Misunderstanding Democratic Backsliding

 

Thomas Carothers is the Harvey V. Fineberg Chair and director of

the Democracy, Conflict, and Governance Program at the Carnegie

Endowment for International Peace. His most recent book is Democracies

Divided: The Global Challenge of Political Polarization (2019,

coedited with Andrew O’Donohue). Brendan Hartnett is a James C.

Gaither Junior Fellow in Carnegie’s Democracy, Conflict, and Governance

Program.

Cases of democratic backsliding have been multiplying across the

globe in recent years, triggering a restless search among Western policymakers

and aid providers for causal explanations. One of the most common

ideas to have emerged in this policy community is that the blame

lies in democracies’ failure to deliver: When democracies do not provide

their citizens with adequate socioeconomic benefits, the thinking goes,

many of those citizens will lose faith in democracy and embrace antidemocratic

political figures who, once elected, will undercut democratic

norms and institutions. From this idea follows the policy conclusion that

to stop democratic backsliding concerned policymakers and aid providers

must help new or struggling democracies do better at delivering the

goods to their citizens, such as jobs, higher wages, food security, or access

to education. Writing in Foreign Affairs last year, Samantha Power,

administrator of the U.S. Agency for International Development, argued

that “to . . . swing the pendulum of history back toward democratic rule,

we must break down the wall that separates democratic advocacy from

economic development work and demonstrate that democracies can deliver

for their people.”1

It seems almost inarguable that governments of any political character

that provide strong socioeconomic results for their citizens will, on

average, be more stable and long lasting than those that do not. Yet it is

much less clear that a failure to deliver on the part of democracies is a

major cause of the wave of democratic erosion that has washed across

multiple regions in the past twenty years. There are some cases where

it seems to be at least partly true. In Tunisia, for example, accumulated

popular frustration with the socioeconomic performance of every government

that ruled the country after its transition to elected, civilian rule

in 2011 clearly factored in President Kais Saied’s decision to carry out

a self-coup in 2021, overturning the country’s democratic experiment.

Yet there are also cases where the “democracy-not-delivering” argument

seems much less true, if at all: Poland, for example, enjoyed an enviable

economic record in the years prior to the onset of its democratic

backsliding in 2015. Clearly, an empirically well-informed picture is

needed to probe the complexities lurking within this intuitively appealing

but sweeping idea.

To develop such a picture, we examined twelve countries that experienced

democratic backsliding during the past twenty years from a

starting point of either electoral democracy or liberal democracy: Bangladesh,

Brazil, El Salvador, Hungary, India, Mexico, Nicaragua, the

Philippines, Poland, Tunisia, Turkey, and the United States. This group,

drawn from the approximately 25 to 40 countries (depending on the specific

measures used) that are usually identified as having experienced

backsliding in the time period, includes some of the most internationally

prominent cases of backsliding—Brazil, Hungary, India, Turkey,

and the United States—while also being representationally diverse

along multiple dimensions, including population size, level of wealth,

and geographic location.2 The countries in this group are also diverse in

terms of their level of democracy prior to the onset of backsliding: Some

were relatively well-consolidated democracies, such as Poland and the

United States; others were much less so, including Bangladesh, Nicaragua,

and Turkey. The backsliding trajectories of these countries also

vary: Democratic erosion has been severe in some, such as Bangladesh

and Hungary, and mild in others, including the United States; likewise,

backsliding is ongoing in some, for example, El Salvador and India, but

appears to have been reversed (at least for now) in others, including

Brazil and Poland.

For each country, we analyze a range of socioeconomic and political

developments to answer three key questions: First, was the election

of the leader who ended up engaging in backsliding clearly linked to a

failure of democracy to deliver? Second, did that leader campaign on an

antidemocratic platform? And third, when the backsliding unfolded, did

citizens embrace it? Our findings raise substantial doubt about the utility

of the democracy-not-delivering thesis as a broad causal explanation of

the recent wave of democratic backsliding. That thesis has some explanatory

power in certain cases, but little in others. Even where the thesis

does apply, it involves considerable empirical complexity and requires

nuanced interpretation. Our findings reinforce on a global basis a crucial

insight drawn from the important study by Larry Bartels of democratic

26 Journal of Democracy

trends in Europe—that democracies “erode from the top” rather than at

the behest of their citizens.3 Policymakers and aid providers seeking to

stem democratic erosion globally should direct their energies and resources

accordingly and prioritize bolstering institutions and norms that

can constrain predatory powerholders.

Performance Realities

To examine whether poor socioeconomic performance precedes the

election of leaders who engage in backsliding, we looked at three core

economic indicators—inequality, poverty, and growth—in the five years

prior to each of the pivotal elections in question.4 The idea that anger

over rising inequality drives citizens to embrace illiberal politicians is

perhaps the most common form of the democracy-not-delivering thesis.

Indeed, some prominent Western observers concerned about rising inequality

around the world point to it as a major cause of democracy’s

global travails.5 Yet, although rising levels of domestic inequality undoubtedly

pose many problems for democracies (and other types of political

systems), the link to backsliding is not apparent.

In eight of the twelve cases under study, inequality was trending

downward in the five years before the elections that brought to power

leaders who ended up moving against democracy, averaging an almost

7 percent drop in the country’s Gini coefficient. In the four countries

where that was not the case—Brazil, India, Tunisia, and the United

States—inequality was roughly stable in the relevant period. A fallback

argument for the democracy-not-delivering thesis might be that what is

crucial is not the direction of inequality but the overall level. But here,

too, the hypothesis falls short: In most cases, inequality in the backsliding

countries was not higher than average when compared with their regional

peers. With respect to poverty rates, the picture is similar: In five

years before backsliding began, poverty levels decreased in nine of the

twelve countries—substantially in some cases, such as India and Poland.

What about economic growth? Has the election of antidemocratic

leaders typically been preceded by decreasing or low growth rates? In

most of the cases, growth was relatively stable for at least five years

prior to the onset of backsliding. And in some—Bangladesh, India, the

Philippines, Poland, and Turkey—growth was not just stable but relatively

high in regional and global terms. In the years leading up to their

backsliding, these countries were widely considered developmental

stars rather than laggards. Three of the cases did see a major economic

downturn prior to the pivotal election—Brazil, Hungary, and Tunisia.

In Brazil and Hungary, the downturn began several years before the

election; in Tunisia, the covid-19 pandemic led to a sharp recession the

year before Saied’s self-coup, adding a harsh edge to the long-mounting

anger among Tunisians about the country’s poor economic performance

in the decade after the democratic revolution. The United States, meanwhile,

experienced an economic slowdown in 2016, the year of its pivotal

election, but that was after relatively strong growth (at least compared

to democratic peers in Europe) from 2010 to 2015.

Although the democracy-not-delivering thesis usually focuses on

these core elements of socioeconomic performance, what about other

elements of governance performance? Two appear to have played an

important role in shaping the outcomes of the pivotal elections in some

of the twelve cases: corruption and crime. In Brazil, the vast Operaç~ao

Lava Jato (Operation Car Wash) corruption scandal involving the Brazilian

Workers’ Party and other mainstream Brazilian political actors

in the mid-2010s was a major driver of many citizens’ alienation from

the country’s traditional political elite and their turn to Jair Bolsonaro

in the 2018 election.6 Perceptions of significant levels of corruption

also contributed to the losses by incumbent parties in India in 2014 and

Mexico in 2018.7 With respect to crime, El Salvador’s extended period

of extremely high levels of violent crime played a central role in citizens

giving up on the two mainstream political parties and embracing Nayib

Bukele in the 2019 election (although it is worth noting that the country

had actually seen a significant decrease in homicides in the three

years prior to Bukele’s election).8 In the Philippines, Rodrigo Duterte

centered his successful presidential campaign around his claim that the

country’s crime situation was out of control and that he would fix it.

Yet even with this widening of the lens, the search for what led to the

election of these various leaders requires looking beyond performance

shortcomings that fit easily into the democracy-not-delivering framework.

In Nicaragua, Daniel Ortega managed to return to power in 2006

largely because the two main parties on the other side of the political

aisle split their vote, rendering Ortega’s 38 percent vote share enough to

win in the first round.9 In Poland, a complex mix of factors—including

shrewdly targeted campaign promises about new governmental benefits

and a somewhat vague desire for change on the part of many voters—put

the populist Law and Justice party over the line in 2015.10 In Turkey,

Recep Tayyip Erdoğan’s effective pushing of certain hot-button pro-Islamist

issues helped him and his Justice and Development Party (AKP)

to gain power in 2002 and remain in power in 2007.11 And in the United

States, it was not primarily economic backlash from “left behind”

working-class Americans facing financial hardship that drove Donald

Trump’s 2016 victory, but rather his skillful tapping of cultural animosities

and anxieties—about immigrants, racial minorities, and China, for

example—as well as other factors, such as the strong personal dislike of

Hillary Clinton among certain voter groups in key swing states.12

Moreover, even when governmental performance becomes a major

issue in a pivotal election, perceptions rather than the realities of performance

tend to matter most. As Christopher Achen and Larry Bartels

28 Journal of Democracy

have shown, voters’ perceptions of governmental performance may be

only loosely tied to the objective facts about governmental performance.

This sobering reality about elections and democracy weakens the assertion

that the electoral success of backsliding leaders can be traced to

specific performance shortcomings of their incumbent opponents.13 Underlining

this reality, charismatic electoral challengers in recent years

have often demonstrated notable skill in negatively shaping citizen perceptions

about the incumbent’s performance, a phenomenon perhaps fueled

by increasingly fast-moving and manipulable media environments.

In the Philippines, for example, the anticrime emphasis of Duterte’s

2016 campaign was a successful demagogic tactic to stir up public

fears: “framing the country’s minor drug problem as a major social one,

effectively manufacturing a crisis.”14 Similarly, the 2016 Trump campaign’s

emphasis on the dangers of immigration flows into the United

States amplified voters’ fears that their status in society was under threat

and resonated especially with non–college-educated whites, who overwhelmingly

backed Trump.15

In short, the outcomes of the pivotal elections in the twelve cases

under study were determined by a dense mix of social, cultural, political,

and economic factors. A sense of grievance on the part of voters

was often present, but their grievances ranged widely across different

domains and often had little to do with governments’ poor socioeconomic

performance. Moreover, contingent political factors, such as the

Nicaraguan opposition splitting the vote in 2006, and subjective voter

perceptions often counted significantly alongside more structural factors.

While democracy’s failure to deliver the socioeconomic goods is

relevant to some degree in some of the cases, according it a central explanatory

role for democratic backsliding represents an imposition of an

overly simple idea on a highly complex reality.

Embracing What?

The other main component of the democracy-not-delivering thesis

is the proposition that when citizens are disappointed with the socioeconomic

results of democratic governance, they knowingly embrace

antidemocratic alternatives—in effect voting against democracy out of

frustration with it. Here, too, the record of the twelve cases under review

does not provide strong support. Instead, it generally bolsters another

crucial finding of Larry Bartels concerning the emergence of political

illiberalism in Poland and Hungary: When Poles voted in Law and Justice

in 2015 and a majority of Hungarian voters chose Fidesz and Viktor

Orbán in 2010, they were voting for change but not consciously for illiberalism:

Only after the elections did it become clear that the winners

were bent on radically reshaping the political system, resulting in “illiberalism

by surprise.”

In the pivotal elections under review here that brought to power leaders

who ended up undercutting democracy, citizens were voting for

change, as citizens in democracies often do in elections. But for the most

part they were not voting for political

figures who were openly promising

antidemocratic action or threatening

to undo democracy. In Turkey, for

example, Erdoğan did not campaign

on a platform of political illiberalism

or antidemocracy in either 2002,

when he was first elected prime minister,

or in 2007, when he was reelected

for the first time. He and his

party instead positioned themselves

at the intersection of the center-right

and Islamic traditionalism, offering

a democratic alternative to the mainstream

parties that had long dominated

Turkish politics.17 In his first term

as prime minister, Erdoğan initiated some democratic reforms, such as

lifting restrictions on broadcasting and reporting in Kurdish, gaining

praise from Western governments for his prodemocratic actions.18

In Bangladesh in 2008, when the people voted into power the Awami

League, a political party that would go on to significantly curtail democracy

in the country, they were not making a purposely antidemocratic

choice. The Awami League’s record both in and out of power

was not unblemished—for example, the party had at times encouraged

violence among its supporters.19 Yet compared to its main rival, the Bangladesh

Nationalist Party (BNP), the Awami League was arguably the

more prodemocratic choice at the time given the BNP’s multiple prior

undemocratic actions, such as ruling by executive ordinance and rigging

some by-elections in the 1990s.20

Narendra Modi’s historic victory in India’s 2014 elections was driven

mainly by his record of economic success in his home state of Gujarat

and his reputation as a no-nonsense, business-friendly leader who

would drive through economic reforms and fight corruption.21 The many

disgruntled Tunisians who voted for Kais Saied in 2019 likewise were

not opting for an antidemocratic figure. Saied was a constitutional-law

scholar who had helped to draft the country’s post–Arab Spring constitution,

and he promised prodemocratic reforms, such as devolving power

to local governments and incorporating direct elements of democracy,

including recall elections. His opponent in the runoff, Nabil Karoui, appeared

as the greater risk to democracy, having enjoyed close ties with

the country’s former dictator, Zine al-Abidine Ben Ali (1987–2011).22

Some of the leaders elected in the case-study countries who later

In many backsliding

contexts, it would be a

serious mistake to say

that large shares of voters

are going along happily

and knowingly with

democratic erosion—many

are being bludgeoned,

bribed, or manipulated via

illegitimate means into

acquiescing.

30 Journal of Democracy

drove democratic erosion had acted undemocratically in their political

pasts. Yet their campaigns were not rooted in that fact, and their

victories do not appear to have been either. Mexico’s Andrés Manuel

López Obrador, for example, refused to concede his loss in the 2012

presidential election. But when he ran again in 2018, he did not base

his campaign on attacking the institutions or norms of Mexican democracy.

It was oriented instead around his promise to make Mexico a more

fully inclusive country by achieving fundamental pro-poor economic

changes.23 In Poland, Law and Justice’s commitment to democracy was

undoubtably blemished going into the 2015 elections—party leader

Jaros³aw Kaczy´nski had, after all, praised Orbán’s model of illiberal

governance and expressed a desire to implement it in Poland. But the

party kept these antidemocratic positions out of view in the 2015 campaign,

softening its image to appeal to voters who were simply tired of

the “boring” incumbent party.24

Prior to running for president of El Salvador in 2019, Nayib Bukele

had engaged in some illiberal actions in his capacity as mayor of San

Salvador, for example, ordering cyberattacks on newspapers critical of

him. Yet in his 2019 presidential campaign, he positioned himself as

the choice to save Salvadoran democracy—as a reformer in between

the country’s two longstanding parties on the right and left, who would

return power to the people through anticrime and anticorruption measures.

25 Daniel Ortega entered Nicaragua’s 2006 election with a clear

antidemocratic past, having presided over the authoritarian Sandinista

regime from 1979 to 1990. Yet in the intervening years he had participated

in the country’s messy multipartism. For his 2006 bid, Ortega

did not position himself against the country’s political establishment

but rather as a candidate of “love, reconciliation, and forgiveness” who

would do more for the country’s poor than his two center-right rivals.26

In only three of our twelve case studies could it be said that the winner

of the pivotal election in question ran on a campaign that specifically

promised systemic antidemocratic actions—though not with the stated

intention of undermining the entire democratic system: Duterte in the

Philippines and Trump in the United States, both in 2016, and Brazil’s

Bolsonaro in 2018. Before being elected president, Duterte had ruled

Davao City with an iron fist, ordering extrajudicial killings of suspected

criminals, drug users, and communist rebels. Instead of downplaying

this on the campaign trail, Duterte “shrugged off accusations of humanrights

abuses, promising to implement his Davao model nationwide.”27

He also threatened to disband, quiet, or ignore Congress, the courts,

and human-rights activists if they got in his way, and he praised former

dictator Ferdinand Marcos.

Donald Trump, at 2016 campaign rallies, led cheers calling for his

main opponent, Hillary Clinton, to be “locked up” and said that he would

respect the results of the election only if he won. Similarly, Bolsonaro,

who dismissed accusations that he was a threat to Brazilian democracy

as “fake news” despite being a well-known apologist for the country’s

1964–85 military dictatorship, said during the 2018 campaign that he

would wipe his political opponents “off the map.”28 Even in these cases,

however, antidemocratic threats or promises were part of much larger

platforms that candidates stood for or pledged to fulfill.

In short, the idea that voters in these twelve countries were purposefully

choosing an antidemocratic path in electing leaders who ended up

working against democracy once in power does not hold up to empirical

scrutiny. In most of the cases, voters were embracing the promise of

significant, even disruptive change, whether in economic, social, or political

domains. Some of their preferred candidates represented a risky

path, democratically speaking, given past antidemocratic incidents or

records. But except in a few cases, voters were not throwing their support

behind political figures who were advocating an end to democracy

or openly questioning the value of democracy per se. Most of the voters

in these pivotal elections were not turning against democracy; in fact,

many appeared to be trying to save democracy by giving it a pointed but

reform-oriented reset.

Once Backsliding Is Underway

Can it be argued that, even if voters in these pivotal elections were

not knowingly voting for the dismantling of democracy, once those

leaders gained power and began to undercut democratic norms and institutions,

their supporters embraced these moves? In other words, if

citizens in backsliding countries are not choosing democratic erosion

before the fact, do they become contented enablers once such deterioration

is underway?

Some of the backsliding leaders in the case-study countries have enjoyed

notably high levels of popularity while in office. López Obrador,

Bukele, Duterte, and Modi stand out in this regard. López Obrador

enjoyed approval ratings above 60 percent in the final months of his

single, six-year presidential term, and his chosen successor, Claudia

Sheinbaum, easily won Mexico’s June 2024 election. Bukele’s approval

ratings among Salvadoran voters have stayed sky-high, at times topping

90 percent since he became president in 2019, and he was reelected in

2024 with almost 85 percent of the vote. Duterte’s popularity remained

high throughout his presidency as well, and his party swept the midterm

elections in 2019, giving it control of both legislative houses. Modi and

his party expanded their support during their first ten years in power.

These leaders have sustained their popularity at least in part by giving

their supporters what they want, whether it is hard-nosed anticrime policies

in El Salvador and the Philippines, pro-poor economic policies in

Mexico, or pro-Hindu social policies and vigorous nationalism in India.

32 Journal of Democracy

In line with Matthew Graham and Milan Svolik’s findings regarding the

United States—that in a highly polarized context voters are unlikely to

turn against undemocratic politicians when they generally agree with

their policies—many people in these four countries appear to be willing

to overlook or tolerate a backsliding leader’s antidemocratic moves.29

For example, Filipino voters were “willing to put up with extrajudicial

killings, political repression, and the gutting of liberal institutions because

they [saw] Duterte as a strong leader. They question[ed] his methods

but not their effectiveness.”30 In some cases, it is likely that supporters

of backsliders do not just overlook the democracy factor but embrace

the illiberalism of their leader—whether it takes the form of systematic

discrimination against certain groups or attacks on the courts—because

they view it as justified, sometimes as payback for perceived wrongs of

the past.

It is important to be cautious, however, about attributing the popularity

of some backsliding leaders to genuine voter satisfaction based on

promises delivered. Most of these leaders manipulate public opinion in

various illegitimate ways. The leaders of Bangladesh, Nicaragua, and

Turkey, for example, have cracked down harshly on those who dissent

or criticize the ruling party. The leaders of almost all twelve backsliders

employ substantial control over the media to shape the information

space in their favor. In the 2022 electoral campaign in Hungary, for

example, the main opposition candidate for prime minister was given

just a few minutes on the state-dominated television to make his case

to voters, in contrast to months of wall-to-wall adulatory coverage of

Prime Minister Orbán and his party.31 Likewise, before the 2023 elections

in Turkey, the state channel TRT gave Erdoğan’s campaign 32

hours of coverage, compared to just 32 minutes of coverage for his main

challenger.32

Backsliding leaders also frequently use state resources to fund their

own parties’ political campaigns and to coopt important businesses and

other key groups, while harassing opponents with regulatory retaliation,

tax inspections, and other punitive measures. During Ortega’s first

decade in power, he simultaneously maintained support from the poor

through patronage and backing from the Nicaraguan business community

in exchange for lucrative contracts.33 In many backsliding contexts,

it would be a serious mistake to say that large shares of voters are going

along happily and knowingly with democratic erosion—many are

being bludgeoned, bribed, or manipulated via illegitimate means into

acquiescing.

In the United States, Brazil, and Poland, backsliding leaders lost

enough popularity, and faced strong enough institutions of countervailing

power, that they ended up losing power in failed reelection attempts.

Their loss of popularity had a mix of causes. In the case of Trump, for

example, the first year of the covid pandemic hit his presidency hard,

both its negative economic effects and the weaknesses it exposed in his

governance style. Although the antidemocratic actions of Trump, Bolsonaro,

and Law and Justice appear to have played some role in weakening

their popular support, it is not a straightforward story.

In all three cases, it appears that the leaders’ antidemocratic actions

did not much faze core supporters, in line with Graham and Svolik’s

finding mentioned above. Survey research ahead of the 2020 U.S. presidential

election showed that even before ballots were cast, there was a

strong desire among Trump supporters for the incumbent to fight against

an electoral defeat. The leading justification provided by those who supported

Trump’s effort to undermine the results was not concern about

the integrity of the election, but simply that they supported Trump and

disliked Biden.34 These supporters either did not perceive their antidemocratic

actions as that harmful or were willing to accept them for other

policy gains.

At the same time, polling data indicate that the antidemocratic actions

of Trump, Bolsonaro, and Law and Justice did hurt them on the

periphery of their supporter circles—among voters who initially gave

them a try but were not necessarily firmly committed. Moreover, their

democratic transgressions helped to mobilize their opponents and get

them to the polls in even greater numbers. Law and Justice’s assaults

on Polish democracy, for example, cost it support from among previous

backers and sparked large-scale antigovernment protests in the

run-up to Poland’s 2023 elections, which produced an alternation of

power.35

Prioritizing Constraint

The idea that a failure of democracies to deliver is a major cause

of the recent wave of democratic backsliding is intuitively appealing.

Many democracies are struggling to provide secure, prosperous lives

for their citizens, and many are struggling to maintain solid democratic

norms and institutions. Yet the evidence for a causal link between these

two phenomena in twelve prominent contemporary cases of democratic

backsliding is mixed at best. In most of the cases, inequality and poverty

were heading downward in the five years before the election that

brought a backsliding leader to power, while growth was largely stable

or even high. Sharp economic downturns did seriously damage the popularity

of incumbent democratic governments in three of the cases (Brazil,

Hungary, and Tunisia), and citizen anger over high levels of crime

and corruption was a key factor in bringing an antidemocratic leader to

power in at least Brazil and El Salvador.

Overall, however, it was a wide range of factors, not just economic

and governance grievances, that led voters in these countries to elect

leaders who ended up eroding democracy. These factors include broader

34 Journal of Democracy

sociocultural anxieties, the impressive electoral and narrative skills of

some of the leaders in question, specific features of electoral laws, the

new fluidity and corruptibility of media environments in many democracies,

and the frequent appeal to voters

almost everywhere of change for

change’s sake.

In addition, most of the winners of

these pivotal elections did not campaign

on promises of dismantling democracy.

Some had illiberalism in their

political pasts, including Daniel Ortega

and Narendra Modi. But even those

for whom that was true focused their

electoral strategies on other issues,

such pro-poor economic policies in the

case of Ortega or economic reform and

dynamism in the case of Modi. Some

appeared to be the more democratic choice compared to their main opponents,

as in Bangladesh in 2008 and Tunisia in 2019. The illiberalism

that emerged in the twelve case-study countries was thus more illiberalism

after the fact than illiberalism by promise. Most voters in these

countries could not be said to have consciously turned away from democracy

in making the choices they did at the ballot box.

What this suggests is that pointing to poor socioeconomic performance

by democratic governments and citizen anger about that performance

as the primary drivers of democratic backsliding is a mistake.

The onus for backsliding belongs on those leaders who gain power for a

wide range of reasons, including in many cases by promising to renovate

democracy, but then once in power relentlessly amass unconstrained

power by overriding countervailing institutions and undercutting basic

democratic norms and procedures. Backsliding in these cases is thus less

about a failure of democracy to deliver than about a failure of democracy

to constrain—that is, to curb the predatory political ambitions and

methods of certain elected leaders.

This conclusion does not absolve the outgoing democratic governments

in these and other cases from their many shortcomings in terms

of how they have governed and what they have delivered to citizens,

which have unquestionably caused or aggravated countless hardships

and wrongs. Indeed, such shortcomings have contributed to the resurgence

in recent years of the old debate—dating back to the 1960s and

before—about whether autocracies are better able than democracies to

deliver economic benefits. But our findings point the primary finger of

blame for democratic backsliding at the politicians and political parties

that have acted antidemocratically and the weakness of democracy’s institutional

guardrails in these countries.

analytic conclusion points to an overarching policy conclusion.

The community of public and private organizations working internationally

to slow or reverse the tide of democratic backsliding should

emphasize bolstering the independence and strength of those institutions

that serve as guardrails against antidemocratic encroachments by political

figures determined to steamroll countervailing institutions and gain

unchecked power. On the side of public institutions, this may include

courts, anticorruption bodies, electoral-management bodies, and those

parts of local government that have preserved some autonomy from national

political control. It may also include work with national parliaments,

if that can be done in a way that augments the role of parliament

as a check on executive power. On the nongovernmental side, this will

often mean support for independent media and independent civic groups

that seek to hold power accountable whether by protecting political and

civil rights, increasing government accountability, or countering political

misinformation and disinformation.

Focusing on fortifying constraint is not the same as a generalized

strategy of political institutional support—it needs to be carefully tailored

to reinforce democratic guardrails and avoid soft-edged governance

programming that may inadvertently help to strengthen an overweening

political power structure. Such a strategy will ideally include a

mix of well-coordinated elements of aid programming, diplomatic interventions,

and economic carrots and sticks. None of the various constituent

elements of such a strategy are new for democracy supporters, but

they should be joined together more purposely into integrated efforts to

buttress prodemocratic constraints on power.

Prioritizing such an approach does not mean international supporters

of democracy should forsake ongoing or potential new efforts to help

new or struggling democratic governments deliver better socioeconomic

results for their citizens. Those efforts are worth doing in and of themselves

to contribute to a better life for people across the democratic

world. And in some cases, they may over time help to alleviate sociopolitical

pressures that could encourage citizens to take risks with disruptive

and potentially undemocratic alternatives. But such efforts should

be viewed as a complement to, not a substitute for, a primary strategy

of strengthening domestic sources and structures of prodemocratic constraint

and disincentivizing politically predatory leaders from deepening

their antidemocratic quests.