Skip to main content
What Happens When Democrats are “Fighting Wall Street With One Hand, Unions With the Other,” While Republicans are Fighting Unions With Two Hands?

Matthew Yglesias noticed something interesting in a political story today that reminds me of one of our arguments in Red State, Blue State. Yglesias quotes a Washington Post article on Blanche Lincoln returning to the U.S. Senate after surviving a primary challenge from a candidate supported by organized labor:

Lincoln was embraced by her colleagues . . . Sen. Chuck Schumer (N.Y.) held up two fists and said of her primary campaign: “Fighting Wall Street with one hand, unions with the other.”

Yglesias points out a fundamental asymmetry here:

Schumer, who’s become something of a national leader among Senate Democrats, celebrates this ideal [of governing in a manner that’s equidistant from rival interest groups], but there’s not a single member of the Republican Party–much less a leader–who’d say anything remotely similar. Schumer is basically describing polyarchy or interest-group pluralism. But the imbalance between Schumer and his rivals on the other side of the aisle reflects what Charles Lindblom ended up criticizing as the “privileged position of business in polyarchy.”

Before getting to my argument about geography and voting, let me set aside the specifics of the interest groups here. For one thing, “Wall Street” is not the same as “business,” and, for that matter, I imagine that many Republican senators would be happy to describe themselves as fighting Wall Street. I’m sure that a mirror-image conservative version of Yglesias could give a long list of policy dimensions in which Democrats are more consistent with Republicans being more likely to compromise.

But these details don’t really affect Yglesias’s point, which is that there are some asymmetries between the two parties in the Senate: it’s certainly not simply pro-business Republicans vs. anti-business Democrats. And you can see this in a variety of issues, including, for example, that there’s been lots of talk about cutting spending to rein in the deficit but not much about raising tax rates in the upper brackets. (Again, I’m not arguing the policies here—that’s not my area of expertise (to say the least)—I’m just elaborating on Yglesias’s point and attempting to detach it from specific, debatable points about Wall Street or whatever.)

I think Yglesias is on to something which relates to the different support of Democrats and Republicans in different states.

In short, Democrats win in Democratic states with the support of all the voters, while Republicans win in Republican states with the disproportionate support of upper-income voters. A Democratic senator such as Chuck Schumer of New York gets the votes of the rich and the poor in roughly equal proportions. (I haven’t looked at Schumer’s votes in particular, but based on general patterns, I suspect he does a bit better among the poor than the rich, but not so much.) In contrast, leading Republican senators from conservative states do not need to concern themselves so much with lower-income voters.

Again, this is more of a statement about relative than absolute positions, and it’s a statement about economic more than social issues. And other things are going on too, notably that there are a lot more Democrats than Republicans in the Senate right now.

Here’s the key graph:

And here’s the full story (from chapter 9 of Red State, Blue State):

As a thought experiment, let’s imagine some alternative scenarios in which voting patterns are different, and see how the Democrats and Republicans might react, we sketch the scenarios in figure 9.8. In each case, the Republicans win the poor states and the Democrats win the rich states, but the scenarios differ in the pattern of income and voting within states.

– In Statesworld, voting varies by state income but not at all by individual income within state. In this world, states that vote more Republican do so uniformly at all income levels.

– In Reverseworld, rich people everywhere favor the Democrats. This is the scenario that people may assume after seeing the red-state, blue-state map: a world in which the Democrats win the rich vote, and the Republicans the poor vote, in every state.

– In Votesworld, rich people tend to vote Republican, with income predicting vote choice in the same way across all states. This is how we had imagined the nation to be, after checking out the patterns of income and voting at the state and national level, but before analyzing income patterns within states.

– Finally, in the Real World, richer people vote Republican in every state, but the relation between income and vote choice is much stronger in poor states than in rich states.

How would politics play out in each of these worlds? In Statesworld, we would expect politicians to have strong geographic loyalties: Republican officeholders would fight for Mississippians of all income groups and Democrats would lobby for the richer states. There would be no particular reason for the parties to disagree on rich–poor issues, except to the extent that these intersected with states. For example, Republicans might favor higher federal transfers to the poor (who are more likely to live in low-income states), and Democrats might favor reducing the top rate of income tax (which disproportionally hits the residents of rich states).

In Reverseworld, the connection between party and income is even clearer, with Republicans dominating among low-income voters in Mississippi and other poor states, and Democrats relying on richer voters in rich states. In this world, it would make sense for the yuppie Democrats to strongly resist the graduated income tax (which takes money from high-income people in rich states) and for the middle-American Republicans to support higher taxes and income transfers.

How would this work? Politicians are motivated to give benefits to their voters. Part of this is simply the overlap between voters, political activists, and politicians themselves. Beyond this, candidates generally get into Congress from primary elections, at which point they need to appeal to the voters of their party in their state or congressional district. We have already discussed in chapter 8 how the two parties in Congress are ideologically divided; here we are relating this to divisions between high and low-income voters.

Continuing with our hypotheticals, neither the Statesworld or Reverseworld stories are plausible; it is the Democrats who favor higher transfers to the poor and Republicans who prefer lower tax rates. This makes sense in Votesworld, a scenario in which the Democrats consistently get the votes of poorer voters and Republicans win richer voters throughout the country. In Votesworld, the Republicans rely on rich votes in Mississippi and the Democrats, symmetrically, dominate among the poor in Connecticut.

Votesworld is not far from the truth, but, in the Real World, there is a much steeper income–voting divide in poor states than in rich states. What might this imply about politics? In Mississippi, the Republicans win with a coalition of middle- and upper-income voters, and as such we would expect them to support policies that directly benefit these groups. On the other hand, in richer states such as Connecticut, the Democrats are the leading party, but with the support of all income groups. Thus, unlike in Votesworld, it would not make sense for the Democrats in Connecticut to simply dismiss the concerns of high-income voters.

Add this up at the national level and there is a net benefit for the rich and middle class: they are represented in the winning coalitions in both rich and poor states, whereas lower-income voters are only supporting the winning party in the rich states. This oversimplifies—a graph of all fifty states looks much messier than the picture of Mississippi, Ohio, and Connecticut—but it captures an essential asymmetry between the patterns of support for the Democrats and the

P.S. David Weakliem adds the following alternative explanation:

Sometimes Republicans win predominantly Democratic states, or Democrats win predominantly Republican ones. The Republicans from Democratic states would get support pretty evenly from across income levels, so you’d expect them to try to balance their interests, and that seems to be true. But when a Democrat manages to be elected from a Republican state, most of their votes would be from low-income people. So you might expect Democrats from Republican states (like Blanche Lincoln) to be all-out supporters of the lower class. That doesn’t really seem to be the case.

Here’s a somewhat different way the state differences in income effects may be relevant to representation. If higher income voters are more sensitive to the candidates’ ideology (or candidates believe they are) then candidates who became too closely identified with the interests of lower-income groups would lose support from the top, without gaining as much from the bottom. In effect, the people in higher income groups would be the swing voters, so candidates would pay more attention.

Of course, this is going beyond the data here–I [Weakliem] am just proposing it as a possibility.