David Brooks omits context on economy

David Brooks is pulling out all the conservative pundit tricks in this passage from his column today:

If you’ve paid attention to the presidential campaign, you’ve heard the neopopulist story line. C.E.O.’s are seeing their incomes skyrocket while the middle class gets squeezed. The tides of globalization work against average Americans while most of the benefits go to the top 1 percent.

This story is not entirely wrong, but it is incredibly simple-minded. To believe it, you have to suppress a whole string of complicating facts.

The first complicating fact is that after a lag, average wages are rising sharply. Real average wages rose by 2 percent in 2006, the second fastest rise in 30 years.

The second complicating fact is that according to the Congressional Budget Office, earnings for the poorest fifth of Americans are also on the increase. As Ron Haskins of the Brookings Institution noted recently in The Washington Post, between 1991 and 2005, “the bottom fifth increased its earnings by 80 percent, compared with around 50 percent for the highest-income group and around 20 percent for each of the other three groups.”

Brooks later claims that "the Democratic campaign rhetoric is taking on a life of its own, and drifting further away from reality... [C]andidates now compete to tell dark, angry and conspiratorial stories about the economy."

Another way to frame this might be to say that pundits compete to tell optimistic, context-free fables about the state of the economy.

The first trick that Brooks uses is to present an out-of-context statistic that sounds promising about the state of the economy under Bush, the same technique used by, for instance, the Wall Street Journal editorial board, NRO's Jerry Bowyer, and the administration (here and here).

He writes that "after a lag, average wages are rising sharply. Real average wages rose by 2 percent in 2006, the second fastest rise in 30 years." But the "lag" Brooks briefly mentions means that workers haven't gained much during the current expansion -- here's what real average wages look like since 1993, for instance:

In general, the current recovery does not compare well to the post-WWII average (or the 1990s), as the Center on Budget and Policy Priorities points out:

Brooks's second claim is also misleading. In it, he uses the gains made under President Clinton to make the status of the poor look better than it is, writing that "according to the Congressional Budget Office, earnings for the poorest fifth of Americans are also on the increase." He then cites statistics for the 1991-2005 period.

However, if you actually look up the CBO report in question (PDF), you'll find this graphic, which illustrates that income hasn't really been "on the increase" for the lowest quintile of households with children under Bush:

The earnings gains that Brooks highlights have tailed off particularly dramatically under Bush:

Of course, it is true that the economic situation is complicated. But Brooks is giving short shrift to concerns about an economy that is currently doing very little for most Americans.

Update 7/31 8:56 AM: Brooks ran this clarification today:

Last week I cited data on rising earnings among the working poor. I should have made it clear that the data referred to poor households with children, since poor households without children did not enjoy those gains.

No word, however, on the larger issues with his column.

Comments

[Rob]

Why do you cite statistics for the lowest quintile of households with children rather than statistics for the lowest quintile of households?

[Brendan Nyhan]

Those are the same statistics Brooks uses -- see the Haskins op-ed. I've clarified this slightly above.

[Porkopolis]

Hmmm...An alternate "glass is half-full" look at things:

1. A look at the historical data leads one to conclude that all households should be striving to get out of the lowest, second, middle and fourth quintile and into the highest quintile.

Capitalism and limited/minimal government provides the highest probability for someone that is so inclined to achieve that goal. In other words, Income Mobility is alive and well.

2. Figure 8 (page 8) of the cited report notes that female headed households with children had lower income when compared with all households during the same period between 2001 to 2003.

Conclusion: Don't have children outside of a marriage or stay married.

3. An argument can be made that illegal immigration has hurt the lowest quintiles the most because lower-skilled jobs compensation rates stay depressed with an ever increasing supply of cheap illegal labor.

[Barry]

"In other words, Income Mobility is alive and well."

In Europe. In the USA, not so much.

[Patrick Meighan]

Particularly in a column purporting to rebut the notion that CEOs are seeing their salaries skyrocket while middle-class workers are getting squeezed, isn't it misleading to cite a raise in *average* wage as purported evidence to the contrary? Wouldn't a look at the *median* wage be much more illustrative?

To wit, in a hypothetical scenario where Worker A, Worker B and Bill Gates comprise the wage pool, and Worker A and Worker B go without a single penny's worth of wage increase while Bill Gates rakes in a $1B raise, average wages will have gone up $333M dollars, while median wages will have stood utterly stagnant. The former stat, average wage (which is what Brooks uses), would be what you'd use if you wanted to try to depict this inequitable scenario as some sort of robust economy that is benefitting all. The latter stat, median wage (which is what Brooks doesn't use), would be what you'd use if you wanted to accurately depict what the scenario clearly describes: C.E.O.’s seeing their incomes skyrocket while the middle class gets squeezed.

Patrick Meighan
Venice, CA

[Alfred E. Neuman]

US Census Bureau income statistics give a good picture of what has been going on in the recent years. In table A3 (page 40), you can see how the shares of household income have changed. For example, the lowest quintile share of income was 4.2% in year 1980 (before Reagan), 3.6% in 2000 (before GWB) and 3.4% in 2005. The highest quintile share was 44.1% in 1980, 49.8% in 2000 and 50.4% in 2005.

The ratio of 95% lower limit to the median income has gone up all the time. For instance , in 1980 the ratio (calulated from the table A-3) was 2.68, in 2000 3.46, in 2004 3.54, and in 2005 3.58.

Well, maybe the US Census Bureau has a liberal bias, like just like the media :-)

[Alfred E. Neuman]

A typo: the ratio in 1980 should be 2.86 instead of 2.68.

[Kandis]

Superb job, thank you for your research on this.

[David Nieporent]

You don't really refute Brooks' main point; you just try to spin it differently. Putting aside that "hourly wage" is far less relevant than total earnings, it's true that the growth in these wages hasn't been great. But it has been growth. This is real growth, not nominal growth. So it's a positive thing, not a negative one.

Why is that "doing very little"?

It's also misleading to talk about quintiles over time, since they aren't the same people. You can't ignore immigration and then talk about the performance of the lowest quintile.