Populism and Identity Politics

Over one-third of humanity lives under populist regimes—and many of those regimes are turning increasingly authoritarian. It is a worldwide challenge to liberal democracy. The conventional wisdom is that bad economics is to blame: the losers from globalization are angry and voting populists into office is their revenge. The policy implication is a kind of technocratic fantasy: fix the economy and populism will fade away. That view has weak empirical foundations, since many emerging countries that are clear winners from globalization have recently elected populists. In this essay I argue that we cannot understand the surge in populism without understanding the rise of identity politics around the world. Identity is the intermediate stopover in the two-way feedback between economics and politics. A focus on identity politics has important practical implications. One of them is that, to succeed in the fight against populism, democratic politicians have to learn to practice identity politics, but of the right kind. The challenge is to build national identities based not on nativism or xenophobia, but on liberal democratic values.

poster children for liberal-democratic transitions. In Southern Europe, democracy and economic growth were flourishing again. Autarky, nationalism and military coups in Greece, Spain and Portugal seemed the stuff of decades past. Turkey was a working democracy and would soon become, many hoped, a member of the European Union.
In South Africa, the hideous apartheid regime was crumbling. A negotiated political settlement would soon allow Nelson Mandela to move from prison to high office. In the New World the news was just as inspiring. Argentina, Brazil, Chile and Uruguay had recently returned to democracy. Soon, countries from Mexico to Peru would take steps toward this greater freedom.
How distant that moment now seems. Venezuela has slid back into dictatorship and Nicaragua is almost there. The shadow of far-right authoritarianism has reappeared in Italy, Spain, and Greece, while all over Southern Europe nationalists and demagogues call the European democratic enterprise into question. Something far more dramatic is underway in Central and Eastern Europe. Poland and Hungary are quasi-autocracies that trample on civil rights and pack once-autonomous institutions with government cronies. The same is true of Turkey.
Even long-established democracies are under stress. In the United States, Donald Trump has repeatedly clashed with Congress and the courts. Narendra Modi, India's prime minister, openly disdains India's secular constitution. The Economist recently criticized him 'for his apparent determination to transform India from a tolerant, multi-religious place into a chauvinist Hindu state'.
As Steven Levitsky and Daniel Ziblatt wrote in their 2018 book How Democracies Die [2], we tend to think the end comes with rolling tanks and machine gun rattle. But it need not be that dramatic: 'Democracies also die at the hands not of generals but of elected leaders (…) who subvert the very process that brought them to power. Some of these leaders dismantle democracy quickly (…). More often, though, democracies erode slowly, in barely visible steps.'

Defining Populism
So populism is a threat to liberal democracy. But what is populism? Economists, unsurprisingly, have defined the phenomenon in exclusively economic terms. Dornbusch and Edwards (1991) provided the now-classic definition: populism is ' an approach to economics that emphasizes growth and income redistribution and deemphasizes the risks of inflation and deficit finance, external constraints, and the reaction of economic agents to aggressive nonmarket policies' [3].
But this definition seems ill-fitting when we consider most of the populist regimes that we see in place today. Instead, there is something we can call political populism, distinct from economic populism. Muller [4] and Mudde and Rovira [5] provide a useful definition: populism is a way of doing politics in which 'the people' are pitted in conflict against others-various ' elites', local minorities, immigrants, foreigners. Muller stresses populists' moralistic interpretation of politics: those on the side of the people are moral; the rest are immoral, doing work the of a corrupt elite. This means populism is not an ideology. It does not pretend to offer ' complex [or] comprehensive answers to the political questions that modern societies generate', and so both right-wing and left-wing types of populism are possible.
Instead, populism rests on a triad: denial of complexity, distrust of pluralism and anti-elitism. Most of us believe that social choices are complex, and that the existence of plural views about what to do is a natural consequence of this complexity. Populists deny this. As Ralf Dahrendorf once put it, 'Populism is simple; democracy is complex' [6].
Inevitably, then, populists do not believe in pluralism. Since there is only one right view-that of the people-so there is only one view deserving of political legitimacy. It follows that the complex mechanisms of liberal democracy, with all that delegation and representation, are unnecessary. Instead, populist leaders make the claim that they alone can represent the people, unchecked by other institutions or individuals.
Populism is also-crucially-a rebellion against various elites, including, of course, traditional political elites. 2 In Politics as a Vocation, his famous lecture of a century ago, Max Weber warned that a key risk for modern democracy was that a political class would arise, disconnected from voters and the common people. Well, that political class did arise. Now people are revolting against it.
The standard refrain is that citizens vote for that politician with whom they would like to have a beer. But rather than sharing a drink with the average voter, leading politicians spend too much of their time with others like themselvesbankers, business people, top civil servants, high-flying academics. To ascertain which politicians can be successful today, Yascha Mounk calls for a 'inverted likeability test': voters do not prefer the candidate they would rather have a beer with; they prefer the candidate who would rather have a beer with them [7]. Too many conventional politicians fail this test.

Politics Trumps Economics
What is behind the rise of populism? Why this new and powerful threat to the liberal-democracy, a political system that just 30 years ago towered triumphantly above all else?
The standard answer takes the economic perspective and focuses on the pocketbook. In countries like the United States and the United Kingdom, the distribution of income has worsened and the top 1% is reaping the lion's share. In places left behind by technological change and globalization, people have lost their jobs and their patience. The 2008 global financial crisis not only caused much pain; it also reinforced the conviction that Wall Street is the enemy of Main Street. No wonder politics has become confrontational and populists have the upper hand.
If this narrative is right, the policy conclusion is simple: tax the rich, redistribute more income, and throw out the rascals who did the bankers' bidding. Populism will eventually fade away. This is an appealing story, but is it right? Should we base policy on it?
There is no shortage of empirical papers that have answered yes, purporting to show that, at least in North America and Western Europe, the forces behind populism are mostly economic. In their influential paper on 'China shock', Autor, Dorn, and Hanson contend that local US labor markets with a bigger trade exposure to China suffered large job losses, decreases in labor market participation and persistent unemployment [8].
Autor, Dorn, Hanson, and Majlesi found evidence that congressional districts with larger increases in import penetration became more politically polarized [9]. 3 In a companion paper, the same authors [10] related the change in the county-level Republican vote share to the growth in local labor markets' exposure to the China shock. They found rising import competition made Republican vote share gains more likely.
Using a similar methodology, but applied to Western European data, Colantone and Starnig argued that voters in regions with higher exposure to China shock were more likely to vote for a far-right candidate [37]. In the UK, argued the same authors in a later piece, more local trade exposure meant an increase in support for Leave [38]. There is also some evidence that increases in unemployment help explain rising votes for populist parties across Europe [11].
But that is not the end of the story. There is also an abundant supply of papers that single out culture and values, not economics, as the key explanatory variables for populism. In the aftermath of the 2016 US presidential election, American statistics guru Nate Silver noted that Hilary Clinton improved on Obama's 2012 performance in the overwhelming majority of the best educated counties, but lost ground in the least educated counties. Diana Mutz similarly concluded that 'Status threat, not economic hardship, explains the 2016 presidential vote' [12]. The title of another influential paper [13] points in the same direction: 'Vote Switching in the 2016 Election: How Racial and Immigration Attitudes, Not Economics, Explain Shifts in White Voting. ' In the UK, research by Becker, Fetzer and Novy, examining 382 local authorities, concluded that while education and demography are good predictors of who voted to leave the European Union, exposure to trade and the extent of budget cuts are not [14]. And evidence in favor of the ' cultural backlash' thesis is not limited to the US and the UK, argue Norris and Inglehart, who studied the performance of populist parties in 31 European countries [15]. They conclude: 'Overall, we find the most consistent evidence supporting the cultural backlash thesis. ' It is unlikely that the debate between the ' cultural backlash' and the ' economic insecurity' hypotheses will ever be fully adjudicated, and not just because of standard econometric difficulties related to identification. Disentangling direct and indirect effects is particularly challenging. It could well be, for instance, that economic shocks lead to changes in cultural values, which in turn increase support for populist parties. Conversely, changes in social or cultural normsincreased tolerance of labor market discrimination against immigrants or ethnic minorities is an example-could have economic consequences, which in turn could affect political outcomes.
Moreover, participants in this debate are not always very precise about what these competing hypotheses are supposed to explain. As Margalit has compellingly argued, the debate often confuses outcome and explanatory significance [16]. It could be that economic shocks shifted 4% of the UK vote toward Brexit. That is outcome significance, in that it focuses on the determinants of those few marginal votes that triggered the outcome. But is that what we need to explain? Perhaps not. Margalit is adamant about this: 'The overall phenomenon to be explained is why 52% of the electorate voted to leave the European Union.' So far the bulk of the formal evidence concerns the possible sources of populism in the prosperous countries of North America and Western Europe. Formal empirical research into the causes of populism in emerging nations is much scarcer. But the informal evidence available suggests a story that is rather different from that often told about the rich nations.
In the rich-country narrative, economic stagnation and the frustrations of the 'left-behind' take centre stage. In the emerging world, by contrast, right-wing populism is thriving in countries with strong economic performanceswhich is just the opposite of what the 'economic insecurity' hypothesis would predict. India, the Philippines and Turkey have grown at rates between 6.5 and 7% since 2010. Poland barely suffered the effects of the European financial crisis and has been Europe's growth champion, with an average per capita growth rate of more than 4% since 1992. The story in Hungary is similar: per capita income has been converging quickly with Western European levels. Or consider the neighboring Czech Republic, where unemployment is the lowest in the EU and the economy grew 3.5% in the five-year period ending in 2019. The country has few immigrants and no refugee crisis to speak of. Nonetheless, populist parties attracted four of every ten voters in the most recent election-a tenfold increase in two decades. So in these countries populism seems to have been the offspring of economic gain, not pain! India, Turkey, Poland or Hungary are winners from globalization, yet they are going populist too! There is one last prickly fact to consider: if surging populism reflected a demand for redistribution, we would expect the surge to be on the left, not the right. The left has seen some success, with left-wing populist parties governing today in Argentina and Mexico, while Podemos has joined the cabinet in Spain. But in much of the world the story of leftwing populists is one of electoral failure, not success-including Jeremy Corbyn's flop in the recent British elections. The spectacular success is that of right-wing populists, who often promise and enact policies that are likely to worsen the distribution of income, yet middle class and working class voter cheer them on.
Dani Rodrik has proposed an explanation to this conundrum [17]. He suggests that political consequences depend on 'the forms in which globalization shocks make themselves felt in society'. So in Latin America, where globalization has involved massive capital flow volatility and frequent financial and debt crises, the populist backlash has been on the left. In North America and Europe, by contrast, where trade and migration have provided the central cleavage, populism is of the right-wing variety.
The hypotheses is intriguing, but it raises as many questions as it answers. Given the depth of the 2007-2009 financial crisis in North America and Europe, why did it not generate Latin American-style left-wing populism? 4 Why have countries like the Philippines and Turkey, which look positively Latin American in their macro and financial instability, become poster children for right-wing populism? There is also the fact that Brazil, a country long affected by financial instability, is now governed by a right-wing populist. His economic agenda involves cutting back pension benefits, privatizing state-owned enterprises and making Brazil more economically globalized.
None of this means to deny the intensity of economic grievances, whether in the north of England, the American rust belt, the shanties of Manila or the favelas of Brazil. The point is different: economics matters, of course, but politics and culture dictate how people process the experience of economic success and failure. The main conceptual shortcoming of the economic insecurity hypothesis is that it assumes a simple (and monotonic) mapping between economic outcomes and political behavior. Such a mapping does not exist. Pre-existing social and value structures can cause economic ups and downs to have very different political consequences-for instance, if an adverse economic shock causes a rise in unemployment, which prompts a turn toward populism in a divided society, but not in a cohesive one. A key role of politics is to manage grievances, economic and otherwise. The turn toward populism and authoritarianism suggests a failure of democratic politics to handle those grievances effectively. There is a one-word reason for that: identity.

The Identity Roots of Populism
Katherine Cramer is a political scientist who visited dozens of small towns in Wisconsin and spoke with hundreds of people in an effort to understand why the state was so politically polarized. She wanted to know why voters in traditionally left-leaning Wisconsin were supporting Scott Walker, a Republican governor with populistic tendencies. What she found surprised her: Perhaps issues are secondary to identities; perhaps when people vote for a candidate their overarching calculation is not how closely this person's stances match my own, but instead, is this person like me? Does this person understand people like me? The answers to those questions include a consideration of issue stances, but issue stances are not necessarily the main ingredient.
Scott Walker had built political capital by picking a fight with the state's public sector unions. Cramer found that most rural residents supported Walker not because of concerns over the budget deficit or the quality of public services, but because they viewed public sector workers as urbanites, who could not possibly have the interests of rural residents in mind. A man milking cows blurted out: 'I'm glad Walker did what he did. It is about time someone takes something away from those bastards'. The bastards in question were public employees. After many conversations like that one, Cramer decided to call her book 'The Politics of Resentment' [39].
Wisconsin is not alone in the central role identity plays in politics. Look around the world today and you see identity politics everywhere. What Brexiteers, Catalan separatists, Russian nationalists and Islamic fundamentalists all have in common is that their politics are all about identity. India's Modi and Israel's Netanyahu have both found political profit in pitting one local identity against another. And what is the massive backlash against immigration if not the assertion of one identity over another? The more globalized the economy becomes, the more politics around the world is driven by the very local identities.
Of course many political parties in the West have long understood-and practised-identity politics. To be electorally successful, parties have differentiate their 'product' from that of their competitors. And in the past two or three decades, that differentiation has come not so much from economics but from other issues that are natural markers for identity: in the United States, Democrats became the party of racial equality, abortion rights, gay marriage, and liberal immigration policies, while Republicans the party of nationalistic pride, right-to-life, traditional values and tight immigration controls.
The reason why identity matters for politics is that identities are shared. In a recent book, Francis Fukuyama argues that 'individuals often want not recognition of their individuality, but recognition of their sameness to other people' [18]. We also want that identity recognized and respected. Fukuyama reminds us that philosophers from Aristotle to Hegel placed the desire to be treated with respect at the centre of human motivation: 'identity politics is everywhere a struggle for the recognition of dignity'.
Populism has a great deal to do with this. To the definitions given above one can add: populism is a style of politics that manipulates and exacerbates identity cleavages for political gain. For the late Venezuelan autocrat Hugo Chávez, anyone who opposed him was an enemy of the people and an agent of the corrupt elite. Change corrupt elite for menacing foreigners, and that is also the rhetoric of Donald Trump. So populism is a kind of identity politics. It is always us against them.
Identity concerns also explain the anti-elitist element in populism. Elites have also been arrogant, often dismissive of the national identities that much of the electorate holds dear. Hillary Clinton's description of Trump voters as a bunch of ' deplorables' did not help her campaign. In Latin America, left wing intellectuals routinely depict middle class voters who lean right as consumerist social climbers who have sacrificed class solidarity in the altar of money-grubbing individualism. Recall Fukuyama's definition of identity politics as a demand for dignity. Well, elites have not treated some citizens with respect and dignity.
Identity politics is not an easy subject for economists. Until recently, economic theory did not make room for identity. Humans were supposed to have preferences but liking this and disliking that did not amount to a coherent whole that we could call an identity. Akerlof and Cranton set out to change this [19][20][21][22][23]. They argued that, in a wide range of contexts, preferences are structured by individuals' choices of a social identity, and studied the economic implications of those preferences.
The identity approach helps us understand why people are willing to pay steep costs, pecuniary or otherwise, to buttress their identities. For instance, in American high schools [20], students who identify as nerds will study hard, while students who identify as jocks or burnouts will fail to study and underperform, even if that is costly, because such behavior helps reinforce their identities and their self-esteem.
Acemoglu, Egorov and Sonin [24] and Di Tella and Rotemberg [25] argue that populist politicians adopt extreme and ultimately unsustainable policies as a way of signaling to voters that they (the politicians) are not in the pocket of powerful elites. So self-defeating economic behavior is quite understandable once identity is accounted for. And populism certainly involves plenty of self-defeating economic policy choices.
Identity also creates feedback loops between individuals' beliefs and actions. For instance, as the share of people that identifies with a certain group goes up, so does the social pressure to identify with that same group and follow its codes of conduct. Alternatively, as in Gennaioli and Tabellini [26], identifying with a group can cause people to slant their beliefs toward the group's prevailing opinion. Or as in Shayo [27], people may choose the group they identify with and, once there, choose their actions in order to minimize the distance between their own attributes and the group's average attributes.
In this context of strong complementarities, small economic or cultural shocks can cause sizeable changes in patterns of identification and hence in political preferences. This helps explain, for instance, sudden and sharp shifts in support away from traditional political parties and toward populist movements. Besley and Persson [28] study these issues in the context of a fully dynamic model, in which there is two-way feedback between identities and policies. When they allow for endogenous institutional changes, like the entry of new populist or nationalist parties, outcomes exhibit path dependence, with temporary shocks having persistent effects on the share of support for populist or nationalist politicians. 5

Can Liberal Democrats Practice Identity Politics?
If identity is key to populism, and populism is central to contemporary politics, what can democratic politicians do in response.
To begin with, they can focus on some important issues they have long neglected. Take, for example, the plight of cities where de-industrialization has destroyed jobs. Previously, the standard advice to residents of Akron, Ohio, or Gary, Indiana, was to move to California, where high-paying jobs are plentiful. Today, we understand that can be unsound advice, and not only for the obvious economic reason that the most educated and enterprising move away, leaving behind communities that struggle to sustain businesses and make ends meet. 6 The combination of job losses and outward migration also weakens the local community, challenges shared identities, and causes the kind of malaise on which populists and demagogues feed. That is why place-based policies must be an essential component of the toolkit of a democratic policymaker. 7 Identities also matter for the way policies are perceived. Take Emmanuel Macron's flop over gas levies in 2018. He did what any reasonable policymaker would have done: concerned with both global warming and local pollution, he proposed taxing diesel more. Before he knew it, the country was up in arms. That was the beginning of the movement of the gilets jaunes, who complained that the president and his friends live in Paris and ride the subsidized metro, while they live in the countryside, drive trucks and pay the taxes that finance Parisians' privileges. They felt that Macron simply did not understand them and their way of life. Warnings from the Elysée Palace about planetary responsibility exacerbated the feeling of disconnect. A leader of the gilets jaunes griped that the president was fretting about the end of the world while they worried about getting to the end of the month.
Convincing middle-class French voters that higher fuel prices were actually good for them was always sure to be an uphill battle. But Macron's background and style made it even tougher. Maybe it was the inevitable consequence of the president's background as an investment banker, his imperial style, or of the abolition of the wealth tax as the initial priority of his administration. Macron could have promised to return the fuel tax revenue to middle class families and businesses, but he did not. What might have been a narrow taxation row became an unwinnable clash of identities. In the end Macron had to back down. It was his biggest loss.
What else can liberal democrats do? They can also abandon the vain hope that simply by tweaking economic policies populism will go away. Better and bolder policies to improve income distribution and enhance social mobility are the beginning of the road, not the end of the road. Populism is a political problem; it requires political solutions.
That is why the way forward cannot be merely technocratic. The title of a recent article by Sheri Berman [31] is spot on: 'Populism Is a Problem. Elitist Technocrats Aren't the Solution.' Populist politicians are capitalizing on the public's dislike and distrust of technocrats. In the midst of the Brexit debate, minister Michael Gove exclaimed 'people in this country have had enough of experts'. Donald Trump has said worse. Having experts and technocrats lead the countercharge against populism is exactly what populists hope for.
Katherine Cramer discovered in Wisconsin that voters look for a candidate who inspires trust, one who would make the choices they would have made if only they had had the time, knowledge and inclination to study and understand the issues. They hope for a candidate who is 'like them' when it comes to values and preferences. Macron comes up short here. So do most other liberal politicians. Liberal democracy has a personnel problem. Democratic parties need a revamped Human Resources Department with a new mandate: hire better and diversify your recruitment sources.
But a better HR policy alone will not do the trick. Something else must change: democrats must learn to practice identity politics-but identity politics of the right kind. Human beings cannot and will not do without narrow identities, which are the most firmly rooted. But there also exist broadly shared identities, which can serve as the basis for the sense of shared destiny that is at the core of good politics. 8 As Michael Ignatieff [32] has observed, 'national identity is a continual contest about who belongs to the national we'. Democrats must provide an expansive definition of that national we.
According to Paul Collier [33], the United Kingdom built that shared identity in the battlefields of the world wars, 'an immense common endeavor in which leaders had crafted narratives of belonging and mutual obligation'. The legacy was to turn the nation 'into a gigantic community, a society with a strong sense of shared identity, obligation and reciprocity'. But, laments Collier, in recent decades much of that was lost. Highly educated professionals in London came to feel they had more in common with their peers in Amsterdam or Paris than with working-class Britons in Sheffield (Collier's home town), who in turn sought refuge in anti-EU English nationalism.
In the United States the process has been similar, but perhaps even more radical, with prosperous residents of the coasts looking down on the rest as mere 'flyover country', while rural dwellers and Southerners fall prey to a nativism based on 'blood and soil'-which is what white supremacists (the very same ones who Trump described as 'very fine people') chanted as they marched down the streets of Charlottesville, Virginia.
The only alternative to this chasm is a shared identity, a love of country based not on a misplaced sense of racial superiority, but on the fact that our homeland stands for noble universal values. Emmanuel Macron calls himself a proud French patriot because France gave the world liberté, égalité, and fraternité. Justin Trudeau likes to say that inclusive diversity is what Canada and the Canadian spirit are all about. These are examples of what the philosopher Jürgen Habermas has called constitutional patriotism. Yes, patriotism. Liberals need not be frightened by the word.
As early as 1945, George Orwell explained the difference between nationalism and patriotism: 'By Nationalism… I mean the habit of identifying oneself with a single nation or other unit, placing it beyond good and evil and recognizing no other duty than that of advancing its interests…. By "patriotism," I mean a devotion to a particular place and a particular way of life, which one believes to be the best in the world but has no wish to force on other people' [36]. Nationalism is toxic; patriotism is not. And the best kind of patriotism is one based on age-old values such as liberty, dignity, and mutual respect. Why not call it liberal patriotism?
Now the key is to root these abstract concepts in everyday experience. If leaders talk about inclusion but the everyday experience of citizens is one of discrimination, then the rhetoric will be of no consequence. American political philosopher Martha Nussbaum, in a book entitled Political Emotions [34], has argued that the key is to elicit positive emotions toward democratic institutions, and to do this through very concrete actions, words and rituals.
Think of Lincoln and Gandhi, suggests Nussbaum. The words they uttered, the clothes they wore and the rituals they designed fostered a broad and inclusive sense of republican we. Or think of Mandela: he donned the green jersey of the white rugby team for precisely that reason. It the best example imaginable of liberal patriotism-and of democratic and healthy identity politics.

The Way Forward
The standard account of the rise of populism-call it the economic insecurity hypothesis-is an inadequate description of reality. It does not fit the facts in emerging nations such as Poland or Turkey. It has little relevance in Israel and India, where right-wing populism has clear ethnic and religious overtones. And it cannot explain the full story in the US and the UK or other advanced democracies, where support for nationalist and extremist forces goes well beyond people 'left behind' by globalization.
The conventional wisdom also fundamentally misunderstands the nature of populism. Only once we understand the identity basis of populist politics can we single out the policies-place-based policies are only one example-that can be effective in fighting populism. A focus on identity also reveals that liberal democracy needs not only a better message but also better messengers, with whom voters can plausibly identify. Liberal democrats must not spurn identity politics; instead, they need to reinvent it, helping build strong national identities based on shared liberal values like dignity and respect. Put it this way: Joe Biden's only chance against Donald Trump is to beat him at his own game.