The Evidence from Iowa (so far)

The one question that has consumed Dems since this whole race began a billion years ago is this: Which of these pretenders can oust Trump from the White House? In survey after survey, Dems across the nation have made that amply clear. This is usually used as a stick by centrists to beat the left with. But that stick is based on the assumption that something approximating the median voter theorem holds. Fortunately or unfortunately, things are not so simple. Just ask yourself where Trump stood with respect to the GOP’s center in 2016. Instead of one-dimensional left-right positioning, what we need is a diagnosis of the present predicament. You can’t fix it unless you know what’s broken. In order to defeat Trump, you need to know why he won. And, once you have a confident diagnosis, ie once we have a firmer handle of American sociopolitical reality, then you can proceed to ask yourself: Which of these pretenders can defeat him?

The Policy Tensor has offered a diagnosis of the catastrophe of 2016. The proximate cause of 2016, given the geography of the electoral college and so on, is the breakdown of elite-mass relations in the United States. The past twenty years have seen a progressively intensifying confrontation between the college-educated professional class that predominates in the affluent zone, and the high school-educated working class that predominates in what the former contemptuously call “flyover country.” Confidence in national institutions — a proxy for elite-mass relations — is at an all time low. What we have is a clash of working class and elite ideologies.

In the grip of Boasian antiracism, American elites look down upon the working class as bigoted and racist. The contempt is mutual. The white working class does its own boundary-making work — one that places what are perceived as status-seeking, corrupt, self-serving, and self-congratulatory elites, beyond the pale. Whence the culture wars, the rise of Fox News and right-wing radio, and, of course, Trump. This is why the Democratic party has become a coalition of college-educated whites and racialized minorities, while Trump’s GOP has emerged as the party of the masses. Moreover, it is precisely the grip of Boasian antiracism on the minds of the professional class that made the now debunked racial resentment hypothesis so compelling to the professoriate and the scribes. The real problem is neither the racism of working class whites, nor just the vanishing of broad-based growth, but the contempt with which the elites hold the masses. Elite-mass relations cannot be restored without both a revival of broad-based growth and elites checking their class privilege. The general implication of this diagnosis is that liberals need to shed their discourse of self-congratulation fast; what is needed is earnest class diplomacy. The particular implication for 2020 is that we need a candidate that is both a champion of working class interests and seen as such — not an antiracist warrior.

In what follows, we look at the pattern of support for the four front-runners. We begin, however, with the Trump swing that we’ve looked at the county level before. Here we look at Iowa counties only — the working class revolt that put him in the White House was across the board and not confined to any particular region. In Iowa, Trump won more votes than Romney in counties that have a lower college graduation rate, lower population, higher overdose death rate, and lower per capita income.


Let’s turn to the pattern of the Iowa caucuses held yesterday. Table 1 shows the correlations between the candidate’s share reported by FiveThirtyEight this evening, with 62 percent of the precincts reporting, and what we found to be the strongest predictors of the Trump swing.

Table 1. Spearman’s correlations (Iowa counties, N=99).
College Graduation Rate Population Income Net Migration Rate Overdose Death Rate Trump swing
Sanders 0.17 0.36 -0.30 0.03 0.32 -0.29
Biden -0.27 -0.14 0.08 -0.25 -0.18 0.36
Buttigieg -0.13 -0.09 0.01 -0.07 0.02 0.24
Warren 0.35 0.25 -0.02 0.34 0.15 -0.38
Estimates in bold are significant at the 5 percent level. 

Warren has perhaps the greatest policy chops of the four. But she is also clearly the candidate of the elite. The correlations between Warren’s performance and college graduation rate (r=0.35, p<0.001), population (r=0.25, p=0.014), net migration rate (r=0.34, p<0.001) are high and significant. Not coincidentally, she does best where Trump did worst (r=-0.38, p<0.001). This is due to the geography of class in Iowa. As a Warren supporter, I feel sad that the most compelling candidate, from a straight policy perspective, has failed to make inroads in the all-important working class. But that’s what this data shows.

Sanders’s performance is also anti-correlated with Trump’s. And he is even stronger in densely populated counties, which tend to be younger. But the Sanders coalition is different from Warren’s. He did particularly well in poorer counties (r=-0.30, p<0.001). He does even better in counties with higher overdose death rates (r=0.30, p<0.001). And he is the only one to do so. Moreover, his performance is not significantly correlated with college graduation rate — meaning that he is less of a candidate of the elite than Warren.

Buttigieg is difficult to pin down. The only significant correlation we find for him is with the Trump swing (r=0.24, p=0.015). Even though they are insignificant, the signs of the correlation coefficients suggest that Buttigieg is closer to Biden than either Sanders or Warren.

Biden has the strongest correlation with the Trump swing (r=0.36, p<0.001). More importantly, he has the strongest anti-correlation with the college graduation rate (r=-0.27, p=0.007). This suggests that Biden enjoys strong support in working class communities. Moreover, the anti-correlation with the net migration rate suggests that he does well in places that are literally being left behind.

So the results suggest that either Sanders or Biden may be our guy. But let’s dig in further. In what follows, we look at the relative support these four candidates received in Iowa counties. We sort the counties by each of our conditioners and arrange them in quintiles. The pattern can then be interpreted as their relative strengths in particular blocs.

We begin with the most important conditioner, class, proxied here by the college graduation rate. The results are congruent with those we deduced from the correlations above. Warren does best among the elite; Biden does best in the working class. The gradients for Sanders and Buttigieg are weaker than those for Warren and Biden respectively.



Next, we look at per capita income, a noisier proxy of class. The pattern here is also congruent with the correlations. Sanders is the only one whose support rises as the per capita income of the county falls.


Biden does better when we look at net migration rate; Warren does the worst. Recall that communities that are bleeding people swung hard towards Trump. If you want to win them back, Biden may be your guy.


But wait, Sanders does much better than Biden if we pay attention to overdose death rates — the strongest predictor of the Trump swing beyond class proxies. Warren displays the same pattern as Sanders, although it is not as strong. Biden scores lower, and Buttigieg is all over the map as usual. If Trump is in the White House because large parts of the country are in serious trouble, perhaps Sanders is the man to oust him.


Buttigieg does well across the population density spectrum; Biden’s all over the place. Meanwhile, both Warren and Sanders do much better in already blue, densely populated counties.


The progressives fare much worse relative to the centrists in counties where Trump did well. This too is congruent with the correlations. Both Biden and Buttigieg do well, but the former’s gradient is more pronounced.

Trump.pngThe results should be taken with a pinch of salt. But the picture that emerges suggests that Warren has so far failed to make inroads into the working class; that Sanders does well in troubled communities; and that Biden enjoys strong support among working class communities. Pete is an astroturf political entrepreneur.

So, who should we bet on to oust Trump? If the pattern evident in Iowa holds, Biden and Sanders may both be viable against Trump. Biden is viable because he is working class and working class folk can tell that he is one of their own just by the way he talks — recall that class is passed on at your parents dinner table. As I suspected, the Biden tendency is the shadow of the class war on the Democratic primary. Sanders is viable because he does well in communities that are struggling. If you think that Trump is in the White House because large parts of the country are in trouble, and he has done little to help them, Sanders is your man. If you think that only a man who can out-blue collar Trump can oust Trump, Biden’s your guy. If progressives want Sanders instead of Biden because the former can be expected to demolish the neoliberal political economy, they must begin by losing the Boasian scold.

Ultimately, the governing question is whether culture or economics is more important to the meaning-making of the working class. For at issue in what Lind calls the New Class War, is not just the vertical and spatial polarization of value-added, income and wealth, but the concentration of symbolic production and the cultural desertification of vast swaths of America. Intellectuals have for too long paid attention to the former at the cost of the latter. It is time to pay attention to the historical sociology of the white working class — the dominant strata of American society. And to bring geography back to the center of political analysis, where it belongs.

2 thoughts on “The Evidence from Iowa (so far)

  1. The New Class War
    Saving Democracy from the Managerial Elite
    By Michael Lind

    Yes, and also saving democracy from the central bank that provides free money for pseudo-capitalists aka the “managerial elite” to buy back their own shares with zero interest money (before Reagan stock buy-backs were illegal, imagine that. It is primarily that the system is rigged, not that growth is uneven, that accounts for inequality.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s