The Political Marketplace

In November 2020, Donald Trump became America’s first one-term president for nearly three decades. True, he didn’t vacate the White House without an ugly fight, and American politics have probably been left more polarized as a result of his presidency. But, even if only just, the American electorate delivered Trump a rebuke that is unusual in recent political history. In the postwar era, George H. W. Bush, Jimmy Carter, and Gerald Ford are the only sitting presidents to have lost their bid for reelection. Before that, we’d have to go back to Herbert Hoover’s Depression-era loss to Franklin Delano Roosevelt in 1932. One interpretation of Trump’s defeat might be that Americans got to have a good look at what populism has to offer and said “thanks, but no thanks.” Another is that despite a catastrophic twin public health and economic crisis that would normally have devasted an incumbent’s reelection hopes, Trump only lost by the narrowest of margins, with more votes than any losing presidential candidate in history. It could well be, in other words, that whatever kind of politics he represented is here to stay.

Was the Trump presidency just a blip best consigned to the history books? Or was his election the harbinger of a more fundamental shift in politics in America, and perhaps, beyond? To answer these questions, we need to better understand why populists like Trump are successful in the first place. And to do that, this book proposes that we should follow Trump’s lead and think less like political philosophers and more like CEOs. Populism, as I see it – and as I think Trump would see it too – is not a set of moral values or specific policies, but a low-cost political strategy based on direct communication with voters. This strategic approach to understanding populism may not be everybody’s cup of tea. But what we’ll see is that it provides a parsimonious explanation for when politicians will use populism to win and keep power: Populism will be most prolific when it is a more cost-effective strategy than its alternatives.

In June 2015, when Trump made his way down one of the lobby escalators at his eponymous New York skyscraper to announce his presidential candidacy, I was pulling long hours trying to finish up my first book on populism. Populism in the economically advanced West wasn’t my main focus back then, but this potential bit of political theatre had my interest piqued. What could Donald J. Trump – real estate magnate, celebrity game show host, propagator of the Obama “Birther” myth, a man with zero experience in government – possibly say to make himself look like a viable presidential contender? As he labored through his speech, there was little on show to convince me that I was watching the future Republican candidate, never mind the future president. He had neither the easy, folksy charm of a George W. Bush, nor the infectious optimism of a Barack Obama. Trump was pugnacious. He was dark; nasty, even. In his heavily improvised speech, he painted a world of economic desperation, looming terrorism, and rising crime. He called Mexicans rapists and promised to build a wall to keep them out. All politicians like to talk about their accomplishments, but Trump’s self-puffery smacked more of insecurity than authenticity: “I’m really rich, I’ll show you that in a second,” he said.²

This combination of negativity and braggadocio hardly seemed likely to win him many supporters. Unsurprisingly, the mainstream media lampooned his controversial – and frankly inarticulate – speech. However, it was precisely because what Trump said was so outrageous, so beyond the pale, that his candidacy would become such a sensation. Trump was portraying himself as the outsider, the man on horseback, who would fix a broken political system. Trump would be the anti-politician. He launched into the Democrats, of course, but he didn’t have many kind words for his own party either. He set out his stall against a political establishment he said was failing the people on trade, on immigration, on jobs, and on security. “How stupid are our leaders? How stupid are these politicians to allow this to happen? How stupid are they?” he said.

Trump beat this antiestablishment drum again and again on the campaign trail over the next fifteen months, turning the liability of his total lack of experience in government into an asset. Made for the Twitter age, Trump had – and has – a way with one-liners. He fired off epithets for Republicans and Democrats alike: “Liddle Marco” (Rubio), “Low energy Jeb” (Bush), “Lyin’ Ted” (Cruz). He dispatched Kentucky senator Rand Paul at the first televised Republican primary debate with a summary shot of condescension: “You’re having a hard time tonight,” he said. When Rubio and Cruz retaliated, the latter suggesting links between Trump and the mob, and the former repeatedly calling Trump a “con man,” the maestro of insults had the perfect comeback for an age of mass distrust in the political class: “They can say what they want; at the end of the day, they’re just establishment guys.” Goodbye Rubio and Cruz. “Establishment” former Republican governor of Ohio, “1 for 38” John Kasich, got the same treatment. The objections of Republican Party operatives and conservative public intellectuals under the Never Trump banner bounced off the Trump juggernaut like BBs from the hull of a Panzer. “Crooked Hillary” Clinton, the consummate beltway insider, was the perfect foil for his marauding campaign.³

Trump’s policy agenda was notoriously vague on details and his campaign lacked the sophisticated “ground game” of more seasoned candidates. Yet his trademarked pledge to Make America Great Again resonated. Although it later emerged that his campaign spent millions of dollars on a social media operation developed by the consultancy firm Cambridge Analytica, the bread and butter of his bid for office was the old-school mass rally. Trump’s rambling, parenthetical speeches are made to be seen and heard in the flesh, not read in a press release. Even though Trump trailed Clinton in fundraising, he held twice as many rallies as she did, often in the kinds of less densely populated places neglected by other candidates. Donning their red baseball caps, Trump devotees chanted in support of his pledges to “Build that wall!,” to “Drain the swamp!,” and to “Lock her [Clinton] up!” It didn’t matter whether he made fun of the disabled, mocked former POWs like John McCain, or even disparaged the Gold Star mother of an American Muslim soldier; nothing or no one was sacred. Confirming the aphorism that any publicity is good publicity, no matter what Trump said, to his supporters he could do no wrong. “I don’t, frankly, have time for political correctness,” he declared at a 2015 GOP primary debate. When the infamous Access Hollywood tape – a 2005 off-camera recording in which Trump boasted that when you’re a celebrity, women will let you do “anything,” even “grab them by the pussy” – hit the airways just a month before the election, the normal laws of political gravity didn’t seem to apply. Thumbing his nose so openly at polite society only bolstered his outsider status. Trump beat the odds and the establishment to take the Republican Party nomination and the presidency itself.⁴

Demanding Populism

We hardly lack explanations of the Trump phenomenon or of the rise of populism in general. For sure, each account has its own slant, but a kind of consensus has emerged: Trump’s success, like that of other populists, was based on a long-simmering conservative-authoritarian backlash by voters against liberal democracy and the economic and cultural globalization that has gone with it. Over the past three or four decades, technological change, international trade, and increasing inequality have pushed the working and lower middle classes into ever more precarious economic straits. At the same time, mass immigration and the growing political assertiveness of long-marginalized ethnic minorities have raised the anxiety of working- and lower-middle-class white majorities who fear greater competition over an ever-shrinking economic pie and resent the associated decline in their relative social status. According to this version of events, populism is on the rise because of mass disenchantment with a political establishment that has forced through this agenda of economic and cultural globalization against their wishes. The liberal democratic values that undergirded the postwar political order no longer hold sway. As a result, resentful voters have turned to populists like Trump in droves.⁵

Populism, according to this interpretation, is a distinct way of understanding the political and economic world. It is a political ideology reducible to a simple dictum: the people versus the elite. This idea, philosophy, or worldview – whatever you want to call it – underlies people’s political preferences. And what the people demand, astute political leaders will deliver. According to this view, which we might call the “product differentiation” model of politics, success is determined by the ideas and policies – the qualities of the product – offered by competing political leaders and parties. When parties of the left or right gain power in a democracy, that’s because this is what the people, or at least what a majority of the electorate, wants. If populists are successful, they too must be offering something that the people desire.⁶

In part, the ascendency of this approach stems from our reliance on the omnipresent public opinion poll. Like the drunk searching for his keys under the streetlight even though he probably lost them somewhere else, political analysts are drawn to where the data are available. Because we have mountains of figures on voters’ preferences, popular demands are an obvious basis to look to explain the rise of populism. However, the prevalence of this approach is not solely due to biases in modern research design. It has a much longer lineage in political thought that goes all the way back to classical Greece and Rome. If the masses want grain, or peace, or war, well then that is what the political elite should deliver. Philosophers like Aristotle and Cicero dismissed such popular appeals as crass, even dangerous pandering, but each conceded that political leaders needed to be cognizant of something we’d now call public opinion. If this approach to understanding politics is correct, it follows that the greater the number of people who adhere to the populist worldview, and the more intensely they do so, the more likely we are to see populists in power. Populism, by this way of thinking, is successful because voters want it; or in economic terms, what matters is the demand side.⁷

Intuitive as this kind of explanation may be, it has several pitfalls. First, it is unclear what exactly the populist ideology is, or how it works to affect political outcomes. The most common efforts to define populism as an ideology either make it so ordinary as to be indistinguishable from democratic politics in general, or they make it so egregious as to equate it with dictatorship. If populism simply means being for “the people” and against whoever is not “the people,” how many democratic politicians would not be populist? If instead populism is understood as illiberalism or anti-pluralism, given that the ability of the opposition to freely contest elections is a minimal requirement of democracy, how different is this meaning of populism from outright dictatorship? Second, even if we did agree on an understanding of populism as a set of values or attitudes, measuring them is extraordinarily difficult. Is populism a single coherent belief, or an amalgam of several different attitudes together? If it’s the latter, how should these distinct attitudes be aggregated? If populism is manifest in appeals to “the people” against “the elite,” is a single speech appealing to “the people” enough? If not, how often must a politician invoke “the people” to qualify as populist? How can we reconcile populism as a coherent set of values when it can take such wildly different forms as Trump’s xenophobia on the one hand and Chávez’s redistribution of wealth on the other? Third, the best evidence shows that when factors like a voter’s personality, policy preferences, and other political attitudes are taken into account, so-called populist values have at best a marginal effect on vote choice. What exactly is it that populist beliefs by themselves do? If populism is just being used as a synonym for nativism or socialism, what is the concept adding to our understanding of politics? Last, even if we put these conceptual and measurement issues aside and accepted that populist attitudes or policy preferences might explain why one person is more likely than another to vote for a populist, this still would not account for change in the relative success of populists over time and in different countries. Given that people’s values change slowly, how can we account for the swift and sometimes erratic shifts in populists’ vote shares? Why do similar grievances not produce the same degree of populist success in different cases?⁸

In this book, I’m largely going to set aside the worries, beliefs, ideologies, and policy preferences of voters – the demand side – that animate most accounts of populism. Following that well-known principle of economic analysis, ceteris paribus – all else equal – my approach is to hold the demand side constant and see just how much can be explained by looking at what happens when there are changes to the supply side of the equation. In other words, rather than asking why people supposedly want populism, I think we can learn a great deal by examining changes in the options that political leaders supply voters with instead. If, as a result, this book appears one-sided in its focus on populists rather than their supporters, this is not because I believe the demand side is irrelevant. In the concluding chapter, I’ll suggest how we could develop what economists would call a “general equilibrium model” that brings together both the supply and demand sides. My main aim, however, is to restore balance to a field that has become excessively focused on just one side of a complex problem. By examining the supply side, this book will show that populism has a clear economic logic. But before we get there, we need to be clear about just what it is we’re trying to explain. What exactly is populism?

Populism as Strategy

Populism is a famously, frustratingly disputed concept. Although it would be tempting to believe that disagreement over the meaning of populism is due to the post–Trump surge in interest in the subject, the problem of definition has been around for a long time. Back in 1967, a group of prominent social scientists got together at the London School of Economics to try to distill from a wide range of national and historical experiences a shared understanding of populism. The published collection of papers that emerged from that conference is full of insights and still repays reading, but as the editors of the volume acknowledged, they could not establish the conceptual common ground on which future writing on populism would build. In his contribution, Peter Wiles wrote “to each his own definition of populism, according to the academic axe he grinds.” Fast forward half a century and the fact that one of the best-selling books on the subject is entitled What Is Populism? is telling of how little agreement there still is on what populism actually means and on who or what qualifies as populist.⁹

The reality is that there is no true definition of populism. It is, like democracy or justice, one of those essentially contested concepts about which philosophers will forever argue. Yet this doesn’t mean that we should just pick a definition at random. Ask a poet and a neuroscientist to define love and you’ll get two equally true but very different answers. What we need is a definition of populism that is useful, and fortunately, there are ways of deciding what this would look like – at least for the purposes of political scientists and economists if not philologists. Obviously – but I would also say, trivially – a useful definition of populism should allow us to distinguish populists from non-populists, to separate full populists from partial populists, or to say whether one politician or party is more or less populist than another. But just as importantly, a useful definition should facilitate a better understanding of populism’s causes and consequences. It should help us to make clear, testable predictions about the conditions under which it will be successful, the effects it will have on democracy or the economy, and so on; even better, a useful definition will lead to policy remedies. It is with these purposes in mind that I define populism as a political strategy, in which the leader of a personalistic political movement appeals directly to the people through mass communication to win and/or keep power. Populism, in short, refers to certain actions or practices, rather than to a set of beliefs or doctrines. It is something that politicians do rather than something they believe.¹⁰

Although this understanding of populism as a political strategy is not currently the predominant approach in academia or mainstream punditry, it has a dignified pedigree, going back to one of the fathers of political economy, Max Weber. Weber famously argued that there are three main sources of political authority: the bureaucratic, the patrimonial, and the charismatic. Bureaucratic authority derives from its dependence on rules and procedures, which are, at least in theory, open and unbiased. This kind of authority is viewed as legitimate because of its procedural fairness. Patrimonial authority instead is based on tradition; the authority of kings, for instance, depends not on talent or justice, but on heredity. Such a system may be less open, but it has the benefit of being predictable. In contrast, charismatic leaders depend on neither rules nor tradition for their authority. Charismatic authority instead rests on a direct relationship between leader and follower, where mass belief in the unique qualities of the leader forms the basis of his power.¹¹

In what remains for me one of the most insightful analyses of populism published to date, Greek sociologist Nicos Mouzelis wrote that populism is best understood as a type of relationship between party leaders and voters – or what he called a “mode of incorporation.” Mouzelis argued that the people don’t just exist as some abstract mass of humanity that shows up at the ballot box of their own volition come election time. The public is deliberately “incorporated” or brought into the system by political leaders. Politicians, as we well know, persuade, cajole, and even coerce. Drawing on Weber’s three sources of authority, Mouzelis argued – as I do here – that there are basically just three ways of organizing the pursuit of power in a democracy: programmatic, patronage, and populist incorporation. Mouzelis stresses that populists communicate directly with the people, rather than working through intermediaries as in the case of programmatic and patronage-based party leaders. He put it like this: “As a rule, populist leaders are hostile to strongly institutionalized intermediary levels … The emphasis on the leader’s charisma, on the necessity for direct, nonmediated rapport between the leader and ‘his people’ as well as the relatively sudden process of political incorporation all lead to a fluidity of organizational forms.” Within the movement or organization, power is vested in the person of the leader. The leader’s authority is essentially arbitrary, in that it is only minimally constrained by rules, roles, or procedures – populist parties are organizationally “fluid.” Outside of the party, populism implies a direct relationship between leader and supporter, which, as much as possible, is unfiltered by party officials, local elites and bosses, newspaper editors, and other intermediaries. The way in which political movements are organized, or what we might call their corporate structure, is critically important to understanding the utility of these programmatic, patronage, and populist strategies.¹²

Programmatic parties are complex and usually large bureaucratic organizations, with regular procedures governing internal promotion and candidate selection, professional staffs, permanent offices, and a generally high level of institutionalization. Internally, authority in the bureaucratic party rests in roles or offices – party chairman, whip etc. – rather than persons. As much as any firm, programmatic parties are professional organizations. Programmatic parties provide career paths open to talent – including, of course, the talents of scheming and manipulation. Programmatic parties typically have well-established links with social and economic organizations such as unions, farmers associations, and churches. As a result, they’re often identified with particular interest groups and policies. Party leaders’ links with voters are also heavily mediated by a dense organizational ecosystem that includes party workers, civil society organizations, and the state bureaucracy itself. Additionally, bureaucratic parties have often been mass membership organizations, funded by member dues, although this is less the case today – and has always been less the case in the United States than in Western Europe or the Antipodes. Programmatic parties take a great deal of time to build, and have a corporate personhood that extends beyond the term of any individual leader or cohort. The canonical examples are the Conservative and Labour parties in Britain and the Democratic and Republican parties in the United States.

Patronage-based parties are looser coalitions of political factions or groups. Leaders – or patrons – gain and retain power by judiciously distributing rewards – or patronage – to their supporters or clients. This patron–client form of politics has a long ancestry, epitomized in the pyramidal feudal system of kings, vassals, and peasants in Medieval Europe. In its modern incarnation political leaders win power by buying votes through a network of allied elites and political brokers. At the level of interaction with voters this form of retail politics is often called clientelism, money politics, or just vote buying. Providing jobs in the public sector was how the legendary Tammany Hall political machine in New York maintained its power, with a third of Democratic voters holding a Tammany job in the 1910s. Similarly, as late as the 1960s, Chicago Mayor Richard J. Daley’s Cook County – or Crook County – organization traded votes for some 30,000 public sector jobs. In this book, however, we’ll be more concerned with the higher-level integration of the leaders of rival but functionally similar political factions. Just as voters are bribed to cast their ballot, individuals who control blocs of votes – brokers – are in turn courted by party leaders. Ministerial appointments, government contracts, and other sinecures are the currency of patronage party loyalty. Leadership within the patronage party is governed by the strength of rival factions of patrons, brokers, and clients. Factions will come together to gain and keep power, but the association is an instrumental one, borne out of self-interest rather than out of a deep sense of loyalty or shared ideology. As we’ll see in Chapters 3 and 4, the distribution of patronage among office-seeking elites was a major occupation of political leaders in the early American republic. In this, Americans were continuing a practice perfected by the famous eighteenth–century British Whig leader and prime minister Sir Robert Walpole. With the demand for patronage always exceeding its supply, Walpole had to judiciously allocate places and pensions to build and keep his majority in the House of Commons – a strategy he executed successfully for some two decades.

Instead of climbing the rungs of the party ladder or forging transactional alliances with supporters, populists gain power by directly mobilizing a mass support base. That is, they communicate directly with voters rather than mobilize them through intermediaries. In populist organizations, memberships and offices are often poorly defined and subject to arbitrary change from above. Preferring to target free-floating or independent voters, some populist parties, such as Geert Wilders’ Party for Freedom (PVV), have no membership system at all. Internally, in a direct inversion of the programmatic party structure, individuals matter more than roles. Pure populist parties, moreover, do not have regularized procedures for leadership replacement or succession. In short, a populist leader is unconstrained by rules or by dependence on factional support, which creates a very different relationship between a leader and his political associates than in the case of more deeply institutionalized bureaucratic or patronage-based parties. Even the most established of populist parties are, by definition, the tools of their charismatic leaders.

In practice, some leaders will mix these programmatic, patronage, and populist approaches, and the composition of their strategic portfolio may change over time. For instance, two-time Greek prime minister (1981–89 and 1993–96), Andreas Papandreou, came to power as the charismatic leader of the Panhellenic Socialist Movement (more commonly known as PASOK), but once in power he depended more and more on the distribution of patronage to maintain himself in office. Going in the other direction, the subject of my first book, Indira Gandhi, prime minister of India from the mid- 1960s, shifted from a patronage-based to a populist strategy to retain power after a faction of her party attempted to oust her from power. In places like the United States and the United Kingdom, where two main parties have usually exhausted the political space, successful populists have typically adopted a mixed strategy – for instance, populism to gain control of the party apparatus but then the use of programmatic or patronage-based mobilization to succeed in a general election. However, as noted previously, money and time spent on one strategy cannot be spent on others. There are, in economic terms, opportunity costs to any chosen strategy. Aspiring leaders must therefore trade off a concentration on one approach against another. Populists depend mostly on the use of a highly personalist organization that makes direct appeals to voters through whatever the mass communication media of the time happen to be.

Populism in this sense is a matter of degree. Determining whether an individual leader is a “populist” means we need to set a somewhat arbitrary threshold for what “mostly” means. I take a relatively restrictive approach, but there is no reason that a more permissive one couldn’t be used. For any given leader, we want to know how much their strategy approximates the populist ideal type just outlined. I’ve previously suggested several practical questions we can ask of any given leader to help make this judgment. As mentioned, populism has both an internal and external dimension to it. Internally, populists have essentially arbitrary authority within their own personalistic political organization, while externally, they link with voters directly rather than through intermediaries. Along the internal dimension, the critical questions are: Is the movement or party one that the leader formed as a personal electoral vehicle? Is authority within the leader’s party or movement arbitrary – completely at the discretion of the leader – or rule based? Does the leader control appointment decisions or is leadership/appointment determined by ballot or some other collective procedure? The main questions to ask with respect to the external dimension are: Does the leader’s movement or party rely primarily on mass rallies, mass media, and social media to mobilize electoral support directly, or does it rely primarily on mobilizing voters through its membership, allied unions, churches, or other organizations, or on systematic clientelism? Is the leader himself/herself the primary object of a campaign or is it a party’s historical political/group/ethnic linkages to a constituency? Answering these questions, and perhaps others like them, allows us to build up a picture of how much a given politician relies on the populist strategy to win and keep power.¹³

Understood in this way, the strategic approach to populism is a good fit for most of the usual suspects: Donald Trump, Geert Wilders, Alexis Tsipras, and Silvio Berlusconi among others would all qualify as highly populist; so would less frequently examined populist leaders such as Charles de Gaulle, Huey Long, or Wendell Willkie. Others, including Andrew Jackson, David Lloyd George, and Jimmy Carter, would also qualify as at least partly populist by these criteria. Consistent with the typical understanding of populism, in my estimation, the criteria would exclude party leaders such as Ronald Reagan, Barack Obama, Margaret Thatcher, or Tony Blair as populists, however personally telegenic or popular they might have been.

It is also the case, however, that other leaders or parties in the contemporary European far right, who are often classified as populist by other scholars, would not count as populist according to the strategic approach. For example, the strategic approach would not classify parties such as the Alternative for Germany (AfD) or the Sweden Democrats as populist; for these parties, the organizational structure is too collective or corporate for them to qualify as populist. Given the tendency to use populism and nativism interchangeably in both academic and popular writing, this omission may bother some readers. However, the problem with the critique that it is “simply impossible to apply [the strategic definition] to European populist parties such as [X, Y, and Z],” is that it begins with the premise that parties X, Y, and Z are in fact populist. This has the problem of making and validating concepts backward. We cannot know if parties X, Y, and Z are populist until we have a definition! Scientific concepts need not exactly resemble their folk equivalents. Populism is a term thrown about so casually that to begin from the view that everything ever labeled as populist is populist would be very problematic. From here on, then, I treat the objection that “you do/don’t include [insert party name here] as populist” as specious.¹⁴

It is also important to note that if populist parties are “personalist” parties, they are not merely so. Critical to the populist strategy is a reliance on mass communication with supporters that need not be true of personalist parties in general; the latter can primarily exploit kinship networks, patronage, or even more coercive techniques. This strategic approach also differentiates populism from authoritarianism. In the same way that we typically distinguish between democratic and authoritarian government more generally – by whether or not a regime has free and fair elections – we can distinguish between populist and authoritarian leaders. To the extent that coercive tactics – such as censoring the media or arresting opponents – predominate over mobilization through mass communication, a leader is better classified as authoritarian rather than populist. For instance, while Adolf Hitler captured power through an essentially populist strategy, once in power, the Nazi regime became brutally authoritarian. Even if Hitler himself retained much of his mass support, following the passage of the infamous Enabling Act of March 1933, he is better classified as a “popular dictator” than as a populist.

Contrary to other approaches, the strategic understanding of populism makes no assumptions about the content of a leader’s appeals or the audience to which they are directed. Decisions about content and audience, we might say, fall into the realm of tactics rather than strategy. The strategic approach is basically consistent with one that classifies populists according to their supposed ideology, but it views discourse – talk – as cheap. Anti-elitism has become so common today as to be totally unexceptional. None except the most politically suicidal leaders would now dare to admonish the masses to leave matters of government to their betters. At the same time, appealing to a virtuous if vaguely defined “people” makes good sense for a politician without strong and stable links to a body of supporters. Sometimes the “people” applies to the lower class, sometimes to the native born, and at other times to the law abiding. In other instances, its application is even less precise. Vague appeals to the people are all too easy to make. There is nothing particularly moral or democratic about them. A “dangerous ambition,” Alexander Hamilton wrote, “often lucks behind the specious mask of zeal for the rights of the people.” No particular group of political leaders, moreover, has a monopoly on this kind of rhetoric. In the words of an anonymous Columbia County, New York, contemporary of Hamilton: “[I]n a representative government, where every act is the act of the people, to talk of ‘an appeal to the people’ is nonsense. They are random words, used by superficial politicians and designing men … in hopes of political advantage.” This is a remarkably insightful critique in itself, but what’s most notable is that it came not from a paper of Hamilton’s elitist Federalist Party but from a paper of its democratic Jeffersonian opponents! The anti-elitism typical of populist rhetoric is little more than a politically acceptable, catchall way of demonizing incumbents, while people-centrism is the ideal way of appealing to an ill-defined constituency through the thin but broad approach of mass media. It’s a cynical view perhaps, but appealing to the people against the elite is, we might say, just good marketing.¹⁵

In this sense, more important than what populists say is how they say it. Or to use the adage of Canadian communications theorist Marshall McLuhan, “the medium is the message.” The technology of mass communication has varied enormously over history. In Ancient Greece and Rome, populists almost physically connected with crowds of people in the assembly and the forum. Public oratory and lavish parades were the way a man could demonstrate his eminence to the masses. By the time mass democracy reemerged in revolutionary America and France in the late eighteenth century, the diffusion of printing technology and the development of more efficient postal services meant that populists and their entourages could also use pamphlets and newspapers to communicate cheaply with larger numbers of people than ever before. Mass communication took on added sophistication in the interwar period, nowhere more so than in Germany. Hitler’s Mein Kampf presents a sophisticated if nefarious understanding of the role of mass propaganda in politics. Effective political speech, he argued, must be plain, direct, emotive, and, above all, repetitive. Technological changes served to make direct communication more efficient. In power in the 1930s, the Nazis exploited government control over mass media and the arts to great effect; even the airplane was turned into a symbolic form of communication. Perhaps no brief radio address has so indelibly marked a political career as Charles de Gaulle’s Appel de 18 Juin, 1940, in which he announced himself as the exiled head of the resistance to the Nazi occupation of France. In the United States, as early as Dwight D. Eisenhower’s election campaign of 1952, the television political “spot” – the vacuous thirty-second soundbite still so prevalent in politics today – had become a reality. Not all populists are great orators. Sometimes image speaks louder than words. The politician who didn’t look good on the box – think of the febrile Richard Nixon sweating through the first presidential debate with John F. Kennedy in 1960 – paid the price. Even though television – an expensive medium – was co-opted for a time by well-organized and well-financed political parties, eventually outsiders such as Ross Perot and Donald Trump became rich enough to afford to use it for themselves. As political scientist Martin Wattenberg could conclude in the 1990s in the fifth edition of his seminal book on American political parties, “[w]here once candidates for public office had to rely on mustering organizational strength to communicate with voters, it is now increasingly possible for them to establish direct contact through the media.” The arrival of social media, especially Twitter, along with other emerging technologies – perhaps most unusually, holography in the case of far–left French presidential candidate Jean-Luc Mélenchon – has allowed populists direct access to our homes and minds like never before.¹⁶

Maybe discourse doesn’t count for much, but might it be that populists target specific groups for support – the “white working class,” for instance? Not in any consistent way. A person’s level of education is negatively associated with support for populist politicians in some cases but positively in others. Income, too, is a weak predictor of populist support. The urban working class might be a likely base of support in a country without a strong socialist party, but in other places, the suburban middle classes could be a more receptive audience. In turn, a policy message stressing economic inequality or immigration might play better in one state, while a law-and-order appeal could resonate in another. Thus, those who have interpreted populism as a set of socialistic economic policies – common in the understanding of populism in Latin America in the 1990s – or as a set of hard-line law-and-order policies – sometimes called “penal populism” – may be right, but only for special instances. No individual policy is definitive of populism in general. Populists, as power-seeking strategists, tailor their appeals to the context. Racism, nationalism, socialism, chauvinism, and other such persuasions are all vital issues, but we shouldn’t muddy the waters by making them synonyms for populism. Populism is a political strategy that relies on mass communication, whatever the technological medium, to win and keep power; no more, no less. As I’ll describe further in the next section, populism in this minimalist strategic sense matters deeply for things we care about, for example, civil rights, economic prosperity, and perhaps even democracy itself. Moreover, if my explanation for why the populist strategy worked for Trump and others holds water, we’ll have some potentially actionable ways to reform our political institutions in the decades to come.

Transaction Cost Politics

Although other scholars have already made a strong case that populism can be best understood as a distinct political strategy, their approach hasn’t gained as many followers as it might have. The reason, I think, is that they haven’t yet provided a fully satisfying explanation for why aspiring political leaders would adopt the populist strategy over others. Each strategy – the programmatic, the patronage, and the populist – comes with its unique set of costs, risks, and potential rewards. Why populism? is not a question of ideas but of a hidden strategic calculus.

If votes are the currency of democratic politics, political leaders, like entrepreneurs in a marketplace, want to grow their share at the lowest cost. If this analogy is correct, then there is a lot we can learn from the economics of how and why firms – and their leaders – succeed. Ford Motors, the pioneer of the mass-produced automobile, provides an apt illustration. Ford’s Model T was the first vehicle to be made by the assembly of identical and interchangeable parts. Ford sold 16.5 million units of the Model T, making it still one of the best-selling automobiles of all time, and in the early decades of the twentieth century, Ford Motors one of the most profitable firms in the world.

We tend to think that success in the marketplace is driven by offering a superior product – the product differentiation rationale noted earlier. But price is just as important. Ford thrived because it vigorously drove down the costs of production. Compared to the previous practice of crafting vehicles by hand, the reduction in the price of producing an additional car on the line – or its marginal cost – was staggering. In 1929, Ford could turn out a finished automobile every ten seconds; just two decades previously, it had taken fourteen hours. Over that period, the price of a car fell from the equivalent of two years’ worth of the typical worker’s salary to the amount he could earn in just three months. Ford’s vehicles were no better than those of its competitors, such as Daimler, but they were significantly cheaper. Costs matter critically for success in any marketplace – the political one should be no different. The more cheaply a politician can win votes, the more successful he’ll be.

For our purposes, Ford’s story is illustrative in a second, more subtle, sense too. It demonstrates why some ways of organizing a firm are more cost-effective than others. One of the means that owner-manager Henry Ford used to reduce costs was to bring most elements of the production process in-house, in a process known as vertical integration. Ford Motors famously made almost all of its own parts, assembling them at its massive production facility in Highland Park, Michigan. This allowed Ford to apply his principles of efficiency to lowering the cost of each input. However, the enigmatic Ford still wasn’t satisfied. Why was Ford Motors still outsourcing the production of some of its parts? Why, for instance, was it buying tires when it could make them itself? Why buy rubber on the open market when it could grow it too? Ford took the idea of vertical integration – bringing all parts of the supply chain within the firm – to its logical conclusion. Frustrated by what he perceived as the exorbitant price of rubber on world markets, Ford bought up 140,000 hectares of rain forest in Brazil and set up his own rubber plantation. But the endeavor was a miserable failure. Why?

The answer is suggested in a remarkable 1937 paper by economist Ronald Coase. Producing something like an automobile entails two kinds of costs. The first, the direct costs of production, are easy to understand. These are the costs of the parts and labor that go into the end product: the metal; the rubber; the leather; and the wages paid to managers, factory workers, and salesmen. The second, indirect costs, or transaction costs, are somewhat harder to grasp, but they turn out to be very important to understanding the political trade-offs at the heart of this book. In economic terms, transaction costs include the search costs, bargaining costs, and enforcement costs incurred as part of an exchange between buyer and seller. The time spent finding an appropriate supplier, negotiating an acceptable price with them, and then ensuring that the product or service is delivered as ordered all incur costs on a business. What Coase showed was that these transaction costs set optimal limits for the size of the firm.¹⁷

Coase recognized that in a world without firms, almost any complex economic activity, such as making and selling a car, would necessitate multiple transactions conducted between countless individuals in the marketplace. The transaction costs involved in all of this marketplace activity would be astronomical. In contrast, in a vertically integrated firm like Ford, an employee is contracted to perform a wide range of tasks as demanded by his employer. This reduces the transaction costs involved to a single contract. Costs, in short, are lowered by bringing many processes in house, favoring huge, vertically integrated firms like Ford. However, as Ford’s Brazilian misadventure shows, sometimes it is more economical to outsource activities to the marketplace. Rubber could be procured more cheaply on the market than it could in-house. In part in recognition of the comparative advantage possessed by suppliers with specialized knowledge or unique resources, today Ford outsources the production of about 70 percent of the parts, components, and other services that go into its vehicles. Firms grow to the point where they minimize transaction costs, and these costs change over time, sometimes favoring integration, sometimes separation. Not product, not philosophy. The shape and size of the firm are determined by cost.¹⁸

In the political marketplace, the “buyers” are politicians while the “sellers” are voters. Although other writers have used this terminology before, most existing approaches to political behavior assume what economists would call a frictionless market. In this approach, it is only a slight oversimplification to say that the share of votes that various parties receive is assumed to be axiomatically equal to the proportion of the voting population who favor those parties. In this sense, political outcomes can be reduced to a quantitative matching of policy offerings and voter preferences. In reality, however, not only does casting a ballot represent an opportunity cost in terms of time and effort for each individual voter, but every vote won by a party requires the expenditure of human and capital resources to mobilize voters. In Coase’s terms, the process of getting votes carries substantial transaction costs. Together, the direct and indirect costs of winning votes are central to understanding the prevalence of populism across time and place.

For the patronage-based party, the monetary cost of winning votes is on crude display. In some cases, candidates and their middlemen – brokers – distribute cash, merchandise, and of course, booze on and before election day. In George Washington’s Virginia, “swilling the planters with bumbo” was a near universal practice before an election. Writing to one of his agents on the eve of his first successful election to the House of Burgesses, Washington’s concern was less with the propriety of bribing voters than that his agent would fail to spend enough to slake the voters’ thirsts. In some cases, a vote has a more explicit monetary value. As a campaign manager in New Deal–era Texas, prior to Election Day a young Lyndon Johnson laid out stacks of five-dollar bills on a big table in a room of the San Antonio Plaza Hotel. Mexican-American men would come into the room, tell Johnson a figure, and Johnson would dole out that number of five-dollar bills. Five dollars was the cost of a Chicano vote in 1930s San Antonio. Somewhat more subtly, political machines like those of Tammany in New York or Mayor Richard Daley in Chicago thrived through to the 1960s by maintaining a lock on jobs in the public sector that they reserved for their supporters. Patronage politics is the classic quid pro quo. A direct cost, the bribe, is exchanged for a vote.¹⁹

From an economic point of view, however, we can think of patronage politics as incurring a second type of cost: a transaction cost. The great dilemma for the politician engaged in vote buying – ethical quandaries aside – is that voters cannot be trusted to keep to their end of the bargain. As Washington himself knew, there was no guarantee that a feast of rum and sandwiches would necessarily translate into votes. Voters can and do accept gifts from multiple candidates running in the same election. Politicians therefore have to spend additional resources in monitoring the behavior of voters – enforcing the informal contract in other words. This can be done in any number of imaginative ways, but each one incurs a cost. In post–Reconstruction New Orleans, the electoral commissioners under the control of the Old Regular machine could, by law, accompany voters into the polling booth; if a man had promised to sell his vote, the commissioner exited the booth with his pencil behind his ear to signal to a watcher that the voter should receive his bounty. In 1940s Texas, the managers of the feudal political machines of the state’s oil barons would often pay voters’ poll taxes directly and have a fixer complete the ballots, or they’d also pay off the appropriate officials and count the votes up as they saw fit. Commissioners or brokers had to be paid, and to add a further layer of complexity, these middlemen were hardly any more trustworthy than voters themselves. Brokers too are often happy to work for more than one candidate at once. Whichever candidate gives him the biggest payoff is the one to whom he will deliver his bloc of votes on Election Day. In Gilded Age New York, Joseph Murray, a brawling political fixer in the Democratic Party’s Tammany machine, felt inadequately compensated by his erstwhile patron and so surreptitiously directed his gang of enforcers to provide their “services” to the Republican candidate instead; Murray had found better remuneration with his newly elected GOP patron. Voters need to be monitored; monitors need to be monitored. Bring it all up a level, where brokers at each level get aggregated into municipal or regional blocs, where these blocs in turn comprise factions within a national patronage party, and the escalating nature of these transaction costs becomes clear.²⁰

Yet, expensive as it may appear, most of the money that fuels the patronage strategy is bilked from public sources rather than private ones. For the political leader, we might say it’s a publicly subsidized political strategy. The patronage strategy can provide a low-cost entry into politics, as political hopefuls can promise to deliver the spoils to their supporters if they win. The economic approach predicts that patronage should be the modal form of mobilization in new democracies: If political organizations are expensive and time consuming to build, it makes sense to draw on public resources in the short term. Although the evidence is disparate, there is good reason to think this has historically been the case. From the Americas (North and South), to South and Southeast Asia, to sub-Saharan Africa and the Middle East, patronage-based linkages appear to be by far the most common ones used during the early period after democracy is introduced. Even in some states in postwar Western Europe, most notably Italy, Austria, and Greece, a similar pattern of linkage prevailed; go back a little further, and the same was true of England and the Netherlands.

Yet if patronage has a compelling economic logic, especially in the early years after democratization, there are good reasons why it becomes less efficient over time. As I argued in my first book, Populism and Patronage, there is a steeply increasing cost involved in managing the patronage racket, one that’s hinted at in the often precariously short lifespan of the Mafia Don. Brokers have distinct interests from their patrons, leading to what is called a principal–agent dilemma. Political leaders – the principals – want votes at the lowest unit costs; brokers – the agents – in contrast, aim to maximize the fee they can extract for delivering those votes. Brokers, as we’ve seen, are disloyal, ever willing to sell their votes to the highest bidder. One historian likens the vote captains of mid-Victorian England to condottieri – the military mercenaries of Renaissance Italy. Mercenaries sometimes make effective soldiers, but as Machiavelli noted, they are “disunited, ambitious, and without discipline.” Patronage parties thus find themselves engaged not just in a conflict with rival parties but in a perpetual civil war over the spoils. Even if power can be bought with a patronage party, the patron faces the perennial threat of her subordinates’ defection. As long as the number of brokers and clients is limited and as long as brokers are tightly constrained from above, patronage is often a cost-effective strategy. However, the costs of maintaining a winning coalition of factions tends to rise over time, as faction leaders exploit their pivotal position for personal gain. The more autonomy, power, and resources that brokers accumulate, the more expensive it becomes to buy them off. Patronage-based costs thus have a kind of ratchet quality, almost always tending upward unless more authoritarian measures are used to cut the brokers down to size.²¹

Direct vote buying sometimes still happens in consolidated democracies like the United States or Australia, but it’s unusual now, and when found out, it causes a scandal. As a result, because most of the costs in a typical Western democracy are indirect, transaction costs, it is easy to forget just how expensive it is to seek and keep power. As a result, we typically think about political competition in the West solely in terms of ideological and policy cleavages. In reality, however, the transaction costs facing party leaders in programmatic political systems are qualitatively different but hardly less onerous than those in patronage-based ones. Some transaction costs, such as campaign advertising and staff, are easily measured. In the US in 2016, a staggering $6.4 billion was spent on election campaigning, a third of this total on the presidential campaign alone. In 2020, that total figure more than doubled to $14 billion. Former New York mayor, Michael Bloomberg, reportedly spent over a billion dollars of his own money contesting the 2020 Democratic presidential primary.²²

By themselves, these are massive sums, but many of the transaction costs in programmatic party systems are hidden and may not even be monetary. They include any cost, whether in terms of cash, effort, or time that goes into affecting a voter’s attachment to a given party and her likelihood of turning up at the polling booth. Consider that in late nineteenth–century America, some 5 percent of the entire adult male population was a party worker, each of whom dedicated about ten to fifteen hours to party activity each week. Obama’s 2008 campaign made use of the “free” labor of over two million volunteers. Furthermore, many hidden costs are often historic and built into party organizations themselves. In programmatic party systems, voters are wooed, socialized, and deeply embedded into party networks by professional organizations over many years. Building and maintaining the nationwide organization of party branches and officials needed to accomplish this is a time-consuming and expensive activity. Bureaucratic organizations also need constant upkeep, and, even when voters are sympathetic to a particular party or candidate, getting them to make the additional effort to turn up at the ballot box is a colossal – and increasingly expensive – job. Furthermore, even if bureaucratic parties don’t explicitly give handouts to voters, they use their organizational resources to occupy government at multiple levels. Parties mediate between people and the state. Without doing anything corrupt, parties help citizens negotiate complex bureaucracies, and advise on matters big and small. This investment of time and organization is not money going directly to voters directly but is another indirect cost of mobilizing their support. At a more abstract level, party brands themselves encapsulate tremendous value. Parties that have presided over a victorious war, an expansion of the suffrage, or even just sustained economic growth enrich their historic brands. This kind of reputation cannot be easily quantified or bought.²³

For any political hopeful, the laundry list of operations required in building and maintaining such a party bureaucracy is potentially overwhelming. The costs of building and maintaining these kinds of parties is hard to calculate, with much of their capital embedded in intangibles like their brand, historical political loyalties, and their functional experience as actors in government. University of Washington political scientist Margit Tavits’ description of the dilemma that faced political leaders in Eastern Europe on the transition to democracy in the early 1990s is revealing; she writes “building a strong organization is costly … because it takes resources and requires long-term and continuous commitment.” Aspiring political leaders “would prefer to avoid building such an organization if elections could be won some other way.” It is by definition beyond the means of any individual politician. However, because of the great advantage that programmatic parties have in getting out the vote due to their mass memberships, strong brands, and well-oiled local organizations, the power-seeking politician may do well to aim at helping to build or winning control of such a machine. The rub, for the individual politician is that he must sacrifice his autonomy – the programmatic party is inherently corporate in nature. As Napoleon understood, “[t]o govern through a party is sooner or later to make yourself dependent on it.” The party politician may aspire to leadership; he may even be the favorite to acquire it; but a programmatic party requires rules and procedures that allow for the removal of any individual. The programmatic strategy entails a costly sacrifice of autonomy that some politicians are unwilling to incur.²⁴

By contrast, the populist strategy mitigates the risks and costs of either having to manage rival individuals and interest groups within a complex bureaucratic organization or of having to share the spoils to keep the competing factions in a patronage party satisfied. The modern populist strategy certainly requires some hard cash; direct communication, especially through television and social media advertising, can be costly, and even public rallies require a great deal of staff, incur administrative and regulatory fees, and so on. It is no coincidence that some of the best-known populists – Thailand’s Thaksin Shinawatra, Italy’s Silvio Berlusconi, and of course, Donald Trump – were already fabulously wealthy when they turned to politics. “Brand awareness” is not bought cheaply, which often limits populism to those with some existing celebrity status. Nor are the costs of mass communication an exclusively modern concern. When, early in his career, Julius Caesar put on lavish public games to win over the Roman crowds, the costs were nearly crippling.

Yet by comparison, when all of the indirect costs of bureaucratic or patronage-based party building are considered, populism is still a minimalist political strategy. Relying on personalist political organizations and direct mass communication, populism, Columbia University political theorist Nadia Urbinati writes, “is an affordable politics.” Especially for politicians with some preexisting name recognition – successful generals, mayors, or even celebrities – populism provides a uniquely cost-effective route to power. It involves minimal expenditure on organization and avoids the costs of having to monitor agents and brokers. Critically, populists reduce the in-house costs of seeking votes by avoiding the expense of building a national rule-based party infrastructure (programmatic incorporation) or using intermediaries to buy votes on their behalf (patronage incorporation). Populists, like many modern firms, are outsourcers. More akin to today’s Ford Motor Company than its 1920s predecessor, populists operate relatively lean and low-cost organizations that seek to do one thing well – the exploitation of unmediated mass communication – thereby lowering the transaction costs involved in getting a vote and ultimately, buying power. As Ecuadorian president, José María Velasco Ibarra, boasted, all he needed to win power was a balcony from which to speak to the people.²⁵

Why Populism?

If the programmatic, patronage, and populist strategies each has a distinctive set of costs and benefits, the obvious question is what conditions make the populist strategy more effective than its alternatives? The succinct answer is that populism will be the most efficient strategy when the direct mobilization of a weakly attached plurality of the electorate is sufficient to win and keep power.

The first part of this answer has to do with the preexisting patterns of voter loyalties. Populism is a quick but superficial way of persuading voters. Populists simply throw appeals out on the wind to see what sticks. Based on mass communication, whether public oratory, television advertising, or social media, the populist leader can only really hope to establish weak and transient links with potential supporters. One key condition for populist success is the availability of voters who aren’t already enmeshed in richer relationships with other parties. As Kurt Weyland, professor of government at the University of Texas-Austin, puts it, populists seek “support from large numbers of mostly unorganized followers.” This means that populists do not primarily target the already committed union member and frequenter of the local Labour Party pub, or the public employee who owes his job to the local Democratic Party boss. Direct appeals through impersonal mass communication are rarely sufficient to break such ties – ties that may be as much social as political. Where voters are relatively unattached, however, populist mobilization can deliver mass support at relatively low cost. This, in turn, raises the question of why voters are available in some countries and at some time periods but not others. No doubt, as previous research has shown, events like economic crises or corruption scandals matter in disrupting party allegiances. Such shocks often provide the impetus for a sudden decline in voter loyalties to established parties. However, such calamities are neither necessary nor sufficient to explain voter availability. The diversity of cases that I cover over the coming chapters reveal a more comprehensive set of sources of voter availability than has previously been described.²⁶

The second part of the answer is one that I think is little noted in modern research on populism. Populists, surprisingly perhaps, need not be all that popular. What they rely on, as much as the raw availability of potential supporters, is a divided opposition. In a two-horse race, rounding up a little, a successful candidate needs the support of 51 percent of the votes cast; in a three-horse race, if votes are distributed evenly, the winning total can drop as low as 34 percent; in a four-horse race, as low as 26 percent; and so on. Low turnouts mean that the proportion of support from the whole electorate needed to win can be much smaller than the proportion of valid votes that are cast. Additionally, electoral colleges and other rules that make representation disproportionate can reduce these winning percentages even further. Either way, the more divided the opposition, the lower the proportion of the electorate a populist needs to woo. The fewer supporters needed, the lower the cost. Populists very rarely command the backing of a large, or even bare, majority. Populists most often come to power when a plurality is sufficient. Silvio Berlusconi first became Italian prime minister in 1994 with little more than a fifth of the vote. French president, Emmanuel Macron, polled just 24 percent in the first round of the French presidential election in 2017; Hungary’s Viktor Orbán returned to power in 2010 with a constitutional supermajority thanks to one of the most disproportional electoral systems in the world – with 53 percent of the vote, his Fidesz party took 68 percent of the seats in parliament – and it retained this supermajority in 2014 – this time keeping the same seat share with just 44 percent of the vote.

Trump, of course, lost the popular vote to Clinton in 2016, but more important is the fact that he won the Republican Party nomination with the lowest proportion of the vote since direct primaries became the main means of selecting the GOP’s presidential candidate. The Republican Party of 2016 was unusually fragmented. Trump’s victory, as we’ll see later on, was thus much more contingent than a lot of Monday morning quarterbacking from the political punditry would suggest. Had the support of his opponents cohered behind a single credible rival, it is unlikely that Trump would have become the GOP candidate . Populists in the United States have consistently won or come closest to winning – 1828, 1832, 1896, 1912, 1968, 1976, 1992, 2016 – when the opposition (whether inside or outside a candidate’s party) is fragmented. Populist success thus tends to be greater where the institutional set up allows for direct election to the executive (e.g., presidentialism). In Europe, where proportional representation is the more common form of electoral system, fragmentation works differently. There, when party systems are fragmented, populists often do well enough to gain a share of parliamentary seats, but only rarely have they been able to win sufficient representation to take power themselves or to become the leader of a coalition government. In short, the fragmentation of the group or groups opposing a populist along with the nature of the electoral system will affect the proportion of the vote that needs to be secured, and hence the cost of winning power.

Populism: Past, Present, and Future

This book explores the economics of populism in the West from its roots in Greek and Roman antiquity through to the Trump era. Within the modern West, I focus particularly on the United States of America, France, and Germany, with briefer reference to Britain, Italy, and other cases. This is an admittedly huge volume of political history to hope to cover in a single book. However, even though I won’t have dealt with every case to the satisfaction of specialists, my hope is that the comparative perspective this wide lens allows compensates for its deficiencies. Ultimately, it is only when we look at populism on this longer historical time frame that the value of the supply-side approach becomes apparent. If we focus only on a single point in time, say, post-1990s Western Europe, the factors that affect the big strategic choices between populism, patronage, and programmatic party building are often obscured. In game theory, we’d say that when one strategy is dominant (for instance, programmatic party building), any alternative strategy is going to lead to failure – to the sucker’s payoff. It’s only at moments of disruption, when the returns to these different strategies potentially change, that we are likely to see them in competition. To capture enough of these moments, I’ve cast the net as widely as time and space will allow.²⁷

For the general audience, I hope this wide-ranging historical account, which begins in the next chapter, will be informative and even entertaining. Although much of this history may be familiar, the interpretation I provide should help readers see it in a new light. For scholars, this is a book that can be understood as an exercise in theory development. There will be no rigorous statistical testing of hypotheses, and although I assess the plausibility of alternative explanations at different points in the book, the main goal is to put forward a novel supply–side framework for understanding populism.

Before we go further, one last question needs answering: Why all this worry about a mere political strategy? Populism is not the same as authoritarianism. It still aims to persuade rather than to coerce. As a superficial mass political appeal not catered to developing enduring loyalties, populism, even if temporarily successful, is almost by design a passing fad. In turns out, however, that the very absence of institutionalization inherent to populism makes it an issue of real concern. Populists, as I’ll show in the concluding chapter, face stronger incentives than other types of democratic leaders to erode political and civil liberties. Because they lack a sufficiently large body of organized supporters when they come into power, populists are faced with a great problem in retaining it. They could, as some populists such as Argentina’s Juan Perón have done, consolidate support through the distribution of patronage. If this involves the pillaging of natural resources or the nationalization of industrial sectors best left in private hands, the environmental and economic consequences can be quite detrimental. Alternatively, at least in theory, populist leaders could go on to construct more stable programmatic political parties. However, given the necessary sacrifice of autonomy this implies, it has rarely been pursued by populists themselves. Too commonly, populists complement direct mass appeals with the imposition of restrictions on their opponents instead, sometimes in the process crossing the line into dictatorship.

Yet for all this likely erosion of the liberties we associate with democracy, populists often become, well, popular. Even if they come to power with mere pluralities, even if they clamp down on political opponents, even if they do eventually become dictators, they can and often do become more popular over time. Although some political philosophers find solace in the idea that politics is about values, that if better informed and better socialized, people might not turn to the vague, if sometimes illiberal, promises of populist leaders, economics predicts no such thing. Civic education, although probably intrinsically valuable, is not a panacea. The utopian notion that virtue can be inculcated from above should have died along with Maximilien Robespierre – one of the populists we’ll meet in Chapter 3. Even if everyone, or the vast majority of people at least, agreed on all the policies and principles of running society, the brutal, if usually now bloodless, contest for power would still go on. Sometimes those objectively closest disagree the most – the narcissism of small differences, as Sigmund Freud put it. Accepting the liberal notion that society is comprised of many groups with legitimately different interests, most mainstream theorists and pundits today advocate adherence to the pluralist rules of the game. But this is an aspiration not a solution. We would do well to heed the advice that Gouverneur Morris, one of the American founders, gave to Marquis de Lafayette, then Commandant General of the revolutionary French National Guard: “[M]en do not go into administration as the direct road to Heaven; that they are prompted by ambition or avarice, and therefore that the only way to secure the most virtuous is by making it their interest to act rightly.” Put simply, political actors will agree to the rules when it is in their interest to do so.²⁸

The task facing us is to understand why some political strategies are more effective at some times and in some places than in others. Politicians aren’t robots and parties aren’t algorithms. But just as financial markets adjust quickly to new information – firm valuations and commodity prices moving rapidly in response to economic or political events – the political marketplace also adjusts. As long as populism is cost effective, aspiring political leaders will use it, whatever values or knowledge we might seek to impart to the people. Even good men, as Morris implied, will act badly if we don’t provide the right incentives. If we want to live in a rules-based political system that has the benefits of popular participation, simply wishing it were so will not do. What is needed are policies, programs, and inducements designed to influence the cost-benefit calculus of aspiring political leaders. We need, in other words, to create the economic incentives to make free and fair competition between well-organized bodies of people the only game in town.