Erik Engstrom and Robert Huckfeldt have written a significant book that helps answer some big questions. Why did Trumpian populism resonate so well with a substantial minority of Americans? Why did the Democrats’ New Deal coalition fall apart? Why has US policy responded so feebly to rising economic inequality? And that classic query, Why is there no socialism in America?

The core of their book and its chief contributions involve meticulous, data-rich analyses of roll-call voting in the US House and Senate from the late New Deal to the early 2000s. Race, Class, and Social Welfare pays particular attention to the crucial years (roughly 1954 to 1965) in which civil rights concerns, fueled by Blacks’ “Great Migration” northward, burst onto the two-party policy agenda. Race-related issues broke up the old social-welfare-based unidimensional politics. They drove Southern whites to leave the Democratic Party and become Republicans, depriving Democrats in the House and Senate of some of their strongest social welfare support from the impoverished South. The results were a partisan realignment and a transformation in the meaning of “liberalism” that left many working-class Americans without clear political champions and left us vulnerable to the snake-oil salesmanship of Trumpian populism.

The data-analysis chapters are quite compelling. They first highlight the pivotal moment of the 1957 Civil Rights Act, with Lyndon Johnson’s canny logroll, which temporarily held the Democratic coalition together but diluted new rights protections. The authors then use ADA scores (1948–2008) to chart the demise of the Democrats’ New Deal coalition in the Senate. There was a sharp drop between 1948 and 1964 or so in the liberalism of Southern senators, caused both by conservative Republicans replacing economically liberal Democrats and by the remaining Southern Democrats (most notably Alabama’s liberal icons John Sparkman and Lister Hill) becoming much more conservative. The clever analysis on p. 95 shows that Senators’ voting records (ADA scores) were substantially shaped by the interaction of two factors: the percent of Black population within their state and the national drift of African Americans toward the Democratic Party between the late 1940s and the late 1960s. As Blacks became more Democratic, Senators from heavily Black states became more conservative.

A parallel chapter on House W-nominate scores from 1940 through 1979 produces similar findings. Between pre- and post-civil rights periods, the Democrats lost seats in the South (mostly during and after the 1980s, however; for an engaging account of how Democratic House candidates hampered this realignment, see James Glaser’s Race, Campaign Politics and the Realignment in the South, 1996). At the same time, Southern Democratic representatives moved from being the most liberal to the most conservative members of their caucus. Again, the ideological realignment occurred within as well as between the parties.

In the next chapter (chapter 7), we get a look at the mass-level manifestation of these processes: the way in which the heart of the national Democratic coalition among ordinary voters outside the South shifted away from Franklin Roosevelt’s skilled and semiskilled workers. By 2004 the Democrats had been transformed from a party dominated by the white working class to a diverse coalition of white professionals, African Americans, less affluent whites, and other nonwhite racial and ethnic groups. (The upward-income nature of this shift—particularly important for the money side of the Democratic Party—can be seen more sharply when professionals are separated from managers or when the focus is on postgraduate education levels. Affluent doctors, lawyers, consultants, and the like have tended to become Democrats.)

As Engstrom and Huckfeldt emphasize, the declining importance of the white working class within the Democratic Party coalition is a consequence of both the changing voting patterns of white workers and the demographic decline of the white working class. This, together with the decline of organized representation of workers through trade unions, is profoundly important for understanding the continued erosion of social welfare policies and the weak US response to economic inequality.

Beyond these specific findings, the book is also important for embedding its data analyses in a broad theoretical and historical narrative about the general roles in American politics of race, class, and various sorts of populism. Here the crucial distinction is between “progressive” (left-) and “regressive” (right-) populism. The authors correctly point out that progressive (prosocial welfare) populism has deep roots in nineteenth-century US history. (They do not, however, have much to say about the Farmers’ Alliance, the People’s Party, or the way in which Democrat William Jennings Bryan hijacked the movement; for that, readers should consult Lawrence Goodwyn, Elizabeth Sanders, and others.)

The authors also note the deep historical roots—especially in the South—of reactionary, right-wing populism that has divided the American working class by fanning not only social resentments, especially racial resentments, but also divisions by ethnicity (e.g., against the Irish and then early twentieth-century Southern and Eastern European immigrants). Chapter 8 brings this up to date with Trump’s divisive 2016 rhetoric against Mexican immigrants, Muslims, and others.

One can quibble with some aspects of the multivariate analyses of Trump’s 2016 support using CCES data. Although noting the opposing linear impacts of income (+) and education (-), Engstrom and Huckfeldt seem to miss the curvilinear impact of each: Trump’s strongest support came from the high-school-educated lower-middle class, not from the affluent or the poor. No explicit measures of trade tensions, immigration attitudes, or racial resentments were included. Still, the key findings are convincing. Both economic anxiety and (apparently conflicting) desires to repeal the Affordable Care Act independently contributed to individuals’ Trump support when controlling for partisanship (a major motivator of ACA repeal) and other factors. And white residents of states with substantial minority populations were (again with substantial controls) particularly prone to support Trump. These findings rather neatly fit with the general analysis of regressive populism.

We would offer a caveat, however. Trump’s campaign rhetoric certainly was aimed at mobilizing support based on racism, xenophobia, and sexism. But it did not in fact represent a pure form of regressive, right-wing populism. As a candidate, Trump actually advocated important elements of progressive left-wing populism: a huge, job-producing infrastructure program, reversing the job-killing effects of trade deals and rebuilding US manufacturing, and not touching Social Security or Medicare. Trump did promise to abolish Obamacare, but only with a “better” and cheaper substitute.

These positions—however fanciful or disingenuous they have proven in the four years since—clearly distinguished Trump’s campaign rhetoric from the anti-social welfare, pro-austerity policies of virtually all his Republican rivals, helping him win the GOP nomination. And they defused the Democrats’ usual general-election advantage on social welfare issues, particularly because Hillary Clinton showed little fervor for them and was weighted down by free trade enthusiasm and Wall Street ties. Only after taking office did Trump almost completely abandon left-wing populism (except for trade and immigration) and delegate most of his domestic policy agenda to orthodox congressional Republicans. In analyzing “why Trump?” it is important to remember that his high-water level of support—from nearly half of voting Americans—came in November 2016, when he could plausibly be seen as mixing elements of both left- and right-wing populism. After that his approval rating never rose above the mid-40 percent.

Engstrom and Huckfeldt’s analysis of Trumpian populism—and of regressive right-wing populism more generally—could also profit from a little more attention to the role of money in US politics. The authors are well aware of the growing evidence that “affluent” Americans wield much more influence over policy making than middle-income citizens do. (They might have said a bit more about the collateral evidence that much of that influence may come from the “truly wealthy”: multimillionaires and billionaires.) But they do not integrate this evidence into the role of money in 2016 and other elections. This includes the likely impact of financial contributions on the Democrats: pushing them in a liberal direction on social issues, sometimes more liberal than the general public; but pushing in moderately conservative directions (less liberal than the public) on some economic issues. Nor do they examine the extent to which money from the wealthy has underwritten both Republican orthodoxy and (in a few key cases) Trumpian populism.

Another important matter also calls for more research. To what extent, and in what ways, have wealthy Americans or other elites—as the authors imply—created or magnified the racism and other social resentments that Trump mobilized into electoral support? To what extent have the economic anxieties of American workers that the authors rightly attribute largely to structural factors (international low-wage labor competition, job-killing automation) been deliberately channeled into social resentments by conservative elites?

This is a hard question. Much past research has touched on it in various contexts, from Southern race-baiting, to Progressives’ disparagement of immigrants, to Nixon’s and Reagan’s anti-Black dog whistles. But more could be done to bring the story up to date in our period of rising economic inequality, increased racial and ethnic diversity, and liberalization of racial attitudes (especially among the young)—in which some Republicans seem desperate to maintain support from lower-income Americans whose economic interests they regularly undermine. And more could be done to analyze the generation of social resentments, as opposed to their electoral mobilization.

Only clever research designs and nuanced historical analyses are likely to provide clear results. But such causal analyses, as well as ideas about how to prevent or counteract demagoguery, will be important as we recover from Trumpian misrule and continue to battle American racism.