« Life Without Sovereignty | Main | Sarkozy »

Fallacies of Composition

24 Sep 2007 10:38 am

One of the odder things about recent American politics and political journalism is that a lot of people seem to have convinced themselves that working class Americans typically vote Republican, and the Democratic base is just composed of screenwriters in the Village and their kids. Paul Krugman notes the reality:

In fact, if you look at voting behavior, low-income whites in the South are not very different from low-income whites in the rest of the country. You can see this both in Larry Bartels’s “What’s the matter with What’s the Matter With Kansas?” (pdf), Figure 3, and in a comprehensive study of red state-blue state differences by Gelman et al (pdf). It’s relatively high-income Southern whites who are very, very Republican. Can I get away with saying that rich white trash are the problem? Probably not.

What this reflects, in turn, is the odd fact that income levels seem to matter much more for voting in the South. Contrary to what you may have read, the old-fashioned notion that rich people vote Republican, while poorer people vote Democratic, is as true as ever – in fact, more true than it was a generation ago. But in rich states like New Jersey or Connecticut, the relationship is weak; even the very well off tend to be only slightly more Republican than working-class voters. In the poorer South, however, the relationship is very strong indeed.

Now what is true that things look different if you use education rather than income as your measure of class, so the idea of the upscale liberal wasn't just concocted out of thin air. And there are some compelling reasons to look at education rather than income. However, as I point out in this article on the subject, insofar as people want "working class" to mean "no college degree" that the median income for non-college whites is higher than the national median income.

Share This

Comments (23)

insofar as people want "working class" to mean "no college degree"

My sense is that we don't have a consensus of how "working class" is to be used. I seem to recall some DLC paper noting that Americans with a HHI of 65K thought of themselves as struggling to get by, and my suspicion is that such people would think of themselves as either working class or allied with the working class.

Quite a straw man you start your post off with. Here are the latest stats on party affiliation by income group, from the Pew Center as of the first quarter of '07. Democrats still dominate the lower income classes in party ID, with Republicans having the edge between $75k and $150k. But Democrats regain the edge above $150k.

Fred, 75k a year is NOT working class. There is no way to to really believe it is unless you REALLY don't want to understand what working class means.

You can't compare the 75k a year office worker to the guy who works three part time jobs and makes 20k a year. It's just not a comparable situation in any way, shape, or form.

If you think of 75k a year as 'low income' you're too out of touch to be adding much to this conversation.

The reason why Democrats far poorly amongst so many poor voters is that, for a generation, - despite their lip service to the contrary - the Democrats have failed to deliver on actual benefits to the working class - particularly white working class males.

The reason for this is that with the collapse of Keynesian economics, the fundamental underpinnings for big government solutions to economic problems have become unsound. Stimulate demand and the resulting supply is apt to come from Tokyo or, nowadays, Shanghai.

While this decline began with the oil shortages of the 1970s and the Japanese automobile invasion, the collapse of Hillary care in 1994 marked the tipping point in this. It became obvious that Democrats were not serious about delivering benefits and just wanted some bureaucratic program to provide perks for female lawyers.

During the 1980-2000 time period, through supply-side, the Republicans seemed to be able to respond in a more meaningful manner.

The rub, for the Republicans, has been that, when you stimulate supply, there is also no guarantee that the benefits will be located in the USA,either. This is the time of outsourcing and of downsizing, all subsidized and encouraged by tax cuts.

But the failure in supply side does not therefore herald a revival of Keynes. The same old problem with stimulating demand persists.

There is no simple, easy, or satisfactory answer to this conundrum - at least none that I know of. But we should realize that this means that the North American continent no longer equals "the economy." Washington DC is less and less likely to be a meaningful place from which solutions can be obtained. However, as both Bush and the Democratic Congress have demonstrated, Washington can still manufacture much mischief.

Given what Matt is suggesting, is it true, then, that education and income are more loosely linked in the South than other regions of the country? In other words, are relatively wealthy Southerners less well educated than their counterparts elsewhere? Does anyone know? Or care?

Soullite,

Read my initial comment again, slowly if necessary, and then feel free to respond to something I actually wrote, OK? Because you are taking issue with things I didn't write.

Hah! Fred says it's a "straw man" that "a lot of people seem to have convinced themselves that working class Americans typically vote Republican." But then Duncan Kinder comes along and says: "Democrats far[e] poorly amongst ... many poor voters."

Straw man or not? We report, you decide.

David,

Duncan Kinder can say what he likes, but the Pew Center data shows that at, below $75k in annual income, party affiliation for Democrats increases as incomes decreases. Those results would seem intuitively obvious to most people.

'Middle class' vs 'working class' is the operative distinction. The genius (in a neutral or ambivelent sense) of the American economic culture is that the ambiguities of class are played up: class is a state of mind. Everyone has traditionally - esp. post-war - wanted to be middle class and most think of themselves as such, whether they make 30K or 150k (or 1.5 million for that matter). But median income no longer buys our mental image of 'middle class' - which is apparently based on basic consumer items and a house. Make some consumer products cheaper, and have people work two or three jobs, and the illusion is maintained for a while. But middle class means more than how much stuff you can buy. It's also about optimism; the assumed upward trajectory over time, your kids doing better than you, maybe getting sorta rich eventually, etc. Like the store called 'One Hour Cleaners' which nonetheless takes a week to do your cleaning, 'middle' is just the name of the class. It's not 'middle' in a real sense. The 75k ˜ 150-200k people have lately been the 'imaginary' - but politically real - middle class. Now their illusion, aforementioned, is threatened. I guess that shortly you will have to make more than 200k to be 'middle class' in the traditional sense. I don't see how the GOP doesn't lose quite a few of those voters in the next cycle - unless inducment of a kind of National PTSD works again to any appreciable degree. I'd also question how deep party affiliation runs, as it were.

[Recall Bush's selling of his tax cuts as benefiting 'average' income. You have to hand it to those assholes: they really know how to lie cannily].

There is no simple, easy, or satisfactory answer to this conundrum - at least none that I know of. But we should realize that this means that the North American continent no longer equals "the economy." Washington DC is less and less likely to be a meaningful place from which solutions can be obtained. However, as both Bush and the Democratic Congress have demonstrated, Washington can still manufacture much mischief.
Posted by Duncan Kinder | September 24, 2007 11:20 AM


Duncan,

I think there is a solution and its one that can be found when you compare the USA to the domestic policies in the rest of the OECD.
The solution in a world of open economies with relentless creative destruction and labour displacement is for the state to do two things:

1- provide as much social programs as possible to redistribute from winners to losers in the economic furnace and allow as many opportunities as possible for workers to re-educate, re-train and re-locate. Smooth out the transations and reduce inherent life risk. The key policies would be adult education, universal healthcare and government pension plans not tied to employment earnings. Good public education for children is wrapped into this as well.

2- promote new business formation through a combination of low corporate tax rates and decent supply of educated workers. Note that subsidy, tarrif and sector rebates are to be avoided since they increase the need to maintain either higher general corporate tax rates, higher personal tax rates or deficit financing.

Some call this the nordic model, other talk about ireland, it's basically a high welfare, low corporate tax, high personal tax regime. Note that the personal taxes are concentrated on consumption (sales taxes) rather than income taxes. It has proven very successful and to my mind makes a good liberal agenda for the USA.

Currently the US has a low welfare, high corporate tax, low personal taxes plus deficits regime. And NB when I day 'high tax' I mean Clinton level on the income tax side, reductions in the coroporate tax should be met with increases in sales tax. I have no idea how politically feasable it is to introduce a national sales tax in the USA. If it is just not possible you are still better off with lower corporate and higher personal taxes. Where the personal tax limit is I don't know. Intuition says 45% marginal top rate, but it would have to be modeled.

"Democrats far[e] poorly amongst ... many poor voters."

Everybody fares poorly among poor voters, because poor voters don't vote very much at all, relatively speaking.

I agree with Fred; Soullite's rejoinder makes no sense. Fred observes, per the Pew data (follow his link and look at the chart), that Democrats win all income ranges up to $75,000. This *includes* the very low income ranges that can reasonably be called working-class. [And the party gap, per the Pew data, is most pronounced at the lowest income levels.] I don't think Fred was suggesting that someone making $50-$74K is working class. So what's your point, soullite?

Contrary to what you may have read, the old-fashioned notion that rich people vote Republican, while poorer people vote Democratic, is as true as ever – in fact, more true than it was a generation ago. But in rich states like New Jersey or Connecticut, the relationship is weak; even the very well off tend to be only slightly more Republican than working-class voters. In the poorer South, however, the relationship is very strong indeed.

Very useful stuff from Krugman here. I never spent enough time pondering the issue, but, he lays it out logically. I had always wondered about what seemed to me a contradiction: rich states increasingly vote Democratic, and poor ones GOP. Now it makes sense.

Also, Krugman's facts here seem to imply that the trend observable in blue states is intensifying; ie., the positive correlation between wealth and GOPness there is weaker than it was, say, 30 years ago. What I wonder is: could it ever cross over, so that in fact in Connecticut or New Jersey you'll someday be more likely to vote Democratic the wealthier you are?

1) It is somewhat false to use income ranges in the same manner across the country. Someone making $50K a year is a lot better off in some parts of the rural South than he would be in Manhattan.

2) The thing about the rural Southern countryside is that each town /county tends to be dominated by a local overlord or small elite. In large corporations or large cities -- and in areas with diversified economies -- no one really cares what your political or personal beliefs or sexual practices are --you are just one of millions.

3) In a small town, by contrast, surveillance and social discipline are much stronger. Cross the wrong family and you will have to leave --because you won't find employment anywhere. Seriously annoy them and You will also lose a lot on the sale of your house.

3) This overlord class is adept at the use of deceit and hypocrisy to justify its self-interest. Newt Gingrich, for example, could rant for hours about the tyranny of the federal government --while saying nothing about the far more powerful tyranny imposed by concentrated wealth. The Southern overlord class argued for "states rights" -- in order to preserve human slavery. Southern legislatures have long concentrated power in the state capital in order to serve the rich. They long argued that the Bill of Rights only constrained the Federal Congress --not them.

Northern Virginia outside Washington DC, for example, is a horrible traffic mess because, Even today, the political machine set up by the Byrds in Richmond Virginia protect rich real estate developers from any measures that would improve the life of residents of Prince William and Fairfax counties.

The overlord class controls what is acceptable
political opinion in the South. Acceptable themes in country music is controlled by a few wealthy men in Nashville-- hence , it's mindless acceptance of aggressive imperialism and the betrayal of our soldiers. It lacks artistic creativity because it is corporate propaganda -- just look at what happened to the Dixie Chicks.

The decentralized nature of the Baptist Church --and of similar Protestant churches -- leaves the local preachers at the mercy of donations from the local overlords -- which explains why local preachers always rant at the harmless sins of the poor and ignore the social injustice and sins of the rich.

In the South, religion truly is the opiate of the masses. Consider that the Southern Baptist Convention is defined by its historical tradition of supporting human slavery-- contrary to the words of Jesus. Consider also how much the Southern Baptists did to undermine justified resentment and rebellion within black congregations.

Needless to say, newspaper publishers in those small towns and cities know what to print. And local school boards know what to teach. And local workers know better than to try to create a union for protection.

Which may explain why an entire generation of Southern men sacrificed their lives -- and plunged their families into deep poverty -- in order to defend an aristocracy which screwed them every chance it got. Why the common citizens died to defend the institution of slavery --even though 3 out of 4 Confederate soldiers was too poor to own even one slave.

Northern Observer:

The European model certainly has much to instruct us, but - if what I am saying is correct - we are going to witness problems with them also.

Hints of what I mean include such items as the recent NYT article on how Belgium is being fractured by the Fleming / Walloon issue. Scotland and Wales are devolving, Catalonia and the Basques are issues for Spain, and the Muslim immigrants are an issue not just because of the current terrorist hysteria but also for much the same reason that Mexican immigrants are an issue here. Many things are bubbling beneath the surface in Europe.

This is not to dismiss what you are saying - I am not shooting you down by saying "Oh, but there's a problem in Europe." But I am saying that the nation-state is growing less and less functional. And while a continental-wide European state would be less problematic than the current various European states - a state nonetheless it would remain.

I really do think we are entering an era where some new form of social and economic organization is going to emerge that will supplant the state.

What is Krugman talking about? He misrepresents Gelman et al's findings when he says that lower income Southern whites don't vote much differently than lower income Northern whites. Here's Gelman:

"The Mississippi electorate is more Republican than that of Connecticut; so much so that the richest segment of Connecticutians is only barely more likely to vote Republican than the poorest Mississippians."

Northern Observer brings up some interesting points about the usefulness of looking at what sort of policies have facilitated prosperity elsewhere. I would expand the search beyond prosperous European countries such as Ireland to prosperous countries in other parts of the world, including Asian tigers like Singapore and Hong Kong, as well as Australia.

I would add that it's also useful to learn from some of the mistakes that have been made elsewhere. In a number of European countries where the government provides the equivalent of our country's corporate retirement plans (albeit on a strictly defined benefit basis), they have fiscal crises much worse than our problems with Social Security. In general, though, I think N.O. is right about the merits of lower corporate income taxes balanced by somewhat higher income taxes, if necessary. This sort of model has worked well for a number of first world countries, including Australia.

Also, the Nordic countries were initially able to afford a generous welfare state because their populations are small, highly educable, disciplined, and homogeneous. In an multi-ethnic democracy, where different ethnic groups have broad differences in educational and economic achievements, generous welfare benefits can be a source of disincentives, resentment, and inter-ethnic discord. An unregulated, open-borders immigration policy can aggravate this further, by attracting less economically-productive groups of immigrants, who consume more in government resources than they pay in taxes; this threatens the longer-term economic viability of a welfare state.

As increasingly higher taxes are required to support a growing class of net recipients of government largess, the most productive potential immigrants are discouraged from immigrating, and some of the most productive citizens feel compelled to emigrate. This is why, for example, there are so many more high-achieving Frenchmen working in London and New York than vice-versa.

Somehow I just don't believe Bartel/Krguman/Pew in this. The fact of the matter is that we know that at every level the people in the Southern states are voting more and more for Republicans (at least over the period 1968 to 2004). Blacks in those states are voting for Democrats, and some of the states with the largest %Blacks are voting the most Republican overall.

Now I am sure that the wealthy whites are voting very highly Republican, more so than wealthy whites up north.

But still, there are not enough wealthy whites to overcome middle class and working class (lower-middle and upper-lower) whites if the latter were really voting Democratic, and in the same percents as their class/income/education equivalents up north. They just are not doing so. Bartel's data, or analysis of that data is wrong.

Re: 75k a year is NOT working class.

Some skilled trades go that high, and higher. Beware of conflating "working class" with "working poor". Education may have something to do with it, ditto income, but I would define "working class" by the sort of work done: If you feel the need to shower after work (not before, or not just before) then you are working class.

Re: The reason why Democrats fare poorly amongst so many poor voters ...

Um, the Democrats do not fare poorly among the poor. To the extent that the poor voite, they do tend to vote Democratic. By ther way, I see no discussion so far of union membership in this thread. Union membership is one of the best predictors of Democratic voting patterns.

Re: During the 1980-2000 time period, through supply-side, the Republicans seemed to be able to respond in a more meaningful manner.

Huh? What supply side policies did the GOP enact in the 90s? With Bill Clinton as president and his 93 tax increase set in stone?

Re: I have no idea how politically feasable it is to introduce a national sales tax in the USA.

How about a VAT, which is a kind of sales tax, but one that is almost impossible to cheat on, and has a proven track record of raising revenue with a minimum of economic disruption.

Re:Hints of what I mean include such items as the recent NYT article on how Belgium is being fractured by the Fleming / Walloon issue.

OK, but how is that an economic problem? Ethnic strife is as old as the Roman Empire in Europe.

OK, but how is that an economic problem? Ethnic strife is as old as the Roman Empire in Europe.

As I recall, the article stated that the Flemings are more prosperous than the Walloons, and this disparity contributes to the strife.

But that is not my point. My point is that social unrest and economic prosperity are not hermetically sealed topics but rather can influence one another. And to the extent that a state's economy depends upon the state itself, I would presume that factors that would tend to undermine that state would also tend to undermine its economy.

"Also, Krugman's facts here seem to imply that the trend observable in blue states is intensifying; ie., the positive correlation between wealth and GOPness there is weaker than it was, say, 30 years ago. What I wonder is: could it ever cross over, so that in fact in Connecticut or New Jersey you'll someday be more likely to vote Democratic the wealthier you are?"

Among elites (especially those that are defined as elite by their education and/or involved in international/multinational firms or policy) in the Northeast, whether Democratic or Republican, cosmopolitanism is an important quality to have. It seems to me that West coast and possibly Midwestern elites are rather similar in this regard. Knowing basic things about international cuisine, wine, cinema, etc. are important. Nobody wants to be the only person who doesn't know at least one decent non-European or non-American wine or doesn't eat sushi.

Nobody is really proud to admit they have no real understanding of a foreign language. This is a generalization, of course, but it holds a lot of truth. As the Republicans move closer towards being the party primarily of Southern white Christianist values and shed away the last vestiges of libertarianism into becoming only the hate gays, Muslims and Mexicans party, such Republican elites will become outcasts in their own party. They already live near immigrant communities and historically gay communities (the Village and Spanish Harlem aren't too far from the Upper East Side in the grand scheme of things). They are probably more likely to accept the possibility of a child being gay or being in an interracial relationship than a Trent Lott Republican. The Republican Party's Southern power elite will say more and more that people in favor of open borders and knowing languages besides English are un-American. They will keep on saying that their gay relatives are inherently evil. Being told that you, your family and your lifestyle are un-American has a way of souring one on the Republicans. This is what Jesse Helms speaking out against the Northern "literary class" was about. The Republican war machine will continue to be bad for business for those not working for Boeing or Lockheed Martin as our deficit balloons. Tax cuts eventually will probably just become impossible without thoroughly destroying the American government's ability to do anything, which will also be bad for business, assuming that Republicans are somehow able to hold onto power. I'm not sure Rockfeller Republicans will be able to have a home in the Republican Party as the Republican Party becomes a de facto Dixiecrat Party if current trends hold.

The big problem with Matt's argument is the assumption that since "working-class conservatism" and "latte liberalism" aren't universal, and social issues don't account for the behavior of all white working-class voters, these phenomena don't exist or don't matter.

Bartels cites data that "whites in the bottom third of the income distribution cast 51% of their votes for Democrats, as compared with 44% of middle-income whites and 37% of upper-income whites." Fair enough. But Thomas Frank, who argued for a strong Democratic populism, would ask: Why only 51% of the lower-income folks? Can't the Dems do better with them? (Rather than trying to increase the 44% or 37% numbers by softening their populism, that is.) Toward that end, Frank rightly advocates softening Democrats' social liberalism while emphasizing the need for things like progressive taxation.

I add more here.

It should also be noted that Thomas Frank defended himself against Larry Bartels' charges (which Krugman cited) here. That shouldn't be left out of the discussion.

(Hat tip to Wikipedia)


Comments closed October 08, 2007.

Copyright © 2007 by The Atlantic Monthly Group. All rights reserved.