Of course the unemployment rate is rigged. I know because the agency that calculates it, the Bureau of Labor Statistics, tells me so.
I’ll explain below. First, some background:
Since the Bureau of Labor Statistics announced the official figure for September’s unemployment rate, analysts and commentators have argued about the number’s validity. Some point out that the unemployment rate seems to have taken a suspiciously large downturn just 32 days before a presidential election in which the most hotly contested issue is the incumbent’s performance on creating jobs. Others dismiss suspicions of impropriety as mere “conspiracy theory” and say that the bureaucrats who gather the data and calculate the unemployment rate are career professionals who would never let their biases affect their work product.
Among the skeptics: Jack Welch, former CEO of General Electric. Welch noted in a Wall Street Journal op-ed that the BLS’s calculations include all manner of subjective elements – that, indeed,
The possibility of subjectivity creeping into the process is so pervasive that the BLS’s own “Handbook of Methods” has a full page explaining the limitations of its data, including how non-sampling errors get made, from “misinterpretation of the questions” to “errors made in the estimations of missing data.”
Bottom line: To suggest that the input to the BLS data-collection system is precise and bias-free is—well, let’s just say, overstated.
Even if the BLS’s work were entirely subjective, Welch noted, people should be skeptical based on the fact that the labor-force participation rate, the growth in government workers, and overall job growth all set multi-decade records achieved over the past two months, a period during which the business community noticed no significant economic upturn.
Peter Ferrara, a former Reagan aide and one of the top experts on government statistics, noted in The American Spectator that political commentators (Rush Limbaugh, in particular) long ago predicted the unemployment rate would fall below 8% in October 2012, if only because a rate below 8% would be considered a necessity for the President’s reelection.
Ferrara pointed out that the number of full-time jobs in the BLS report actually fell by 216,000.
So how did the reported unemployment rate fall to 7.8%, down from 8.1% in August, and 8.3% in July? That is because the BLS (might consider dropping the L) also reported that the separate Household Survey, on which the unemployment rate is based, found a giant 873,000 jump in jobs last month, the biggest one month increase in nearly 30 years! That makes sense because the same survey found a decline in jobs of 119,000 in August and 195,000 in July.
Of course, 582,000 of those supposed jobs were part time “for economic reasons.” The BLS defines that as “individuals [who] were working part time because their hours had been cut back or because they were unable to find a full-time job.” That is not all good either. But a part-time job is better than no job at all.
That’s the key to the growth in job numbers for September: the huge number of people (582,000) who gave up looking for full-time jobs and settled for part-time jobs.
Ultimately, the foundation for October’s relatively low unemployment rate (low compared to the previous month, anyway) is that, back during the Clinton administration, the government altered the way that unemployment is calculated. As explained in August by New York Post financial columnist John Crudele:
Back in ’94, the definition of a discouraged worker was changed. Until then, [the Department of] Labor would call people’s houses and ask the adults if they were employed or not. If someone said they weren’t even looking for a job because they were too discouraged, that pre-1994 person was considered unemployed and included in the figures.
The Clinton administration decided that unemployed people couldn’t be discouraged — and not job-hunting — for more than 12 months. If a person hadn’t searched for a year he was simply not included in the U-6 or other measures of joblessness.
And that’s why we know the numbers are rigged – because any calculation is rigged if it is based on criteria that are inherently subjective. Are you “unemployed” if you’re an accountant and you took a job as a greeter at Wal-Mart? How about if you spent one night of the month babysitting for a neighbor? There is no correct answer to such a question. For each question based on such a hypothetical, BLS bureaucrats simply pick an answer, and that answer becomes the correct one by virtue of having been picked.
No conspiracy is necessary to rig the numbers. That the U3 unemployment number is not purely scientific, and not objective as most people understand the term, is clear from the fact that the BLS produces alternative unemployment figures — U1, U2, U4, U5, and U6, all of which are as valid as U3 and all of which are different from the U3. By producing the alternate calculations, BLS is admitting that there is no one, true unemployment number – that, when the government picks one over the other, it is rigging the result.
The official rate, at this writing, is 7.8%. But U6, the broadest government category, puts the unemployment rate at 14.9%. Meanwhile, the SGS-Alternate unemployment rate, produced by private analysts using government data, puts the number at 22.8%, roughly two-thirds of Depression levels. (The SGS-Alternate rate, an alternative to the official government rate, is calculated using pre-Clinton criteria for “discouraged workers” that include those who have been “discouraged” more than a year.)
Don’t get me wrong. Most of the people who work on government statistics really are trained professionals doing the best job they can. Only rarely (e.g., on global warming) do they outright lie. The folks at the Bureau of Labor Statistics try their best to get a real unemployment number. When they poll people, they determine respondents’ employment status indirectly rather than by asking them directly — a method that’s more likely to get a true result, but that is still fraught with problems. To ensure that their polling represents the general population, the BLS pollsters try to weight their results based on age, sex, so-called race, “Hispanic ethnicity,” and other factors, but in doing so they must rely on Census numbers that are themselves partially faked. (For example, “Hispanic ethnicity,” a category based on the language supposedly spoken by one’s ancestors, is no more real than, say, a hypothetical “Anglo ethnicity” that would include, in one group, people with ancestors from English-speaking countries as diverse as Botswana, Ireland, India, and Australia.)
For technical reasons, it’s impossible to determine with certainty a lot of numbers on which policymakers depend and over which the media obsess. These include the poverty rate, the number of people in various racial classifications, the number of species of living things (and how many of those species are endangered), and the earth’s average surface temperature circa 1880, which you must know if you’re going to determine whether global warming theory is correct. The fact that such numbers cannot be determined scientifically never gets in the way of government officials who pretend that they can do these calculations with a level of accuracy down to one part in a thousand or even one in ten thousand.
Often, researchers and analysts and commentators have no choice but to rely on government statistics, flawed as the figures often are. Often, those are the only numbers available, and, often, they’re better than nothing. But anyone relying on government statistics should keep in mind that, usually, they’re no better than SWAG.
SWAG? That’s what scientists call a scientific wild [expletive] guess.