Bermeo, On Democratic Backsliding
Bermeo, On Democratic Backsliding
Bermeo, On Democratic Backsliding
Nancy Bermeo
Journal of Democracy, Volume 27, Number 1, January 2016, pp. 5-19 (Article)
On Democratic Backsliding
Nancy Bermeo
ing six of its major varieties. It illustrates that forms have varied in
frequency over time; that some of the most blatant forms of backslid-
ing are now less common; and that more vexing forms of backsliding
are becoming more common. Ironically, we now face forms of demo-
cratic backsliding that are legitimated through the very institutions that
democracy promoters have prioritized. Overall, trends in backsliding
reflect democracy’s slow progress and not its demise.
A close historical look at the varieties of backsliding reveals that the
classic open-ended coups d’état of the Cold War years are now out-
numbered by what I call promissory coups; that the dramatic executive
coups of the past are being replaced by a process that I call executive
aggrandizement; and finally, that the blatant election-day vote fraud that
characterized elections in many developing democracies in the past is
being replaced by longer-term strategic harassment and manipulation.
According to the Oxford English Dictionary, the earliest use of the
English word “backsliding” dates from 1554, when the Scottish Protes-
tant theologian John Knox (1513–72) employed it in a pamphlet entitled A
Faythfull Admonition to the Professors of God’s Truth in England. More
famously, “backsliding” also appears in the King James Bible (1611),
where it translates a prophetic plea (Jeremiah 3:22) for Israel to drop its
“faithless” or “wayward” habits in order to resume a relationship of loyalty
to God. When linked with the word democratic, the term’s current secular
meaning is in keeping with its origins in that it denotes a willful turning
away from an ideal. But where does backsliding from democracy lead?
Backsliding can take us to different endpoints at different speeds.
Where backsliding involves rapid and radical change across a broad
range of institutions, it leads to outright democratic breakdown and to
regimes that are unambiguously authoritarian. Where backsliding takes
the form of gradual changes across a more circumscribed set of institu-
tions, it is less likely to lead to all-out regime change and more likely to
yield political systems that are ambiguously democratic or hybrid. Dem-
ocratic backsliding can thus constitute democratic breakdown or simply
the serious weakening of existing democratic institutions for undefined
ends. When backsliding yields situations that are fluid and ill-defined,
taking action to defend democracy becomes particularly difficult.
Positive Trends
Democratic backsliding has changed dramatically since the Cold
War. Three of the most dramatic and far-reaching varieties of backslid-
ing seem to be waning. Coups d’état, executive coups by elected leaders,
and blatant election-day vote fraud all have declined in frequency.
The decline of classic coups d’état. Coups are illegal attempts by mili-
tary or other state elites to oust a sitting executive. Historical analysis shows
Nancy Bermeo 7
30
25
Democracies
20
15
10
5
0
Year
Source: Regime data are from Polity IV, supplemented with Polity IV data modified by
Kristian Skrede Gleditsch (see endnote 6). Coup data are from Jonathan Powell and Clay-
ton Thyne’s “Coups d’état, 1950 to Present” dataset (www.uky.edu/~clthyn2/coup_data/
home.htm). A detailed explanation of the axes for this figure and how the data sources
were used may be found at www.journalofdemocracy.org/articles/supplemental-material.
a dramatic decline in all coups and especially the open-ended military coups
that gave rise to long-lasting and brutal dictatorships during the Cold War.
As Figure 1 shows, the probability that a democracy will be targeted by any
sort of coup has dropped dramatically. The probability reached a thirty-year
low after 1995, and although it rose slightly as the first decade of the new
century ended, it is still significantly less than it was during the 1960s.
The likelihood of a democratic government being the target of a suc-
cessful coup has also declined markedly, dropping to nearly zero in the
early 2000s. Though it has recently risen slightly, the drop in the success
rate that began during the Cold War has not been reversed.
12
10
8
Elections
6
4
2
0
1991–95 1996–2000 2001–2005 2006–2010 2011–2012
Year
Source: See source note in Figure 1. Additional election data are from Susan Hyde’s
NELDA project (http://hyde.research.yale.edu/nelda/). A detailed explanation of the axes
for this figure and how the data sources were used may be found at www.journalofdemoc-
racy.org/articles/supplemental-material.
a whole has not diminished, but there is near-consensus that open fraud
on election day has decreased. Figure 2 shows the trend. The extent
to which this trend is being driven by normative change, the rise of
election monitoring, or the deterrent effects of parallel vote tabulation
remains under debate, but the decline itself has been widely noted. Sea-
soned election observers report that cheating has “become more subtle,”
that “blatant manipulation on election day seems less and less common,”
and that fraud in polling stations has been reduced.3 A recent study of
African elections found that count falsification, ballot-stuffing, and bal-
lot-box fraud were relatively rare and that “the vote count process was
the most highly regarded dimension in the whole electoral process.”4 In
the words of another firsthand observer, “Today, only amateurs steal
elections on election-day.”5
Continuing Challenges
The decline in the three varieties of backsliding outlined above is
certainly gratifying. Unfortunately, other varieties of democratic back-
sliding either remain unchanged or are on the rise. These have been
understudied and merit our immediate attention.
When surrogates took their respective places, the surrogate for the coup
leader won 54 percent in the December 2013 runoff, while his party won
a plurality of seats in Parliament. The presidential elections held after
military officers toppled Mali’s democracy in March 2012 were won by
Ibrahim Boubacar Ke¦ta, a civilian politician whose brief membership in
an anticoup coalition did not eclipse his longtime status as a military fa-
vorite. In August 2013, he won the presidency in a runoff landside, while
his party took an overwhelming legislative majority in November.
In a more recent case, Fiji’s 2014 postcoup parliamentary election
was won by Frank Bainimarama, the very officer (and onetime com-
mander of Fiji’s tiny patrol-boat navy) who had headed the 2006 putsch.
Though he ruled by decree for over seven years, his Fiji First party won
59 percent of the vote and 64 percent of the seats in Parliament.
Had coupmakers and their allies fulfilled their promises for improved
democracy, this subtype of backsliding might be said to have an upside.
But an example of democratic deepening after a coupmakers’ victory is
yet to be found. It is too early to know if the new government in Fiji will
be able to claim such an achievement. Regrettably, none of the other
cases mentioned above has even matched (let alone exceeded) the level
of freedom they enjoyed before their coups.8
The promised improvement of democracy has remained elusive even
when coup opponents have won postcoup elections. Promissory coups
in Lesotho (1994), Niger (1996), Pakistan (1999), Thailand (2007), and
Guinea-Bissau (2012) were all followed by elections in which coup op-
ponents proved victorious. Yet in only one case—that of Lesotho—has
a substantial improvement in political and civil rights been recorded. In
2003, Freedom House’s rating system moved Lesotho from Partly Free
to Free, but Guinea-Bissau, Niger, and Pakistan remain Partly Free and
even this status remains precarious.
Freedom House deemed Thailand a Free country before its promis-
sory coup in 2006, but it has since returned to dictatorship. The coup co-
alition that ousted the freely elected government of Thaksin Shinawatra
made good on its promise to hold free elections (in December 2007) and
even allowed Thaksin allies to regain power through the ballot box. But
tolerance was short-lived. The military seized power again in May 2014,
and, ominously, made no promise of elections at all. Unlike other forms
of backsliding, promissory coups sometimes raise expectations at home
and abroad, but these expectations are nearly always dashed.
challenged military and civilian elites with less than perfect democratic
credentials of their own, they cut through the old order with what even
critics describe as “a democratizing edge.”14
The same can be said of many of the initiatives taken by President Ra-
fael Correa in Ecuador. Like Erdo¢gan, Correa has changed the institutions
of democracy in basic ways, but always with an electoral mandate that has
averaged almost 56 percent across three elections spanning 2006 to 2013.
True to his vow to lead a “Citizens’ Revolution,” Correa has made pro-
found changes in a panoply of democratic institutions. These began with his
2007 initiative to hold elections for a Constitutional Assembly. A stunning
82 percent of the electorate endorsed his proposal in a referendum and a
weighty 64 percent of the public endorsed the new constitution in a 2008
plebiscite. Correa cautioned that he and his supporters “had won the elec-
tions, but not power,” and he immediately set out to consolidate the latter.15
Correa convinced the newly elected Constitutional Assembly to force
the seated Congress into permanent recess and to assume legislative
functions itself. Many of the established parties, discredited by corrup-
tion and poor performance, never recovered. Correa sealed their fate
through recentralizing measures meant to undermine conservative oppo-
sition elites in Guayaquil, and through changing the rules for licensing
parties, drawing electoral districts, and allocating seats.16
Coupled with the undeniable success of a series of redistributive pro-
grams that led to unprecedented drops in poverty and income inequal-
ity, these legal initiatives crippled Correa’s legislative opposition. Cor-
rea’s party, Alianza País, won a whopping 73 percent of the seats in the
2013 legislative elections while its strongest competitor won less than
9 percent.17 Since the 2008 Constitution allows full amendments with a
two-thirds majority, the legislature’s December 2015 vote to eliminate
presidential term limits came as no surprise.
Correa has used his strong mandate to make other major changes as
well. Banks and bank shareholders may not own media outlets; broad-
cast frequencies must operate in “the collective interest”; media outlets
and journalists are legally liable if the information they disseminate is
not deemed “true, verified, opportune [and] contextualized”; and no cov-
erage may be aimed at “destroying the prestige of a natural or juridical
person or reducing their public credibility.” Any citizens’ organization
can be dissolved if the state deems it divorced from its original purpose,
harmful for state security, or disruptive of the “public peace.”18
Though both Erdo¢gan and Correa are categorized as populists (and their
parallels with Hugo Chávez and Viktor Orbán are obvious), leaders of var-
ied ideological hues have engaged in executive aggrandizement. Between
2006 and 2008, President Abdoulaye Wade of Senegal curtailed protest
rights, tampered with the electoral calendar, changed legislative rules to
hamper a potential rival, and created a new upper house dominated by
his own appointees. Before his February 2014 ouster, Ukraine’s President
Nancy Bermeo 13
9
8
7
6
Elections
5
4
3
2
1
0
1991–1995 1996–2000 2001–2005 2006–2010
Year
Source: See source note in Figure 1. Additional election data are from Susan Hyde’s
NELDA project (http://hyde.research.yale.edu/nelda/). A detailed explanation of the axes
for this figure and how the data sources were used may be found at www.journalofdemoc-
racy.org/articles/supplemental-material.
Viktor Yanukovych meddled with the courts and police; maneuvered for
a parliamentary supermajority by inducing floor-crossings while banning
multiparty blocs; and engineered a return to the 1996 Constitution so that
he could hire and fire governors and cabinet members (including the pre-
mier) on his own. We find further examples of executive aggrandizement
across a range of countries as diverse as Sri Lanka and Mozambique.
Confronting Backsliding
A change in pace. The decline of coups means that de-democratization
today tends to be incremental rather than sudden. Dramatic breakdowns
will probably still occur, but troubled democracies are now more likely to
erode rather than to shatter—to decline piece by piece instead of falling to
one blow. Democratic erosion may be better than democratic cataclysm
because it is less likely to be violent, but incremental decline still presents
us with important challenges.
A first challenge is scholarly. Social science has focused mostly on
clear cases of democratic collapse—paying “scant attention” to the “in-
cremental”23 regime changes that color many countries’ histories. Re-
search on “hybrid” regimes has been a step forward, but we need to know
more about how the slide backward into hybridity takes place.24 Focusing
on democratic erosion will require more scholars to see that democracy is
“a collage” of institutions crafted and recrafted by different actors at dif-
ferent times.25 It is put together piece by piece, and can be taken apart the
same way. Politicians who engage in executive aggrandizement and stra-
tegic electoral manipulation already know this. Political scientists must
learn it too, or risk their own slide into irrelevance.
Incremental forms of backsliding create profound political challenges as
well. Domestically, alterations in electoral laws, district boundaries, elec-
toral commissions, and voter-registration procedures may seem too arcane
to be the stuff of mass mobilization. Court-packing and media restrictions
are probably easier to frame as dangers to democracy, but the jurists and
journalists who are likely to mobilize in opposition to these maneuvers can
easily be counterframed as “special interests” or tools of a discredited old
order. Civic organizations representing disadvantaged groups of other sorts
can be framed and silenced as tools of foreign forces. The fact that they often
are funded from abroad makes this especially likely and effective. Piecemeal
erosions of autonomy may thus provoke only fragmented resistance.
At a more general level, slow slides toward authoritarianism often lack
both the bright spark that ignites an effective call to action and the op-
position and movement leaders who can voice that clarion call. Executive
aggrandizement takes place precisely where a majority that supports it is
already taking root. Strategic electoral manipulation takes place where in-
cumbents already deem themselves capable of either securing or reinforcing
majority support. Since both forms of backsliding emerge precisely where
oppositions are already weakened by performance failures and internal di-
visions, mustering the power of numbers to reverse them is especially hard.
Even when opposition leaders succeed in mobilizing mass action against a
stolen election, their success is often heavily dependent on foreign allies.26
Nancy Bermeo 15
120 -10
-8
100 -6
-4
80
Authoritarianism
-2
60 0
2
40 4
20 6
8
0 10
1950–1959 1960–1969 1970–1979 1980–1989 1990–1999 2000–2009
Year
Source: See source note in Figure 1. A detailed explanation of the axes for this figure and
how the data sources were used may be found at www.journalofdemocracy.org/articles/
supplemental-material.
and the decline in the longevity and brutality of successor regimes sug-
gest that democracy’s prospects are still good. More systematic thinking
about how to cope with backsliding will make them even better.
NOTES
The author thanks Alexander Gard-Murray, Mario Rebelo, and Adam Brodie for their
research assistance.
4. Max Grömping and Ferran Martínez i Coma, “Electoral Integrity in Africa,” Elec-
toral Integrity Project, 2015, 22, https://sites.google.com/site/electoralintegrityproject4/
projects/expert-survey-2/regional-studies.
5. Diplomat quoted in Brian Klaas, “Bullets over Ballots: How Electoral Exclusion
Causes Coups d’Etat and Civil Wars” (PhD diss., Oxford University, 2015), 2.
7. Tom Harvey, “Junta Chief: Aristide May Return,” Sun-Sentinel (Florida), 6 Octo-
ber 1991.
8. Freedom House, “Individual Country Ratings and Status, FIW 1973–2015,” https://
freedomhouse.org/report-types/freedom-world.
11. Berna Turam, “Turkey Under the AKP: Are Rights and Liberties Safe?” Journal of
Democracy 23 (January 2012): 109–18.
12. Yesim Arat, “The Decline of Democracy in Turkey: Popular Support, Institutional
Decay,” unpubl. ms., Princeton University, March 2015.
Nancy Bermeo 19
17. Jason Eichorst and John Polga-Hecimovich, “The 2013 Ecuadorian General Elec-
tions,” Electoral Studies 34 (June 2014), 364.
18. Catherine M. Conaghan, “Surveil and Sanction: The Return of the State and Societal
Regulation in Ecuador,” European Review of Latin American and Caribbean Studies 98
(April 2015): 14–19.
19. Emily Beaulieu and Susan D. Hyde, “In the Shadow of Democracy Promotion:
Strategic Manipulation, International Observers, and Election Boycotts,” Comparative
Political Studies 42 (March 2009): 393.
20. Susan D. Hyde and Angela O’Mahony, “International Scrutiny and Pre-Electoral
Fiscal Manipulation in Developing Countries,” Journal of Politics 72 (July 2010): 690–704.
21. Judith G. Kelley rejects the tradeoff argument in Monitoring Democracy: When
International Election Observation Works, and Why It Often Fails (Princeton: Princeton
University Press, 2012), but agrees on the latter point along with Sarah Birch, Electoral Mal-
practice (New York: Oxford University Press, 2011), 163; and Alberto Simpser and Daniela
Donno, “Can International Election Monitoring Harm Governance?” Journal of Politics 74
(April 2012): 501–13.
23. Lust and Waldner, “Unwelcome Change,” 4; José Alemán and David Yang, “A
Duration Analysis of Democratic Transitions and Authoritarian Backslides,” Comparative
Political Studies 44 (September 2011): 1125.
24. For work on democratic erosion, see Guillermo O’Donnell, “Democracy’s Future:
Do Economists Know Best?” Journal of Democracy 6 (January 1995): 27; Scott Mainwar-
ing and Aníbal Pérez-Li~nán, “Cross-Currents in Latin America,” Journal of Democracy
26 (January 2015): 114–18; Larry Diamond, “Facing Up to the Democratic Recession,”
Journal of Democracy 26 (January 2015): 147.
25. Nancy Bermeo, “Interests, Inequality, and Illusion in the Choice for Fair Elec-
tions,” Comparative Political Studies 43 (August–September 2010): 1120.
27. Nicolay Marinov and Hein Goemans, “Coups and Democracy,” British Journal of
Political Science 44 (October 2014): 799–825.
28. Aníbal Pérez-Li~nán, Presidential Impeachment and the New Political Instability in
Latin America (New York: Cambridge University Press, 2007).
29. “Se pueden agregar textos a enmiendas, dice Rafael Correa,” El Universo (Guaya-
quil), 21 November 2015, www.eluniverso.com/noticias/2015/11/21/nota/5251623/se-
pueden-agregar-textos-dice-presidente.