 2014,
at A.11 (suggesting that Arias and Linares’ candidacy were in part designed to guard against this risk).
280. See Orlando J. Pérez, Panama: Democracy Under the Shadow of Corruption, 37 REVISTA DE CIENCIA
POLÍTICA [POL. SCI. MAG.] 519, 525 (2017); Scandal in Panama: No Chavez, but No Prize, ECONOMIST (Feb.
5, 2015), https://www.economist.com/the-americas/2015/02/05/no-chavez-but-no-prize.
281. Stefanie Eschenbacher & Elida Moreno, Panama Tribunal Rules Ex-President Martinelli Cannot Run
in Election, REUTERS (Apr. 26, 2019, 8:24 AM), https://www.reuters.com/article/us-panama-election-
martinelli/panama-tribunal-rules-ex-president-martinelli-cannot-run-in-election-idUSKCN1S21Q2.
282. Id.
398 HASTINGS LAW JOURNAL [Vol. 71:359
removals via impeachment since its return to democracy,
283
and many other
crises, but no serious attempts at evasion of term limits. In some cases, such as
during and after the presidency of the popular incumbent Luiz Inacio Lula da
Silva, the possibility of a return to power was clearly relevant. Lula planned on
running for power in 2018, after his successor, Dilma Rousseff, was impeached
and removed. However, his candidacy was stopped by a corruption investigation
that left him in jail.
284
These examples might be added to others, such as the case of Cristina
Kirchner in Argentina, which is considered in more detail below, in which the
prospect of a subsequent return to power seems to have impacted presidential
decisions not to proceed with evasion attempts. These examples are not, of
course, a proper “test” of our theory, and we do not attempt such a test here. An
empirical test would be, at any rate, difficult to carry out in a convincing way,
given the relative rarity of the consecutive model and the large number of factors
that impact evasion attempts and their success.
B. A
RE CONSECUTIVE BANS STRONG ENOUGH?
One response to our argument above may be that weak or temporary term
limits are in fact “successful” in avoiding evasion only because term limits
merely defer, rather than reduce, the dangers to democracy posed by personalist
presidential rule. As we noted above, many presidents facing temporary bans on
reelection go on to stand at subsequent democratic elections.
285
Some have also
ultimately re-assumed the role of the would-be authoritarian leader after
returning to power (for example, Russian President Vladimir Putin, Congolese
President Denis Sassou-Nguesso, and Dominican President Rafael Trujillo in
the 1940s).
286
It remains to be seen whether some recent Latin American
presidents, such as Martinelli, will succeed in recapturing the presidency.
283. See Alexandra Rattinger, The Impeachment Process of Brazil: A Comparative Look at Impeachment
in Brazil and the United States, 49 U. MIAMI INTER-AM. L. REV. 129, 146–150 (2017) (discussing impeachment
of Presidents Collor and Rousseff).
284. See Phillips, supra note 97. We note that some countries with a consecutive term limit have had them
loosened or eliminated in recent years in Latin America. However, these attempts have often involved the
wholesale replacement of the constitution, not a targeted amendment of the term limit. Venezuela, Ecuador, and
Bolivia, for example, all had a one consecutive term limit before the ascension to power of Presidents Chavez,
Correa, and Morales, respectively. Each president, as one of their first acts in office, called Constituent
Assemblies that replaced their constitutions, and these new constitutions, promulgated in 1999, 2008, and 2009,
all allowed presidents to serve two consecutive terms in office. But these constitutional conventions did not seem
primarily aimed at loosening term limits; instead they were aimed at a number of ideological and policy goals
of the different leaders, most importantly the “refound[ing]” of their respective states. See David Landau,
Constitution-Making Gone Wrong, 64 ALA. L. REV. 923, 951 (2013). As noted by Versteeg et al., it is often
difficult to find a change to a term limit as the primary goal of a wholesale constitutional replacement, although
such a process may well change the term limit once it has been activated. See Versteeg et al., supra note 4.
285. See supra notes 91–97 (discussing several former presidents in Chile, Comoros, and Brazil who ran
for non-consecutive presidential reelection).
286. Ginsburg et al., supra note 1, at 1850 (discussing Putin and Trujillo); Congo Leader Nguesso Promises
Referendum over Third Term, BBC NEWS (Sept. 23, 2015), https://www.bbc.com/news/world-africa-34336471;
February 2020] CONSTITUTIONAL END GAMES 399
There are, however, two key dynamics that suggest that a delay in a
president’s ability to seek reelection is likely to reduce the risk of a slide toward
authoritarianism, and not simply to postpone its occurrence.
First, a delay creates a window of opportunity for alternative leaders to
emerge. When a president steps down from power, his or her successor will often
begin to build up power, thus blunting the consolidation of power of the prior
incumbent.
287
An important benefit of non-consecutive term limits, therefore, is
that term limits can provide a window of opportunity for the strengthening of
political parties as structures independent from the individual president or
leader.
Sometimes, this occurs because the president’s chosen successor, lacking
the advantages of the incumbent, loses the next election. Consider Panama,
where, as noted above, President Martinelli stepped down and his chosen
candidate lost to Juan Carlos Varela, his former vice president, turned
adversary.
288
Varela took power in an increasingly difficult economic and social
situation, and his approval ratings declined fairly rapidly.
289
However, he took
advantage of the presidency to develop policies that built on, but were quite
distinct from, Martinelli’s.
290
While he continued anti-poverty projects and
large-scale public infrastructure projects, such as the expansion of the Panama
Canal, he argued for increased transparency and action to combat global money
laundering in Panama.
291
Furthermore, he introduced new policies designed to
increase food affordability, and lobbied for rebuilding democratic and social
institutions.
292
As a result, support for Varela’s Democratic Revolutionary Party
(PRD) increased and support for Martinelli’s Cambio Democratico (CD) party
declined.
293
Between 2015 and 2018, 119,660 voters chose to leave the CD
party, leaving it with only 338,842 registered members, compared to the PRD
with 462,395 members.
294
In effect, non-consecutive term limits allowed Varela
to build his own base of support, while Martinelli’s eroded.
Conor Gaffey, Congo’s Denis Sassou Nguesso Extends 32-Year Rule, NEWSWEEK (Mar. 24, 2016, 12:13 PM),
http://www.newsweek.com/congos-denis-sassou-nguesso-extends-32-year-rule-440356.
287. See infra notes 288–299 and accompanying text.
288. Lees, supra note 267.
289. See Varela Seeks to Boost His Declining Popularity, ECONOMIST (Jan 11, 2016),
http://country.eiu.com/article.aspx?articleid=843838468&Country=Panama&topic=Politics&subtopic=Fo_6;
Daniel Zovatto, Panama Stands Up for Alternation, BROOKINGS (May 17, 2014),
https://www.brookings.edu/opinions/panama-stands-up-for-alternation/.
290. See John Murray Brown, Panama’s Juan Carlos Varela Promises to Improve Transparency, FIN.
TIMES (Apr. 7, 2016), https://www.ft.com/content/7f4a0cd6-fc9c-11e5-b3f6-11d5706b613b.
291. See id.
292. See Zovatto, supra note 289.
293. See Eric Jackson, Parties and Presidential Candidates Lining Up, PAN. NEWS (Jul 18, 2017),
https://www.thepanamanews.com/2017/07/parties-and-presidential-candidates-lining-up/.
294. See id.
400 HASTINGS LAW JOURNAL [Vol. 71:359
Strikingly, similar dynamics often hold even where the winner of the next
election was handpicked by the old president. Consider two recent cases from
Latin America: Colombia and Ecuador.
In Colombia, as noted above, the Colombian Constitutional Court issued a
landmark, but highly unusual, decision denying President Alvaro Uribe the
possibility of running for a third term.
295
Uribe left power and supported his ally
and fellow party-member Juan Manuel Santos in the 2010 election.
296
With
Uribe’s support, Santos won the election.
297
However, Santos developed an
independent party base that pursued policies contrary to Uribe’s own agenda,
which eventually made the two men archenemies.
298
Santos, for example,
pursued peace with the FARC guerrilla movement, a policy Uribe detested, and
eventually reimposed an absolute one-term limit on the presidency, reversing
changes carried out by Uribe.
299
Similarly, in Ecuador, President Correa agreed to leave power in 2017 and
backed his close ally Lenin Moreno in the subsequent election.
300
Like Santos,
Moreno won a close election. After developing his own power base, Moreno
quickly turned against Correa.
301
One of Moreno’s first acts was to hold a
successful referendum reimposing presidential term limits, foreclosing Correa’s
planned return to power.
302
The independent political space we explain here will not always emerge.
Sometimes, a former president will seek to continue to exert significant ongoing
influence over the direction of a party, and independent bases of power will not
always develop. They may indeed succeed in placing a puppet or shadow-
president in office, a point we explore in more detail below.
Where presidents have a long-term chance to be re-elected, they also have
an incentive to undermine the rise of credible alternative leaders and influence
the party in a direction of non-independence. Tarnishing the standing of
alternative party leaders, however, carries clear dangers for former presidents.
For example, the party of former presidents may lose control of the presidency,
295. See Corte Constitucional [C.C.] [Constitutional Court], Febrero [Feb.] 2, 2010, Expediente [Decision]
C-141/10.
296. See John Otis, Colombia’s New President: A Win for the U.S., TIME (June 21, 2010),
http://content.time.com/time/world/article/0,8599,1998279,00.html.
297. See Sibylla Brodzinsky, Juan Manuel Santos Wins Colombia Presidential Election, GUARDIAN (June
21, 2010, 2:43 PM), https://www.theguardian.com/world/2010/jun/21/juan-manuel-santos-colombia-president.
298. Reyes L., supra note 88.
299. See José Miguel Vivanco & Juan Pappier, Álvaro Uribe: Colombia Peace Deal’s Unwelcome Critic,
MIAMI HERALD (Aug. 15, 2016, 7:52 PM), https://www.miamiherald.com/opinion/op-ed/article95846637.html;
see also Reyes L., supra note 88 (noting, as well, that the new one-term limit may only be changed via
referendum).
300. See Maggy Ayala & Marcelo Rochabrún, Ecuador Votes to Bring Back Presidential Term Limits, N.Y.
TIMES (Feb. 4, 2018), https://www.nytimes.com/2018/02/04/world/americas/ecuador-presidential-term-
limits.html.
301. Id.
302. Id.
February 2020] CONSTITUTIONAL END GAMES 401
and thus, the ability to influence a range of other important appointments,
including that of prosecutors and judges who may influence a subsequent
prosecution against former presidents.
303
Second, a delay helps counter the effect of certain behavioral biases among
voters that tend to encourage a preference for the political status quo. A
consecutive term limit gives voters an opportunity to reassess the reasons behind
their support for the incumbent without the cloud of incumbency. Voters will
then be in a better position to assess whether their support in fact reflects genuine
political preferences for the incumbent or instead, common behavioral biases.
Behavioral economics and social psychologists have shown that people are
sometimes subject to forms of “endowment effect,” meaning they tend to place
greater emphasis on certain kinds of losses over equivalent gains.
304
The
“endowment effect” can also create a persistent form of “status quo” bias in
individual and collective decision-making.
305
In a democratic context, a “status
quo” bias results in voters preferring political incumbents simply because they
fear losing whatever prosperity and security they currently enjoy.
306
Similarly, people often put different values on their “decision utility”
versus “experience utility,” or have different reactions to imagining things and
living them.
307
When a strong person tells people, “I am essential to your
wellbeing,” it is hard for voters to imagine what it would look like to live under
a different regime. While they may ultimately find that it is not only possible to
live with different leadership, and sometimes fare better under it, imagining this
alternate reality can be extremely difficult.
From this perspective, consecutive term limits may tend to disrupt biases
of this kind. By mandating a transfer of power to a new president, term limits
help counter the effect of endowment effects among voters. By giving voters the
opportunity to experience a new leader, consecutive term limits may also help
reduce the gap between decision and experience utility.
303. See Baker, supra note 8 at 296; Posner & Young, supra note 100, at 135; Roger Tangri & Andrew M.
Mwenda, President Museveni and the Politics of Presidential Tenure in Uganda, 28 J. CONTEMP. AFR. STUD.
31, 33 (2010); Thomas P. Wolf, Immunity or Accountability? Daniel Toroitch arap Moi: Kenya’s First Retired
President, in LEGACIES OF POWER: LEADERSHIP CHANGE AND FORMER PRESIDENTS IN AFRICAN POLITICS 197,
201 (Roger Southall & Henning Melber eds., 2006). In Kenya, for example, many commentators suggest that
President Moi over-reached when he nominated Kenyatta as his successor: Kenyatta was relatively
inexperienced, and thus not seen as a real threat to Moi’s ongoing power and influence. Id. at 201. But this
ultimately also contributed to the party’s loss at the next presidential election, and a major re-alignment in the
racial-ethnic make-up of the Kenyan government (of a kind that was clearly disfavored by and unfavorable to
Moi). See Nic Cheeseman, African Elections as Vehicles for Change, 21 J. DEMOCRACY 139, 144 (2010).
304. See, e.g., Christine Jolls et al., A Behavioral Approach to Law and Economics, 50 STAN. L. REV. 1471,
1484 (1998). See generally Daniel Kahneman et al., Anomalies: The Endowment Effect, Loss Aversion and Status
Quo Bias, 5 J. ECON. PERSP. 193 (1991).
305. Kahneman et al., supra note 304, at 194–99.
306. See Corrales & Penfold, supra note 31, at 163; Ginsburg et al., supra note 1, at 1820.
307. See, e.g., Daniel Kahneman & Richard H. Thaler, Anomalies: Utility Maximization and Experienced
Utility, 20 J. ECON. PERSP. 221, 221–22 (2006); Daniel Kahneman & Richard Thaler, Economic Analysis and
the Psychology of Utility: Applications to Compensation Policy, 81 AEA PAPERS & PROC. 341, 343 (1991).
402 HASTINGS LAW JOURNAL [Vol. 71:359
The evidence on presidential reelection rates lends considerable weight to
the power of these dynamics. In Latin America, for example, Corrales and
Penfold found that between 1998 and 2006, the general reelection rate for sitting
presidents was between 83% and 90%, a massive incumbency advantage.
308
For
former presidents, however, the reelection rate dropped considerably to 40%.
309
This was comparable to the election rate of other major party candidates, so that
the authors found “no detectable advantage for ex-presidents running for
nonconsecutive reelection.”
310
For example, in Ecuador, Correa left office as a highly popular president.
311
Nonetheless, only eight months after he left office, his preferred position was
trounced in Moreno’s successful referendum to reinstate presidential term limits.
The result of this referendum has seemingly blocked Correa from making a
return to power.
312
This example suggests the ways in which distance from
incumbents can erode the formidable advantages they enjoy while still in office.
C. D
ESIGNING CONSECUTIVE TERM LIMITS: ONE TERM OR TWO
As noted above in Table 1, there is important variation in the design of
consecutive term limits around the world. About half the systems that include
this design make presidents leave office at least temporarily after serving only
one term in office, and the other systems require a break after two consecutive
terms. There appears to be a near-consensus that allowing more than two
consecutive terms in office is a bad idea, and only one system in the world with
presidential term limits does so.
The choice between one or two consecutive terms raises a basic tradeoff.
Allowing only one consecutive term in office may help to reduce the risk that a
single president will consolidate enough power to become a threat to democracy
and be able to successfully evade term limits. A constant rotation in office
prevents the construction of overly strong power bases, even where presidents
return to power later, and will instead allow space for competing power centers.
On the other hand, allowing presidents to serve no more than two
consecutive terms in office has well-known advantages, as suggested above in
Part I.
313
A consecutive two-term limit recognizes the significance of executive
experience, provides longer time-frames for policy-making and development,
and gives voters the opportunity to hold a president immediately accountable for
his or her policy promises.
314
In short, presidents who serve two consecutive
308. Corrales & Penfold, supra note 31, at 163.
309. Id.
310. Id.
311. See Ecuador’s Correa Leaves Office with 62 Percent Approval Rating, TELESUR (May 18, 2017),
https://www.telesurenglish.net/news/Ecuadors-Correa-Leaves-Office-With-62-Percent-Approval-Rating-
20170518-0027.html.
312. See Ayala, supra note 300.
313. See Carey, supra note 53, at 127.
314. See Cain & Levin, supra note 17, at 178; Streb, supra note 246, at 193, 204.
February 2020] CONSTITUTIONAL END GAMES 403
terms are more likely to fulfill their policy promises and are more subject to
democratic accountability.
Whether a one or two consecutive term limit will be the better choice
depends on a number of other elements of a country’s constitutional design. The
most obvious one is the length of presidential terms, which globally appear to
vary between four and seven years.
315
The impact of forcing a president to stand
down after four years is quite different from forcing them to stand down after
six. Thus, while a system allowing two consecutive six-year terms is ill-advised,
it may be a good idea to allow consecutive terms if the presidential terms are
only four years, as in Brazil and Argentina.
316
Furthermore, the relative length of terms is an important aspect of design.
The threat posed by longer stays in office may be lessened if the terms of other
key officials, like justices, are longer and more insulated from presidential
control.
317
Thus, it may be reasonable to allow two consecutive terms in office
where presidents have limited powers to appoint or influence officials that are
intended to check their actions.
A final key feature of consecutive term limits design is how long
presidents must sit out for before returning to office. While most systems impose
only a one-term break, some, such as Panama, require a longer break of two
terms.
318
Constitutional designers may be more comfortable allowing presidents
to serve two consecutive terms in office if presidents must then sit out longer
because this helps avoid the consolidation of power and ameliorates the
possibility of ex-presidents continuing to run the show through the installation
of “shadow presidencies,” an issue we explore next.
VI.
AVOIDING THE PROBLEM OF SHADOW PRESIDENTS
A temporary or consecutive ban on reelection will be substantively
meaningless if a president continues to exercise substantial control over the
executive branch as a “shadow” president. In Federalist 72, Alexander Hamilton
famously relied on this point as an argument against presidential term limits.
319
He highlighted the dangers to political stability of “half a dozen men who had
had credit enough to be raised to the seat of the supreme magistracy, wandering
among the people like discontented ghosts, and sighing for a place which they
were destined never to possess[.]”
320
Recent decades have witnessed the rise of “shadow presidents” in a range
of democratic contexts.
321
In some cases, this trend has been unrelated to the
315. BATURO, supra note 67, at 46 tbl.2.1.
316. See Corrales & Penfold, supra note 31, at 159, tbl.
317. See Ginsburg et al., supra note 1, at 1816–17.
318. See Carey, supra note 53, at 125.
319. See THE FEDERALIST NO. 72 (Alexander Hamilton) (Yale Univ. Press ed., 2009).
320. Id; see also Carey, supra note 53, at 120–21; Ginsburg et al., supra note 1, at 1825.
321. See Baker, supra note 8, at 297–98.
404 HASTINGS LAW JOURNAL [Vol. 71:359
existence of presidential term limits and instead stems from other features of
constitutional design. In Myanmar, for instance, limitations on the capacity of
the president to have family members who are citizens of other countries meant
that the National League for Democracy (NLD) leader and Nobel Prize Winner
Aung San Suu Kyi assumed the role of “state counsellor,” which is effectively a
form of shadow presidency.
322
Other cases, however, have been linked to an attempt to avoid the
substantive effect of a presidential term limit. The leading example in recent
years was the transfer of the presidency from President Vladimir Putin to
President Dmitry Medvedev in 2008 in Russia, as a response to Putin facing
constitutional term limits.
323
Putin, in turn, moved from the presidency to the
post of prime minister.
324
While Medvedev took a somewhat different approach
to some issues, most observers believe he remained strongly under the influence
of Putin.
325
Moreover, the independent support that Medvedev seemed to build
from economic liberals and foreign governments during his tenure was
ultimately insufficient to persuade or allow him to seek a second term as
president.
326
Instead, he stepped aside to allow Putin to run again as soon as he
was constitutionally eligible to do so.
327
Not surprisingly, Putin won reelection
by a large margin both in 2012 and 2018.
328
Even with constitutional term limits
in place, Putin served four terms in office with a one-term break that did little to
limit his power.
329
Recent work by Versteeg et al. found that the installation of a puppet ruler
is a fairly common method for incumbents to evade term limits.
330
Out of sixty
322. See ြပည်ေထာင်စသမတြမန်မာိင်ငံေတာ် ဖွဲစည်းပံအေြခခံဥပေဒ [REPUBLIC OF THE UNION OF
MYANMAR] [CONSTITUTION] May 29, 2008, art. 59(f); Melissa Crouch, Democracy and Peace Frustrated in
Myanmar: Remembering U Ko Ni, LOWY INST. (Jan. 29, 2018), https://www.lowyinstitute.org/the-
interpreter/democracy-and-peace-frustrated-myanmar-remembering-u-ko-ni; Euan McKirdy, New Government
Role Created for Myanmar’s Aung San Suu Kyi, CNN (Apr. 7, 2016, 5:03 AM),
https://www.cnn.com/2016/04/06/asia/aung-san-suu-kyi-state-counsellor-role-created/index.html.
323. See A Putin-Shaped Throne, ECONOMIST (Mar. 6, 2008), https://www.economist.com/
europe/2008/03/06/a-putin-shaped-throne; Ginsburg et al., supra note 1, at 1850; David Hearst & Miriam Elder,
How Dmitry Medvedev’s Mentor Turned Him into a Lame Duck, GUARDIAN (Mar. 2, 2012),
https://www.theguardian.com/world/2012/mar/02/dmitry-medvedev-rivalry; The Odd Couple, ECONOMIST
(July 10, 2008), https://www.economist.com/europe/2008/07/10/the-odd-couple.
324. Seth McLaughlin, Medvedev vs. Putin: Who’s Ruling Russia? WASH. DIPLOMAT (Nov. 18, 2010),
https://washdiplomat.com/index.php?option=com_content&id=6194:medvedev.
325. Id.
326. Id.
327. Andrew Osborn, Medvedev Steps Aside for Putin, FR.24, https://www.france24.com/en/20110926-
2011-09-26-1126-medvedev-putin-russia-elections (last modified Sept. 26, 2011, 2:26 PM).
328. See Shaun Walker, 2018 Election is No Problem for Putin— But What About 2024? GUARDIAN (Feb.
6, 2018), https://www.theguardian.com/world/2018/feb/06/2018-election-is-no-problem-for-putin-but-what-
about-2024.
329. See McLaughlin, supra note 324.
330. See Versteeg et al., supra note 4.
February 2020] CONSTITUTIONAL END GAMES 405
evasion attempts since 2000, six such attempts, which they call “placeholder
presidents,” involved using puppet rulers as the primary method of evasion.
331
Perhaps significantly, the “puppet ruler” method has the lowest success rate
as only two of the six attempts succeeded.
332
This highlights the difficulty of
finding a faithful agent who will be successful in the next election without the
incumbent’s reputation and who will remain faithful to the incumbent’s
interests, rather than developing his or her own power base. Overall, an effective
constitutional design should prevent the implementation of shadow presidents
that undermine the value of consecutive term limits.
A. C
ONSTITUTIONAL DESIGN AND SHADOW RULE
Shadow presidencies are more likely to occur where the incumbent can
continue to hold a formal post from which they can exercise power. In Russia,
for example, President Putin managed to retain prominence in part because of
his continued post as prime minister between 2008 and 2012.
333
One of the
potential disadvantages of semi-presidential, as opposed to pure presidential,
systems, in this context, is that it creates two competing executive positions that
often results in an ambiguous and shifting distribution of powers, such that
officials like Putin and Medvedev can play a game of musical chairs with one
another.
More broadly, the threat of shadow presidencies can be ameliorated by
imposing a temporary formal ban on the ability of former presidents to hold a
range of other elected and appointed offices.
334
Some appointed roles, such as a
foreign ambassador, may raise only a limited risk of shadow influence, and be
broadly compatible with the development of a strong party and alternative leader
domestically.
Alternatively, powerful presidents who leave office due to term limits may
attempt to maintain their power by assuming the role of president or chair of
their own party. Shadow political influence of this kind will be a direct threat to
the development of a true alternative leadership structure within the party.
335
Thus, constitutional designers may consider extending constitutional
prohibitions on alternative office-holding by former presidents to formal roles
within their own political parties. While the internal organization of political
331. Id.
332. Id.
333. See Putin Hints at Becoming Prime Minister Again in 2024, TIMES OF ISR. (May 26, 2018, 4:16 PM),
https://www.timesofisrael.com/putin-hints-at-becoming-prime-minister-again-in-2024/; Vladimir Putin:
Russia’s Action Man President, BBC NEWS (Feb. 27, 2018), https://www.bbc.com/news/world-europe-
15047823.
334. See Carey, supra note 53, at 129–30 (suggesting restrictions in other offices).
335. Carey in fact views this as an argument for rejecting non-consecutive term limits. Carey, supra note
53, at 129 (“The perverse incentives within parties generated by punctuated eligibility suggest that if presidents
are to be barred from reelection at all—whether after multiple terms or only one—they should be barred
permanently.”).
406 HASTINGS LAW JOURNAL [Vol. 71:359
parties has traditionally been left to ordinary democratic politics, constitutional
design and doctrine has increasingly moved towards a preference for direct
regulation in the constitutional text.
336
Limiting the effectiveness of shadow rule may require limits on the ability
of family members of an incumbent president to run for high office. Some
constitutions prohibit a sitting president’s “blood relatives” from seeking high
office. Panama, explored above, is one example.
337
For our purposes, the benefit of a rule preventing family members from
running for office as an incumbent is leaving power is obvious. Family members
of existing presidents have historically been an important route through which
ex-presidents continue to rule behind the scenes and prevent the emergence of
new political power bases.
338
Family members are more likely to be loyal to an
ex-president than even close, but unrelated, political allies.
339
In Senegal, at the same time he was seeking to extend presidential term
limits, President Wade was widely believed to be grooming his son, Karim, to
succeed him as president.
340
Many observers believed that Wade’s decision to
introduce the role of vice president was designed to achieve this form of dynastic
political control.
341
Similarly, in Tajikistan, the constitutional changes that allowed President
Emomali Rahmon to stay in office indefinitely also lowered the minimum age
of presidential candidates from thirty-five to thirty. This was viewed as a bid to
allow his 28-year-old son, Rustam, to succeed him as president.
342
In Panama, many observers suggested that Martinelli’s support for Arias,
on a ticket with Martinelli’s spouse, Linares, as vice president, represented an
attempt to maintain an ongoing form of “shadow” influence over the presidency.
Fifty-six percent of voters in Panama agreed with the statement that the
nomination of Linares represented a disguised reelection attempt by
Martinelli.
343
Many experts raised similar concerns.
344
336. See, e.g., Richard H. Pildes, Political Parties and Constitutionalism, in COMPARATIVE
CONSTITUTIONAL LAW 254 (Tom Ginsberg & Rosalind Dixon eds., 2011); see also Stephen Gardbaum &
Richard H. Pildes, Populism and Institutional Design: Methods of Selecting Candidates for Chief Executive, 93
N.Y.U. L. REV. 647, 670–71 (2018); Samuel Issacharoff, Democracy’s Deficits, 85 U. CHI. L. REV. 485, 485, 501
(2018).
337. See Lees, supra note 267; Zamorano, supra note 266.
338. See Versteeg et al., supra note 4.
339. See id.
340. Estelle Cornado, Senegalese President Abdoulaye Wade’s Rise and Rule, BBC NEWS (Mar. 26, 2012),
http://www.bbc.com/news/world-africa-16905528.
341. Yarwood, supra note 170, at 53.
342. France-Presse, supra note 159.
343. Carin Zissis, Poll Update: Panama’s Ruling-Party Candidate Ahead in Possible Close Race, AM.
SOC’Y/COUNCIL AM. (Mar. 28, 2014), https://www.as-coa.org/blogs/poll-update-panamas-ruling-party-
candidate-ahead-possible-close-race.
344. See Archibold, supra note 279; Zovatto, supra note 289.
February 2020] CONSTITUTIONAL END GAMES 407
In Argentina, there have been several instances of presidents seeking to
prolong their rule via the selection of a spouse. In 1951, Eva Perón considered
running as vice president in an attempt to assume the presidency if her husband,
President Perón, died in office.
345
Eva ultimately withdrew from the race due to
ill health.
346
Peron’s third wife, Isabel Martinez de Perón, however, was able to
serve as vice president from 1973 to 1974, and became president upon Peron’s
death in 1974.
347
More recently, the transfer of power between Nestor Kirchner
and Cristina Fernandez de Kirchner allowed the Kirchners to control the
presidency for a twelve-year period.
348
Some family members may have an independent political base, but most
will be reliant on the outgoing president for political support. This can make
them susceptible to ongoing influence by the outgoing president.
In Argentina, for example, even with her prior record as a successful
senator, many commentators believed that President Cristina Fernandez de
Kirchner would have stepped down to allow her husband, Nestor Kirchner to
seek a further term as president.
349
While Nestor Kirchner chose to step down
after only one term to allow Fernandez to run in his place in 2007, some analysts
perceived that he continued to exercise substantial influence over her
presidency.
350
It was only after Kirchner died in 2010 that many commentators
believed she began to exercise power fully in her own right.
351
Thus, effective term limits may require temporary limits on the ability of
close family members of presidential incumbents to run for high office. Limits
of this kind are problematic because they explicitly limit the freedom of political
participation of other individuals, including family members, on grounds we
generally think are quite suspect: family identity and marriage.
For spouses particularly, a ban of this kind may have gender-discriminatory
effects. As the Argentinian and U.S. experience has shown, women have
historically been socially encouraged to prioritize the political careers of their
345. Saga of Eva Peron: 12 Years to Power, N.Y. TIMES (July 27, 1952),
https://archive.nytimes.com/www.nytimes.com/learning/general/onthisday/bday/0507.html.
346. Id.
347. Melissa Gomez Hernandez, Latin American Women in Public Service: Progress, but Not Yet Equality,
in GOVERNING IN A GLOBAL WORLD: WOMEN IN PUBLIC SERVICE 130–31 (Maria J. D’Agostino & Marilyn
Marks Rubin eds., 2018).
348. See Erasing the Kirchner Cult, ECONOMIST (June 30, 2016), https://www.economist.com/the-
americas/2016/06/30/erasing-the-kirchner-cult; First Lady Runs in Argentine Poll, BBC NEWS (July 20, 2007),
http://news.bbc.co.uk/2/hi/americas/6906563.stm; Uki Goñi, Cristina Fernández de Kirchner: Is the Fairytale
Ending for Argentina’s New Evita?, GUARDIAN (Feb. 21, 2015, 7:44 AM),
https://www.theguardian.com/world/2015/feb/20/cristina-fernandez-de-kirchner-argentina-president-political-
turmoil; Nick Miroff, “Kirchnerisomo” Politics Set to Outlast Argentina’s Outgoing President, GUARDIAN (July
21, 2015, 6:00 AM), https://www.theguardian.com/world/2015/jul/21/cristina-fernandez-kirchnerismo-
argentina-election.
349. See Daniel Schweimler, Kircher Death Leaves Fernandes to Run Argentina Alone, BBC NEWS, (Nov.
2, 2010) https://www.bbc.com/news/world-latin-america-11673444.
350. Id.
351. See Miroff, supra note 348; Schweimeir, supra note 349.
408 HASTINGS LAW JOURNAL [Vol. 71:359
spouse, even when they also have political ambitions.
352
Preventing them from
running for president, or at least doing so for a significant period, is thus likely
to have a disparate impact based on gender. On the flip side, as we have noted,
353
power transitions between family members are most likely to raise successful
threats of shadow presidencies.
While there is a strong, but contestable, case for temporary bans on family
members and spouses, we would be reluctant to sweep beyond these well-
defined categories to include non-relatives such as the “associates” of an
outgoing president. Limits on the electoral eligibility of political associates are
more difficult to define with precision than limits on family members. Bans that
focus only on close formal advisors, or members of a president’s staff, may be
readily overcome by a president selecting a successor from among their informal
network of supporters and associates. Conversely, if bans of this kind are drawn
too broadly, they may unjustifiably limit rights to freedom of political
participation and association. More importantly, such bans undermine a
president’s ability to find a successor from within his or her party, and therefore
weaken his or her incentive to leave office at the constitutionally appointed time.
The dangers of shadow rule are also likely to be reduced for associates, as
compared to family members. Non-familial associates have a stronger chance of
developing the kind of independent political reputation and support base over
time that is needed to resist attempts at ongoing influence by a former president.
Recent history includes a number of examples of trusted or handpicked allies
that later turned against their former mentors: consider Correa and Moreno in
Ecuador, or Uribe and Santos in Colombia.
354
We thus would not recommend
extending electoral bars beyond immediate family members of an incumbent
president.
B. C
REATING INCENTIVES TO AVOID PARTISAN BEHAVIOR
Constitutional term limits will work best when they are combined with
other rules that incentivize presidents to leave office and remain outside partisan
politics for the relevant period of non-eligibility for reelection. Some of the most
successful instances of presidential term limit “enforcement” have occurred in
contexts where key players seem to have been mindful of these concerns. As a
reward for staying out of partisan politics and allowing a true democratic
alternative to emerge, former presidents have been appointed prestigious roles
with the UN and the African Union.
355
Thus, it could be beneficial to build on this experience and create more
formal, global roles for former presidents, as opposed to roles limited to only
352. See Schweimeir, supra note 349.
353. See supra Subpart VI.A.
354. Supra text accompanying note 299; Reyes, supra note 88.
355. Roger Southall et al., Former Presidents in African Politics, in LEGACIES OF POWER, supra note 303,
at 1, 12–15.
February 2020] CONSTITUTIONAL END GAMES 409
national governance. Scholars such as Maltz, for example, suggest that
presidential term limits could be strengthened by building on the U.S. experience
and providing former presidents with a range of benefits, such as generous
retirement benefits, assistance in setting up a presidential library or personal
foundation, a role in official state ceremonies, and access to state benefits like
bodyguards, cars, and airport transport.
356
Arrangements of this kind, as Maltz, Southall, Simutanyi, and Daniel note,
are likely to work best if they are conditioned on a president having a track
record of respect for democratic constitutional norms while in office.
357
Otherwise, the entities that offer these arrangements would lose credibility and
undermine their capability to motivate potential future members to relinquish
domestic political benefits.
But at the same time, too strict an insistence on a perfect human rights and
democratic record from former leaders can be counterproductive. If certain
historical wrongs permanently bar a president from being considered for
membership in such an institution, there will be little incentive for current actors
to adhere to democratic constitutional norms.
An even harder question, but one which cannot be avoided, is the impact
of constitutional term limits on domestic and international systems of criminal
accountability for a range of wrongs that presidents may commit while in office.
Too great a weakening of these norms undermines commitments to democratic
constitutionalism, human rights, and the rule of law. But too absolute an
insistence on such norms can undermine domestic attempts to enforce national
constitutional limits and discourage presidents from leaving office.
358
Thus,
there is, at least in some cases, a relationship between these norms and
presidential incentive to evade term limits.
How this balance should be struck is a significant issue largely beyond the
scope of this Article. One possibility is the exercise of prosecutorial discretion,
which takes into account a former president’s compliance with democratic
constitutional limits. While compliance with presidential term limits should not
shield a president from future criminal prosecution, a president’s peaceful
transfer of power may be a relevant factor in deciding whether to ultimately
prosecute them.
VII.
ALTERNATIVE PROPOSALS
We have laid out the case for why systems vulnerable to term limit evasion
might want to consider using consecutive, rather than absolute, bars on
reelection. We have also analyzed some of the constitutional design choices
relevant to that decision.
356. Maltz, supra note 1, at 140.
357. See Maltz, supra note 1, at 129; Southall et al., supra note 355, at 5.
358. See Southall et al., supra note 355, at 18–21.
410 HASTINGS LAW JOURNAL [Vol. 71:359
In this Part, we briefly consider three alternative solutions described in
recent work on constitutional design: the proposal to scrap term limits
completely, the proposal to rely on popular enforcement of term limits, and the
proposal to require that any loosening to term limits may only be made
prospectively so that it does not benefit the incumbent.
A. S
CRAPPING TERM LIMITS
In response to the problem of non-compliance with presidential term limits,
some have argued for the removal of presidential term limits altogether and that
other forms of control should be relied upon. Smith, Crowley, and Leguizamon,
for example, suggest that age and death from natural causes are effective limits
on presidential terms.
359
An alternative approach could be to impose formal age limits on presidents,
although we note that these kinds of provisions can create similar pressures for
evasion as term limits themselves.
360
Many presidents seeking a third or fourth
term in Africa, for example, are already in their seventies and eighties.
Cameroon’s President Paul Biya was seventy-five at the time he sought a third
term in office and Congo-Brazzaville’s President, Sassou-Nguesso, was
seventy-one when he sought another term.
361
In the Congo, in order for President
Sassou-Nguesso to remain in office, the Constitution was changed so that the
two-term limit was removed, and the maximum age was extended for
presidential candidates.
362
Furthermore, reliance on a president’s eventual mortality is an insufficient
response because it does not take into account young, healthy presidents who
are serving during their prime. In Africa, Rwanda’s President Paul Kagame was
fifty-seven when the Parliament amended the Constitution to allow him to seek
two further consecutive terms in 2015.
363
Burundi’s President Nkurunziza was
only fifty-four when voters approved the repeal of term limits that extended the
length of presidential terms in 2018.
364
Similarly, in Burkina Faso and
Madagascar, Presidents Blaise Compaoré and Marc Ravalomanana were in their
early sixties when they reached the end of their second terms.
365
359. See Smith et al., supra note 21.
360. See Turkmenistan Removes Legal Barrier to Leader’s Indefinite Rule, REUTERS (Sept. 13, 2016, 11:45
PM), https://www.reuters.com/article/turkmenistan-president-idUSL8N1BQ0NR. In Turkmenistan, for
example, voters recently voted to repeal the age limit (of seventy) on the presidency, thereby allowing President
Kurbanguly Bedyukhamedov to stay in power indefinitely. See id.
361. Ross, supra note 125.
362. Tull & Simons, supra note 53, at 80, 93.
363. Adam Agustyn, Paul Kagame: President of Rwanda, ENCYCLOPEDIA BRITANNICA,
https://www.britannica.com/biography/Paul-Kagame (last updated Sept. 27, 2019).
364. Dahir, supra note 138.
365. Amy McKenna, Blaise Compaoré: President of Burkina, ENCYCLOPEDIA BRITANNICA,
https://www.britannica.com/biography/Blaise-Compaore (last visited Jan. 24, 2020); Marc Ravalomanana:
President of Madagascar, ENCYCLOPEDIA BRITANNICA, https://www.britannica.com/biography/Marc-
Ravalomanana (last visited Jan. 24, 2020).
February 2020] CONSTITUTIONAL END GAMES 411
The same is true elsewhere. In Latin America, Rafael Correa was only fifty-
two when the Constitution was amended in 2015 to eliminate presidential term
limits.
366
Bolivia’s Evo Morales is only fifty-nine years old, has held power
continuously since 2006, and now will be able to run for office indefinitely.
367
Turkmenistan also recently voted to remove the age limit for the presidency,
allowing fifty-nine-year-old President Kurbanguly Berdymukhdamedov to
continue to run for office in future elections.
368
Ginsburg, Melton, and Elkins suggest that informal norms or conventions
may be a better alternative to formal constitutional limits.
369
For example,
George Washington famously retired from the presidency after two terms,
thereby creating a de facto convention that a president would only serve two
presidential terms.
370
This was well before the adoption of the Twenty-Second
Amendment.
371
Some of the most notable examples of voluntary presidential
retirement in Africa occurred in South Africa and Tanzania before or in the
absence of formal, relevant term limits.
372
In Tanzania especially, this seems to
have contributed to an ongoing “culture” of voluntary presidential retirement.
373
President Julius Nyerere voluntarily stepped down from the presidency of
Tanzania in 1985 after a lengthy period as president.
374
Both Presidents
Benjamin Mkapa and Jakaya Kikwete followed the same pattern and stepped
down at the end of the two-terms permitted by the Constitution in 2005 and
2015, respectively.
375
There are encouraging signs that the current Tanzanian
President John Magufuli will take the same approach and leave office by no later
than 2025, rather than support calls to repeal or remove current term limits.
376
In some contexts, there are good arguments for making constitutional
limitations informal and voluntary, rather than formal and legally mandated.
Government officials may be more likely to comply with certain requirements
when they are trusted to do so, instead of being constitutionally directed or
366. Rafael Correa, President of Ecuador, ENCYCLOPEDIA BRITANNICA,
https://www.britannica.com/biography/Rafael-Correa (last visited Jan. 24, 2020).
367. Evo Morales, President of Bolivia, ENCYCLOPEDIA BRITANNICA,
https://www.britannica.com/biography/Evo-Morales (last visited Jan. 24, 2020).
368. Turkmenistan Removes Legal Barrier to Leader’s Indefinite Rule, supra note 360.
369. Ginsburg et al., supra note 1, at 1859.
370. Ginsburg et al., supra note 1, at 1834; Maltz, supra note 1, at 130–31.
371. Ginsburg et al., supra note 1, at 1834–35; Maltz, supra note 1, at 130–31.
372. Tull & Simons, supra note 53, at 88.
373. Id. This may reflect path-dependencies in constitutional practice. See generally Ozan O. Varol,
Constitutional Stickiness, 49 U.C. DAVIS L. REV. 899 (2015) (generally discussing constitutional path
dependency).
374. Tull & Simons, supra note 53, at 88.
375. Id.
376. Jerry Bambi, Presidential Term Limit Debate Put to the Rest in Tanzania, AFR. NEWS (Jan. 17, 2018),
http://www.africanews.com/2018/01/17/presidential-term-limit-debate-put-to-rest-in-tanzania-the-morning-
call/; Fumbuka Ng’wanakilala, Tanzania’s Magufuli Rejects Calls to Extend Rule Beyond Two-Term Limit,
REUTERS (Aug. 7, 2017, 9:15 AM), https://www.reuters.com/article/us-tanzania-politics/tanzanias-magufuli-
rejects-calls-to-extend-rule-beyond-two-term-limit-idUSKBN1AN1YT.
412 HASTINGS LAW JOURNAL [Vol. 71:359
mandated to act a certain way. This reflects broader human tendencies to
reciprocate expressions of trust and the relatively weak effect of constitutional
constraints.
377
It is dangerous, however, to rely on informal norms to encourage or enforce
constitutional term limits. The fact that presidents are facing the end of their
term, and thus the end of an ongoing political give-and-take relationship, gives
rise to the problem of constitutional non-compliance in the first place. Often,
there will be exogenous factors, such as war, or other political or economic
crises, that may encourage a president to remain in office and depart from
informal conventions of this kind. It was arguably pressures of this kind that led
President Franklin Delano Roosevelt to overlook the norm of a two-term
presidency and stand for a third and then fourth presidential term during the
Great Depression and World War II.
378
Similar arguments about the need for
economic and political stability have been used to justify the repeal of term
limits in other contexts.
Ginsburg, Melton, and Elkins have also suggested that it may be preferable
to replace term limits with a form of “super-majoritarian” escalator, whereby
presidents are required to gain increasing majorities the longer they are in
office.
379
However, such a proposal raises problems of incentives that could be
even more dramatic than those posed by an absolute ban on reelection.
Presidents who win a majority of the popular vote may have weak incentives to
accept the results of an election that they fail to win with a sufficient
supermajority.
380
Provisions of this kind therefore raise the risk of presidents
engaging in forms of constitutional over-staying that are even more dangerous
to democracy than the circumvention of term limits.
B. P
OPULAR ENFORCEMENT OF TERM LIMITS
In recent work, we argue that enforcement of “tiered” systems of
constitutional change, such as eternity clauses that protect provisions like term
limits may be more likely to occur through popular means, rather than resorting
to the courts.
381
Versteeg et al. take this position forcefully, noting that there has
only been one successful example since 2000 of a court blocking an attempt to
change a constitutional term limit.
382
In 2010, the court successfully blocked
377. Compare Rosalind Dixon, Constitutional Drafting and Distrust, 13 INT’L J. CONST. L. 819, 837 (2015),
with Rosalind Dixon, Constitutional Design Two Ways: Constitutional Drafters as Judges, 57 VA. J. INT’L L. 1,
31 (2017).
378. Carey, supra note 53, at 121.
379. Ginsburg et al., supra note 1, at 1861–63.
380. Id.
381. Dixon & Landau, supra note 10, at 507 (“Given the challenges associated with judicial enforcement of
limits on amendment in many common circumstances, constitutional designers and scholars should also be
attuned to the possibility of nonjudicial routes through which amendment tiers might gain force.”).
382. Versteeg et al., supra note 4.
February 2020] CONSTITUTIONAL END GAMES 413
President Uribe’s attempt to amend his presidential term limit.
383
The
Colombian case, however, relied on specific conditions that may not commonly
be present at moments of stress where incumbents seek to eliminate term limits.
First, in that case, the court retained independence and power and had not been
packed or otherwise placed under the thumb of the president.
384
Second, Uribe’s
congressional coalition was quite fragmented and many of his rivals within his
movement welcomed the decision because it created a new opportunity for
them.
385
It is common for courts to support incumbents seeking to stay in office
beyond their constitutionally appointed term by removing term limits.
386
Indeed,
as we noted above, the courts are a fairly common instrument used by
incumbents to stay in power.
387
In contrast, there have been failed attempts to change presidential term
limits because of popular protests. In Paraguay, as noted above, an attempt to
extend term limits in a clearly unconstitutional manner was scrapped because of
massive protests backed by the Catholic Church and a broad swath of civil
society.
388
In Ecuador, similarly, in the face of huge protests, Correa assented to
the crucial provision requiring that he stand down in 2017.
389
Similar dynamics
have worked in Africa, for example in Burkina Faso in 2014 and Malawi in
2002.
390
As a descriptive matter, there is evidence that popular involvement is often
critical to halting attempts to evade or to change presidential term limits.
391
The
trouble is determining how to leverage popular involvement into principles of
constitutional design. Most of the factors that determine levels of popular
uprising are extrinsic to constitutional design. In our earlier paper, we have ideas
for how designers might make popular mobilization around constitutional
protections for term limits especially likely.
392
One solution is to design term limit provisions such that any ambiguity is
removed. Provisions that unequivocally identify the term limit by name and that
clearly require a stringent, heightened procedure to change the term limit can
better counteract the weaknesses of vague constitutional designs.
393
The latter
kind of design is really a delegation to courts through legalistic language, and as
noted, courts are not to be trusted during these situations.
383. See Corte Constitucional [C.C.] [Constitutional Court], febrero [Feb.] 2, 2010, Expediente [Decision]
C-141/10 (Colom.).
384. Dixon & Landau, Tiered Constitutional Design, supra note 10, at 505.
385. Id.
386. See Ginsburg et al., supra note 1, at 1811, 1823.
387. See supra Part III.
388. See supra text accompanying notes 184–187.
389. See supra text accompanying notes 144–148.
390. See Versteeg et al., supra note 4.
391. See, e.g., supra text accompanying notes 171–177, 186.
392. See Dixon and Landau, supra note 10, at 478–79.
393. Id. at 510–11.
414 HASTINGS LAW JOURNAL [Vol. 71:359
A second possibility is to require popular referenda to change term limits.
There are recent examples of popular presidents losing referenda to extend or
eliminate term limits, including President Chavez of Venezuela in 2007
(although Chavez later succeeded in 2009), and President Morales of Bolivia in
2015 (although Morales later convinced a court to remove the limit).
394
Presidents Chavez and Morales were popular but their proposals were less so.
However, referenda are famously malleable instruments in regimes where
democracy is absent or threatened.
395
There are plenty of other cases where
referenda to eradicate term limits passed overwhelmingly. At any rate, voting in
a popular referendum does not have the same effect as the sorts of popular
protests that have successfully prevented a president’s attempt to evade term
limits.
The bottom line is that we have some distance to go in identifying how
constitutional design can support popular mobilization in these kinds of
moments. It is likely that constitutional design can only do so much. Proposals
to encourage popular mobilization, however, are fully concordant with our own
design proposal in this Article.
C. P
ROSPECTIVE-ONLY RULES OF CHANGE
A final alternative solution focuses squarely on constitutional design. One
possibility is for constitutions to include a rule allowing changes to term limits,
but also stating that any changes will only take effect after the incumbent has
left office. Consider, for example, the constitution of South Korea, which states
that “[a]mendments to the Constitution for the extension of the term of office of
the President or for a change allowing for the reelection of the President shall
not be effective for the President in office at the time of the proposal for such
amendments to the Constitution.”
396
This provision states in clear language that any change to the presidential
term length or limit will not benefit the president who is the incumbent at the
time the change is made.
397
Other constitutions may achieve a similar effect
through the use of temporal restrictions on amendments. Where constitutional
changes to term limits take a long time to carry out or require sequential votes
with intervening elections, they may also make it unlikely or impossible for the
current president to benefit from the change.
398
There is a certain appeal to these kinds of rules. They ensure that changes
to term limits occur because of their long-term benefits and costs, and not as a
way to help a current president remain in office. They seek to ensure, in other
words, that constitutional changes to term limits occur behind something of a
394. Id. at 501.
395. Id.
396. 1987 DAEHANMINKUK HUNBEOB [HUNBEOB] [CONSTITUTION] art. 128(2) (Oct. 29, 1987) (S. Kor.).
397. Id.
398. For a discussion of temporal restrictions on sensitive amendments, see Dixon and Landau, supra note
10, at 502–03.
February 2020] CONSTITUTIONAL END GAMES 415
veil of ignorance, such that the beneficiary of the change remains ambiguous or
unknown. These kinds of constitutional provisions may be very useful
throughout constitutional law and remain understudied. The inclusion of a
prospective-only rule for constitutional change is also complementary to our
own proposal. A bar against consecutive terms can be bolstered if placed on a
higher tier with a prospective-only rule for change.
However, we are skeptical that a provision like the South Korean one,
standing alone, will prove to be a robust barrier against term limit abuse. While
it creates an obstacle for presidents seeking to remain in power, it does little to
ameliorate the powerful incentives that presidents often have to evade term
limits. As a result, presidents may turn to substitutes for formal amendments to
achieve their goals. For example, presidents could turn to the courts, which could
reinterpret the provision or remove it. There are arguments, grounded in
principles of equality or on the will of voters, that would potentially allow courts
to do so.
399
Relatedly, we note that prospective application of term limits is also a
double-edged sword that has often been abused by incumbents. Would-be
authoritarian leaders in several countries have allowed new term limits to go into
effect, but have then either included explicit provisions stating that the new
limits apply only prospectively, or convinced courts to adopt similar
interpretations.
In Angola, Equatorial Guinea, and Zimbabwe, for example, term limits
were only added to the constitution on a prospective basis in 2010, 2011, and
2013, when Presidents Dos Santo, Mugabe, and Obiang Nguema had all been in
power for more than two decades.
400
In Sudan, a term limit was introduced in
2005, but was not implemented until 2010. This allows President al-Bashir, who
has been in power since 1993, to potentially remain in power until 2020.
401
In Burkina Faso, President Compaoré, who has been in power since 1983,
succeeded in removing constitutional term limits in 1997.
402
He agreed to re-
introduce them in 2000 but again, only on a prospective basis so he could remain
in office until at least 2015.
403
In Rwanda, the term limit changes introduced in
2015 allowed President Kagame to serve an initial “transitional” seven-year term
399. For example, in Venezuela the Supreme Court held that a legislative effort to shorten presidential terms,
in a way that would apply to the incumbent President Maduro’s current term, would be an unconstitutional
constitutional amendment, using these kinds of arguments. See José Ignacio Hernández G., La
Desnaturalización de la Justicia Constitucional en Venezuela Durante el 2016 [The Denaturalization of
Constitutional Justice in Venezuela During 2016], LA UNIÓN [THE UNION] 1, 22 (2016) (Venez.)
400. Tull & Simons, supra note 53, at 91.
401. Id. at 91 fig.3.
402. Id. at 91–92.
403. Id. at 92.
416 HASTINGS LAW JOURNAL [Vol. 71:359
beginning in 2017, followed by two further five-year terms,
404
thereby
potentially allowing him to stay in office for a total of thirty-one years.
405
Constitutional courts have sometimes adopted a similar interpretation of
term limits at the behest of incumbents. In Bolivia, for example, well before
Morales sought to loosen term limits, the Bolivian Constitutional Court held that
the two-term limit in the new 2009 Constitution did not apply to Morales’s
term.
406
Effectively, this allowed Morales to serve three consecutive terms in
office, rather than two terms, before even seeking to change the limit. Similarly,
in Burundi, the 2015 decision of the Constitutional Court on term limits
effectively postponed their effect so they only became binding after an
additional five years in office.
407
The pervasiveness of these abuses of prospective application of term limits
may reduce our confidence in their viability in constitutional design. Ideally,
there would be a norm mandating that loosening of a presidential term limit must
be prospective such that any tightening of the term limit could be applied to
incumbents. At the moment, however, no such norm exists.
C
ONCLUSION
Timing is one of the key mechanisms available to constitutional designers
in entrenching democracy. We have noted elsewhere that “tiered” systems of
constitutional change that make certain kinds of change more costly work
largely by raising the political costs, or level of political support, needed for anti-
democratic constitutional change.
408
This has the effect of both deterring and
delaying such change. Delay can also be built into constitutional design in a
more explicit sense; double passage requirements and similar devices may slow
the actions of would-be authoritarians, and this may create space for them to lose
power before they can consolidate control.
Delay, by itself, does not guarantee that democracy will survive. It may
simply slow an inevitable process of democratic backsliding. However, delay
can help create a window for broader social or political change, which can
undermine democratic support for would-be authoritarian actors or policies.
A similar logic applies to temporary presidential term limits. Relatively
“weak” bans on consecutive reelection are more likely to be effective than
absolute bans on all reelection because they deal with the constitutional end-
game problem we have identified. They give incumbents hope of a potential
return to power later, which may make those incumbents less likely to seek
404. Rwandans Vote on Allowing Third Kagame Presidential Term, BBC NEWS (Dec. 18, 2015),
http://www.bbc.com/news/world-africa-35125690.
405. Rwanda Changes Constitution to Allow President to Extend His Rules Until 2034, supra note 131;
Rwanda: Paul Kagame Is in Line to Stay in Office Until 2034, supra note 132.
406. Bolivia: New Law Backs President Evo Morales Third Term, BBC NEWS (May 21, 2013),
https://www.bbc.com/news/world-latin-america-22605030.
407. Tull & Simons, supra note 53, at 92.
408. See Dixon & Landau, supra note 83, at 613–14 (arguing that doctrines raising the cost of constitutional
change sometimes act as speed bumps).
February 2020] CONSTITUTIONAL END GAMES 417
extraordinary measures to alter a term limit through either formal or informal
means.
At the same time, such bans protect democracy because they remove
presidents from power at least for a set period of time. This delay, in turn, helps
reduce the actual chances a president will be elected to a third or fourth term by
creating space for opposition figures and independent voices within the
president’s own party. The experience of presidential alternation in Tanzania,
for example, has not been one of alternation between parties, as all three
presidents that have stepped down at the expiration of their term were from the
same political party.
409
However, the experience nonetheless helped build a
stronger ruling party, as well as democratic traditions and norms.
410
Furthermore, even if a president is eventually reelected to a non-consecutive
term, they will likely pose less of a threat to democracy than if they had held
power continuously because the time they spend out of office reduces their
ability to accumulate formal and informal forms of control.
In this light, the U.S. presidential term limit is emblematic only in certain
respects and poorly designed in others. While it lacks the special form of
entrenchment or tiered constitutional design found in many modern orders, the
term limit is nonetheless highly entrenched because Article V itself makes all
formal constitutional change of the U.S. Constitution extremely difficult.
411
We
agree with others who have argued that the formal amendment of the U.S. term
limit provision for the benefit of a would-be authoritarian president is
unlikely.
412
In most political conditions, political minorities would maintain
sufficient power at the congressional or state-legislative level to block the
proposal.
413
Even in the worst case, the legislature would likely slow the
proposal, since the United States’ process is time-consuming.
414
However, this design may actually exacerbate, rather than ameliorate, the
end game problem identified in this article. Would-be authoritarian presidents
cannot formally amend the term limit but will still seek to stay in power. Thus,
they have very strong incentives to seek other means of change, such as
manipulation of the judiciary or wholesale replacement of the constitution. Thus,
the design of term limits in the U.S. Constitution may inadvertently raise risks
of authoritarianism. Perhaps most important, this feature makes the U.S. design
a perilous model for countries abroad. In many contexts, we have argued, a
weaker, consecutive ban on holding power for more than two terms may make
more sense than an absolute bar to reelection.
409. See Baker, supra note 8, at 298; Posner & Young, supra note 100, at 128–29; Southall et al., supra
note 355, at 14; Vandeginste, supra note 139, at 59.
410. Tull & Simons, supra note 53, at 88.
411. See, e.g., Lutz, supra note 195, at 362 (arguing that the U.S. Constitution is the most difficult in the
world to amend).
412. See, e.g., Ginsburg & Huq, supra note 3, at 143–44 (arguing that the rigidity of Article V makes formal
amendment unlikely as part of a program of democratic retrogression in the United States).
413. Id. at 145–46.
414. Id. at 143–44.
418 HASTINGS LAW JOURNAL [Vol. 71:359
***
