September 23, 2003

Income Inequality

So there have been a few posts of note about the growth of income inequality. Paul Krugman recently decried it in the NYT. Jane Galt of Asymmetrical Information and Dan Drezner have weighed in.

There are four effects that I've heard about recently that I think are worth mentioning. I don't know the extent to which these effects account for the growth in income inequality, but I've heard compelling arguments that they have a significant effect:

  1. Immigration — A constant flow of uneducated new arrivals, with no wealth and weak social networks, hold the lowest income deciles down while the higher deciles rise. Any true look at income inequality would have to control for immigration or somehow take into consideration time spent in the country.
  2. Feminism — Working women lead to larger income gaps at the household level. There are two effects here. First, men with higher incomes are more likely (through education, social networks, etc.) to be married to women who can also command high incomes. Thus, any existing gap is likely to be widened. Second, since lower income families are more likely to have had two earners already, the increased participation of women in the workplace is less likely to impact family income in the lowest deciles.
  3. Social Mobility — America is somewhat unique in the amount of mobility between income deciles over time. This spreadsheet at the Economic Policy Institute shows that fully 23.5% of people in the bottom two income quintiles in 1969 were in the top two quintiles in 1994. Another 20% had moved to the middle quintile. This spreadsheet would be better if it offered other years for comparison (to assure us that these aren't stand out years) and if it included a "1969 Not Yet In Country" bucket to help address point #1 above. But this brings up a question: to what extent is greater income inequality still a problem if it's different people in the bottom and top?
  4. Standard of Living — The standard of living of even the lowest income deciles have improved in ways that the CPI (and hence real wages) doesn't capture. Improvements in the quality of consumer goods (appliances, automobiles, ready-to-eat meals, clothing, etc.), new technology products (CDs, computers, cell phones, etc.) and healthcare and workplace safety have raised all income groups in ways that are not captured in standard metrics.
Obviously, everyone worries about income inequality to some extent — the differences between reasonable people seem to be about how much inequality is warranted assuming everyone is (at least slightly) better off than they were before? Conversely, how much better off do the lowest income groups need to be to justify a given level of inequality?

Regardless of where you end up on the philosophical spectrum (egalitarian to Walzerian to Rawlsian to sufficientarian) these effects should inform the discussion of how much is too much. If anyone has more information on attempts to quantify these effects, I'd love to hear it.

UPDATE: I've added some charts below.

Here's a chart I found at Brad DeLong's site. Though only obliquely relevant to this discussion, it is interesting:

Second, here's one that speaks to Mike comment below on global issues:

Posted by richard at September 23, 2003 11:35 PM

I can't find the link unfortunately, but within the last week or so I read a blog about global income inequality. It said that 70% of the variability of income can be accounted for by the country that you live in, and only 30% has to do with inequality within countries. Due to advances by China and India, the overall global income inequality has been shrinking dramatically over the last 25 years. However, Africa and the other poor countries haven't improved at all in that period.

Your original post was aimed at American inequality, presumably, but it's an even bigger global issue.

Posted by: Michael Weiksner at September 24, 2003 01:54 PM

Interesting thoughts. One critique: your discussion somewhat confuses the notion that income inequality exists with the bigger problem -- that it is worsening. There are inoffensive and non-controversial reasons why income inequality exists -- people make more money as they get older, people retire and leave the labor force, small businessmen take write-offs in some years, farmers have bad years due to weather. The thing that Krugman and others are worried about is why is income inequality getting worse?

Your immigration and social mobility arguments aren't really relevant, unless you're arguing that immigration and social mobility are increasing, which I don't think is the case (Krugman has a pretty good critique of the social mobility argument nestled in this longer piece - Likewise, the rate at which women are entering in the workforce, I expect, is decreasing, so you'd expect whatever contribution it makes to inequality to be diminishing as well.

The standard of living argument is interesting. I agree that consumer products are better today than 30 years ago. But is improvement in consumer products adequately compensating for the decline in affordability of things like health insurance, childcare, college education and decent housing? Isn't it true that more families have two parents working not because of "feminism" per se, but because an increasing number feel they need two incomes just to keep their heads above water? What's the impact on standard of living of the loss of the intangibles a stay-at-home parent provides a family?

More questions than definitive answers, but my sense is that these are the concerns that the income inequality crowd is worried about, and I think they deserve some attention.

Posted by: Scott Reents at September 24, 2003 02:59 PM


I'm mostly talking about the income gap "worsening" (i.e. growing larger) too. Sorry if that wasn't clear. I suppose there are different was you could measure "growing larger" though.

The one that I was thinking of that is most relevant to my points: how much more income the highest income quintile has than the lowest quintile. For instance, if in 1990 the top quintile earned 10x more than the lowest quintile while in 2000 they earned 12x more, I would say that the income gap is growing (these are just made up numbers, of course).

With this metric, I think immigration could play a huge role. For instance, assume that a new immigrant today (with little education and no personal wealth) is able to make about the same in real wages as he/she would have in 1990. If, on the other hand, everyone else was able to make 1.1x real wages in 2003 versus 1990, the constant supply of low income immigrants would hold down the lowest quintile and would make the spread (between low and high quintiles) grow. This despite the fact that everyone is better off — certainly the immigrant who chooses to come here and the rest who are all 1.1x richer.

And to your point about feminism, yes, you're right, all of the movement of women to the workplace is not driven by feminism. The point is, though, that some of it is — and more at the high end than the low. If at the household level the highest quintle goes from earning 10x the lowest to earning 11x the lowest, but it's because 10% of women in the highest income households decide that they'll feel more fulfilled with a job, does that really say something about our system getting more unequal? Or is it just turning into monetary terms (through her new salary) an inequality that already existed (i.e. the free time of the rich housewife)?

As far as social mobility goes, you're also right, it's less relevant as to whether the gap is really growing or not. However, it does speak to the effects that the gap might have. If the gap is growing but social mobility remains high and people still move frequently from low income groups to high, then any presumed effects of the income gap would, I would think, have to be mitigated somewhat. As would any social justice questions.

I believe that whether or not you should care about income inequality growing larger depends on whether it is really growing larger and whether it has reached a point where the negative effects (to social cohesion, etc.) outweigh the positive effects (of rising tide lifting all boats, etc.).

Posted by: richard at September 24, 2003 06:11 PM


You're right. China and India have made great strides and brought a huge number of the world's poor up with them (a point that people often miss when they look at income statistics by country instead of by indiviuals). But there are significant parts of the world which are not progressing — which is one of the reasons it's imperative that we KickAAS.

Seriously, getting rid of agricultural export subsidies and textile tariffs would do a lot to help the developing world. Certainly more than giving a few billion more dollars in foreign aid to corrupt African dictators.

I fear that the collapse of the Cancun talks and the recent scares about offshoring/outsourcing of service jobs will lead to a protectionist backlash will rule out these productive measures.

Posted by: richard at September 24, 2003 06:32 PM

Agriculture subsidies are an interesting matter in the US. If I remember correctly, farmers represent only 3% of the US population.

Why do they enjoy such subsidies and political clout? historic? strategic importance of food? bloc of single issue voters? powerful lobby?

I don't have a good answer.

Posted by: Michael Weiksner at September 25, 2003 07:52 AM

Thanks for the response, Richard. I realize you posted this ages ago, but it's been kicking around in my mind all these weeks, so I figure I should at least but it down in writing.

Is income inequality worsening? There are so many ways to answer that with statistics, but the table below presents one of the more compelling, IMO:

Distribution of Income Gains, 1947-89

Percentile/years % Annual Increase

1947-73 2.6%
1973-79 0.4
1979-89 -0.3
1947-73 2.7
1973-79 0.4
1979-89 0.3
1947-73 2.8
1973-79 0.7
1979-89 0.6
1947-73 2.7
1973-79 0.6
1979-89 1.1
1947-73 2.5
1973-79 1.1
1979-89 1.6

again, from:

Two things are apparent. First, income growth fell for all income segments after the first period, 47-73. Second, beginning in 73, income growth stagnated for the bottom 40%, whereas it continued at a slower but still appreciable rate for the top. Thus, while all quintiles shared equally, more or less, in the rapid growth from 47-73, that growth became highly concentrated in the upper quintile between 73-89 (unfortunately, I don't have comparable numbers for the 90s).

So not only does it appear that income inequality has worsened since 1973, but that that worsening broke from the trend of the previous 25 years, during which time inequality stayed roughly the same. Immigration may explain some of that, though the immigration that occured from 47-73 was apparently not enough to worsen inequality (immigration was increasing markedly throughout the period).

I find these statistics troubling for two reasons. First, something changed around 73 that has made income inequality worsen in the years that followed. Even if you believe that some level of income inequality is necessary or good, the conservative in you has to be somewhat worried about a trend we don't really understand the consequences of and that, on the face of it, indicates progress toward a less equitable society. Second, the growing inequality, from far "lifting all boats" has left a segment of the population with income growth that is close to nil.

Posted by: Scott Reents at October 20, 2003 05:47 PM

I can't understand why a person will take a year to write a novel when he can easily buy one for a few dollars.

Posted by: Parmalee China at February 27, 2004 04:08 PM