An alternative hypothesis

by Russ Roberts on September 4, 2006

in Standard of Living

Sebastian Mallaby sums up conventional thinking in today’s Labor Day column:

By now almost everyone agrees that inequality is serious. Economic
growth no longer seems to help the majority of workers; the proceeds
flow to the top fifth or so of the workforce, and the top within the
top has done especially handsomely. But the tough debate is what to do
about this trend. The surprising answer is: tax reform.

If I have time later this week, I’ll discuss whether tax reform is going to solve this problem. But here I want to look at two interesting qualifiers in that paragraph. The first is in the first sentence, the word "almost." Mallaby realizes that there are a few skeptics left out here in the hinterland who haven’t quite accepted this new kind of economic growth that somehow leaves most of us in the dust.

But the more important word is in the second sentence, the word "seems." Why did Mallaby insert that word? Why didn’t he confidently write, "Economic
growth no longer helps the majority of workers?"

Perhaps he is uneasy about making the absurd assertion that there is something new about the extraordinary growth in recent years in America without an explanation of how this has come about. For it is absurd, to presume that somehow, without our noticing, the economic system has
for the first time in American history become a system that grows like topsy without
benefiting the average worker. But it’s more than absurd—it’s dangerous.  You’re in uncharted terroritory. How can you possibly posit a solution to the problem without understanding it?

But there is another way to interpret the word "seems." It doesn’t just express uncertainty. It has an alternative meaning. It is an admission of the possibility that the claim of a stagnant standard of living for the average American is simply an illusion.

What if the average worker is actually benefiting from economic growth? What if the measures of inflation we use are not capturing quality changes (which they surely do not). What if because those measures overstate actual inflation, living standards are actually rising and the average worker is sharing in economic growth, despite statistical claims to the contrary.

What if in fact, all the other data, the data on life expectancy, on the size of our houses, on our purchases of unparalleled luxuries—our iPods and big-screen TVs—and enrollment in college at or near an all-time high—what if these data are more reliable than the official data on wages?

What if things are actually getting better for the average worker, but it merely seems otherwise?

Be Sociable, Share!



38 comments    Share Share    Print    Email


Bruce Hall September 4, 2006 at 11:32 am
Isaac Crawford September 4, 2006 at 11:34 am

Playing the "seems" game is dangerous, your last sentence is especially telling. If it "seems" that the average worker is not doing so well, then that's all that really matters isn't it? If they don't feel as good, then no amount of numbers will change that.

I personally don't trust any of the macro stats, there are too many problems and too many assumptions to make them useful. It reminds me quite a bit of baseball. We need a Bill James type of character to come up with new stats that actually tell us something about actual performance. Till then, everything just "seems" to look the way we want it to…


drtaxsacto September 4, 2006 at 11:53 am

I agree with Isaac. The Macro stats are unreliable. One surprising stat from last week suggested that by focusing on five counties in the US (Manhattan, Kings County (WA), San Francisco, Santa Clara, San Mateo – fundamentally the Silicon Valley and Microsoft and the money center) all of the supposed income distributional issues disappear.

We are going through a major transition in employment for many workers where they have substantially greater roles in determining their futures. For those of us who believe in individual liberty – that is a very good thing. But in order to understand how all those changes affect individual workers we need to think about some new ways to collect the data.

rjh September 4, 2006 at 12:00 pm

I suspect that "Seems" is because it is not mathematically correct, but an expectable emotional reaction. If you trace back to the original BLS statistics, what you find is that over the past 40 years the average improvement rate in total compensation is 1.55% CAGR. For the bottom 80% of the population, it is 0.5% CAGR, and the rate holds at around 0.5% CAGR for all the sub-percentiles. It starts climbing at the 80% percentile. For the top 10% it is 1.5% CAGR and for the top 1% it is 3.5% CAGR.

The long term effect of that difference is that it seems that the top 1% is getting all the improvement. This is a perceptual effect, not a correct mathematical statement. The correct mathematical statement is that everyone is benefitting, but the benefit is conspicuously biased toward the top 1%.

It is also unclear to what extent this reflects a transient from the Internet bubble. Work at the Inequality Project indicates that a significant portion (not all) of the bias is due to the real estate and options gains of the participants. This work is bolstered by some of the recent BLS reports that the top 10% and 1% percentile benefits levels have dropped in the past couple years. That would be consistent with them including transient effects.

As for the "spiteful rich" my experience is that this is more a "nouveaux riche" behavior. The newly successful are triumphantly showing the world that they have succeeded. This is emotionally the same as the successful child jumping up and down saying "look at me". It is not motivated by spite.

Tim September 4, 2006 at 2:19 pm

I have a big concern with statistics showing the growing gap between rich and poor that most people seem unable to address: Is it real?

Imagine our country has 10 people in it. The rich (top 2) get richer, and the poor (bottom 2) see their wealth grow even faster. We may see a convergence in wealth.

But people in Mexico and elsewhere see our poor having this type of opportunity and enter into the United States. They go from making $10/day to $5/hour. Clearly they are better off. All 10 original residents are better off now, too as their cost of living decreased.

But even though the poor of our hypothetical world are getting richer much faster than the rich, our statistics may not show it. If the poorest 2 original citizens made $6/hr, and the new poor people make only $5/hr, then we can say something like, "The bottom quintile of Americans saw their pay [i]decline[/i] from $6/hr to $5/hr, while the rich saw their pay [i]increase[/i]." The insinuation is, of course, bunk. It's a matter of whom we chose to measure in which scenario, not a matter of who actually got richer.

Incidentally, I brought up to my professor that I had the same issue with the implications people take from the Solow growth model (that limiting immigration causes the per capita income to rise, thus making us better off).

John S. September 4, 2006 at 3:25 pm

Let's not waste time on what Times reporters believe, or profess to believe. As soon as there's a Democrat in the White House, the situation will right itself, and those statistics that "seemed" to point to one thing will magically begin pointing toward something else.

Michael Bindner September 4, 2006 at 3:59 pm

You would be interested in another article in Sunday's Washington Post by a fellow of the AEI, who talked about what was poor in 1973 and what is poor now are two different things that the statistics don't capture well.

I will agree that it is not about only the numbers. However, the fact is that quite a few people are only making ends meet because of things like subsidized mortgage loans, no principal mortgages and the balooning of credit card debt. As the Federal Reserve tightens the screws, what the "numbers show" may become painfully obvious to even those trying to justify the current regime. This regime is about to be summarily tossed out, at least on the congressional side and in spite of gerrymandering. It is almost the perfect storm with the economy starting to head south at the same time opinion on the war is as well. Gay marriage politics will not save GOP bacon this year. Many people have put their economic plight on hold for reasons of cultural and defense politics. This won't happen forever.

kebko September 4, 2006 at 4:25 pm

To me, there seems to be an irony in public opinion as a product of human nature. People's satisfaction is really based on a comparison of their situation to their expecations. So, it seems that often, lower satisfaction is the result of higher expectations.

It seems to me that frequently the loudest complaints are a product of unprecedented hope & potential.

Patrick R. Sullivan September 4, 2006 at 4:49 pm

Michael Bindner is talking about Nick Eberstadt whose WaPost article is a much shortened version of this paper

In which he points out that the official poverty statistics move the opposite direction of others that should be complementary. Such as unemployment or educational attainment. It's a question of who ya gonna believe, the poverty statistics or your own lying eyes?

Patrick R. Sullivan September 4, 2006 at 4:54 pm

For some reason the link didn't work to the Eberstadt paper. The url is:

Lisa Casanova September 4, 2006 at 5:40 pm

As economic growth raises standards of living, our expectations for how we should live continue to rise. So even if we get better off, aren't people always stuck feeling like they're not getting any of the benefits of economic growth?

triticale September 4, 2006 at 7:20 pm

I am not made better off by iPods and big-screen TVs. I own neither. I haven't even bothered to replace the AM-FM radio in my car with the MP3-capable CD player I bought.

I am, however, unquestionably made better off by cellular phones. Without them there would be no employment for cellular RF optimization technicians. There's no job security; I work when a provider does an upgrade project, but my annual earnings are around what I made in that hot factory I hated, and I love what I do. Other factors, including the Gingrich-era "tax cuts for the rich" mean that the periods of unemployment between projects amount to leisure time. In short, macro statistics do not even hint at my situation.

ben September 4, 2006 at 8:18 pm


"The long term effect of that difference is that it seems that the top 1% is getting all the improvement. This is a perceptual effect, not a correct mathematical statement. The correct mathematical statement is that everyone is benefitting, but the benefit is conspicuously biased toward the top 1%."

There is still the problem that the top 1% is populated by different people each year, meaning inequality is overstated by such measures.

The debate on inequality is plagued by misuse of statistics.

guez September 4, 2006 at 8:40 pm

I'm not going to argument the technical economic point that some people are making here. Like an earlier poster, however, I find the iPod and Big-Screen TV index disturbing. Conservatives and Liberals (and Communists and Fascists, for that matter) can all agree that the "good life" is not measured by consumer conveniences (or any convenient economic index, for that matteR). People don't merely want toys: they want to work and build lives for themselves and their children. Because of job insecurity (at least perceived job insecurity), expensive health care, etc., many people are finding it harding to reach this goal, and we shoudn't be dismissive.

JohnDewey September 4, 2006 at 11:55 pm

Ben: "There is still the problem that the top 1% is populated by different people each year, meaning inequality is overstated by such measures."

That's a very good point, Ben. Small business owners sell out and realize a one time capital gains. Lawyers receive non-recurring awards from a product liability verdicts. Movie producers hit the jackpot with blockbusters.

It's also true that the bottom 10% are populated by different people over time. A few may remain poor for years. But most move up into the middle income class.

Morgan September 5, 2006 at 12:44 am

"Because of job insecurity (at least perceived job insecurity)…"

Good qualification, guez. The percentage of people who qualify for unemployment insurance filing initial claims for unemployment each week (let's call it the Job Insecurity Index) shows a distinct downward trend since 1971, the first year for which I have data on number of covered lives.

It was near its April, 2000 low early this year, and remains near that low point – about 2.4 of every 1,000 covered people file each week.

Index averages:

Jan 1971-Dec 1979: 5.35 per 1,000 covered lives
Jan 1980-Dec 1989: 4.51
Jan 1990-Dec 1999: 3.25
Jan 2000-Jul 2006: 2.83

The peak was 8.7 per 1,000 covered lives in January of 1975. By this measure, at least, job insecurity has clearly been decreasing for a long time.

Data source here:

Tino September 5, 2006 at 9:29 am

What nonsense.

Look, here is US census data for household income 1970-2003 (remember households have been getting smaller, so this underestimates):

Poorest fifth


Second poorest fifth:


Middle fifth


Second to richest fifth


Richest fifth


Everyone is getting better off, OK?

Maybe this positive development stopped recently after Evil Republicans took over congress, or the “new” economy, or whatever the economically illiterate left is ranting about this week? Wrong. Let’s see what happened just under the last 10 years: 1993-2003

Poorest fifth


Second poorest fifth:


Middle fifth


Second to richest fifth


Richest fifth


I think a part of the problem is looking at the “hourly wage” data that is very biased. According to them the US has not even had growth in AVERAGE wages the last 30 years!!! Since the capital-labor income division has been constant, all the growth must have disappeared into a black hole.

Either that or the data they use is junk. Not even including the increase in benefits accounts for all of it.

Cyberike September 5, 2006 at 1:45 pm

Forget statistics, lets talk about something I know a lot about: me. I make a ton of money, but I have to work my butt off to do so. I have a lot of stuff, but my sense of security is almost zero. I have not been able to take a vacation in 2 years. Do I feel good about the economy? No.

My situation has nothing to do with the political party in power, but some people want to make this a referendum on the republicans in power. Fine. I am a republican, but my feeling is that when you campaign on fear for so many years, it creates a general feeling of unease in the population. I personally am sick and tired of it.

I am much better off than I was a few years ago, and much better off than the general public. But I don't feel good. I live with pressure and stress. I look around and see others in worse shape than I am, and instead of making me feel better it worries me.

The way it is now, "stuff" is not going to make me feel better. I will feel better when I can stop and enjoy it. Factor that into your equations.

panasianbiz September 6, 2006 at 4:52 pm

Of course, we would all feel better to find that life for the average worker is actually better than previously believed! However, I think that the data can be interpreted in various ways, and it may be the agenda of the person utilizing the data that informs the results rather than the numbers themselves.

Previous post:

Next post: