Friday, January 06, 2012

A breaking point for employment? Just a second

The other day, the Economist had an article that suggested that productivity gains in the United States had reached a limit and businesses were going to have to add workers if they expected to keep going.  Sure enough, the BLS announced today that the private sector added a more-than-expected 200,000 jobs and the unemployment rate dipped slightly to 8.5%.

But an inordinate number of those jobs were couriers and messengers (even adjusting for seasonal effects) which are unlikely to endure now that the Christmas season is over.  Meanwhile, over at Zero Hedge, they're insisting that the BLS is cooking the numbers by adjusting down the labor participation rate: "Real jobless rate is 11.4% with realistic labor participation rate."

So it's "hooray" with a lower-case "h."  Let's see what happens once the holiday effect wears off.


$cam said...

Of course the Obama economic numbers have been manipulated to Obama's political advantage.

John Kennedy's statisticians came up with the concept of "discouraged workers" to cut the unemployment numbers. Richard Nixon tried to force the BLS to announce only the lower of its two calculations, adjusted or unadjusted, whichever the lower figure happened to be in a given month, but without revealing which figure it was.

Jimmy Carter's administration deliberately understated the inflation rate by hiding the "underlying rate" that includes random, temporary or seasonal changes. During Reagan's Presidency, the BLS dropped anyone who hadn't actively looked for a job in two weeks. They also eliminated anyone under 20 who'd lost a job. For the Consumer Price Index, Reagan's people started counting the cost of renting a home, as opposed to purchasing one.

The CPI has continued to be manipulated over the years. During George H.W. Bush's time in office, Alan Greenspan and Michael Boskin argued that the index methodology should be revised downward to produce "better" results.

Examples: retail cost was replaced with discount retail cost; "quality improvements" should offset price increases (e.g. a car with airbags is worth more, and thus a higher sticker price should not count); if steak sales drop, it means that consumers have opted to switch to hamburgers, and therefore the price of hamburger should be the new constant. Better still, when prices for an item increase, it's now assumed that fewer people will buy them and thus their weight upon the CPI is deemed to have decreased. Conversely, if an item gets cheaper, the BLS assumes it will become more popular, and it is counted at more than 100% of itself.

These changes also impact the calculation of GDP, again to an incumbent's benefit. They also alter cost of living increases, Social Security payments, variable price contractors, and other areas of the economy that are tied to the CPI by formula. Change the CPI, and you change the payout.

Boskin and Greenspan's proposal was fully adopted during Bill Clinton's Presidency. Clinton also oversaw the elimination of further categories of workers from the unemployment rate, and the reduction of poverty reporting. Those policies were continued under George W. Bush, whose statisticians expanded the calculation of employed persons to include various part-time or "marginally attached" workers: essentially, anyone who worked at least 1 hour a week.

This isn't just an American thing. When Margaret Thatcher was Prime Minister of England, her government's methodology for determining the unemployment rate was revised 24 times. 23 of those times, it resulted in a lower number.

Just something to think about the next time Obama talks about "marked progress," or the Republicans talk about "proven failure."

Eric said...

Wow, that was an extremely comprehensive - and interesting - comment.

I tried to find the link but I heard on NPR yesterday that the typical "courier and messenger" hiring during the holidays is 10-15K but for this one it was >40K. That's a lot of low-paying, temporary jobs that doesn't really point to much of a recovery.