«-- Odds 'n Sods: | Main | Letter Re: Battle Rifle Recommendations for a Californian --»

"Official" Statistics on Population, Employment, Income Levels, Money Supply, and Inflation?

James:
In a recent Odds 'n Sods item, you cited a article published by The New York Times: You stated: "A key data point mentioned in the article: 'The median household [in the US] earned $48,201 in 2006, down from $49,244 in 1999, according to the Census Bureau.' "

That's from changing population dynamics and more careful surveys of low-income families. For comparable populations, income has risen as you ought to expect.
Consider the results for "Worked Full Time, Year Round, Both Sexes, White"...
For 1999 income:
Persons in this group: 81.7 million
Mean income of all persons in this group: $44,854
For 2006 income:
Persons in this group: 88 million
Mean income of all persons in this group: $55,176
The 1999 figure, adjusted for US retail price inflation to 2006, is equivalent to $53,781.
Adjusted for US wage inflation, the number is $53,622.

This is only barely better than staying even, but that's a lot better than the conclusion you drew from the New York Times article, which is that the median income has somehow declined 23% in constant dollars. Since when did you start trusting everything you read in the New York Times? In this case, the [New York Times] author went out of his way to make a clearly false claim: "Most American households are still not earning as much annually as they did in 1999, once inflation is taken into account." Based on the actual facts he presents as his source data, that just isn't true.

From the CIA World Factbook, the US GDP was $9.26 trillion in 1999 and $12.98 trillion in 2006, a 40.1% increase. Tom's price-inflation calculator says $9.26 in 1999 is equivalent to $11.10 in 2006, so the real growth was about 17%.

According to the Census Bureau, the population of the United States grew about 7% in those seven years, leaving us with roughly 10% of growth in per-capita GDP. So that's consistent with the other Census data, and it's reasonable to conclude from these analyses that average individual income did in fact increase faster than inflation during this time. - PNG

JWR Replies: Like you, I am dubious about statistics complied by governments. Journalists with an axe to grind--such as the New York Times writer that you mentioned do indeed distort statistics even further, so this is cause to distrust press accounts of "official" statistics.

In many cases, government statisticians are solving equation with multiple missing variables, so their results are an admixture of mathematics, conjecture and voodoo. Inflation statistics are case in point. The official figures on consumer price inflation have become almost laughable. The "core" inflation rate excludes "volatile" food and energy costs. This makes the "official" consumer inflation figure just about useless to me, since my family's three biggest budget items are insurance, groceries, and gasoline.

Money supply figures cannot be trusted. The figure for electronic "bankers" dollars are perhaps fairly trustworthy, ut figures for printed paper dollars are unreliable, at best. There is no way to account for how many dollars are squirreled away in mattresses, or are in the hands of foreigners. (Although if foreigners have half a brain, they are currently scrambling to exchange into a more stable currency.) One key statistic, the M3 Money Supply Aggregate, got so embarrassing that in 2006 the government stopped publishing it. At least one web publisher, ShadowStats, has attempted to reconstruct the M3 figure, independently. (They charge for access to most of their data and reports.)

Government unemployment figures are also highly suspect. By their own admission, the Bureau of Labor Statistics undercounts the chronically unemployed. Once someone has been unemployed long enough to have their state unemployment insurance benefits run out, they simply drop off the radar. The unemployment statistics also do a poor job of accounting for underemployment. For example, they would in the aggregate count an out-of-work stockbroker (that formerly made $250,000 per year) as "full time employed" if he out of desperation takes a full time job as a waiter, for minimum wage, plus tips.

Census figures cannot be completely trusted. The US Census has become a political football. Most notably, it has become a cause celebre for both homeless advocates and illegal alien advocates. These advocates can be found both inside and outside of government. They have attempted to manipulate data for political ends. How many illegal aliens are there in the US? Nobody really knows. The estimates that I've read range from 10 million to 22 million. But again, it is guesswork.

The bottom line is that "official" statistics are not be trusted. I'll close with an unattributed quotation: "Most people use statistics the way a drunk uses a lamp post, more for support than enlightenment."

Permalink