Econbrowser: Conspiracy theorists are wrong about unemployment rate

Econbrowser has a post up today thoroughly debunking the conspiracy theories surrounding last Friday’s BLS estimate that the unemployment rate fell to 7.8 percent last month: Watergate! Iran-Contra! BLS 2012.10.5?

The Bureau of Labor Statistics is staffed by career civil servants. Based on a survey of about 60,000 households each month, they estimate the number of people who say they are working, the number who say they are part of the labor force, the number who are working part-time but would prefer full-time, and so forth.

The resulting data is quite noisy from month to month, but over time it tracks closely with the monthly estimates from a separate survey of payroll establishments. It’s that second survey from which the BLS estimates the number of payroll jobs.

You can see all the data in last Friday’s release, or click here for my brief summary.

Since the estimates were released last week, op-ed pages and internet sites have been fueling the idea that the books are cooked — and even that the White House has direct influence in the improvement in the unemployment rate. The Nixon administration once tried to alter the data, but it’s collected and analyzed in ways that make today’s conspiracy theories borderline ridiculous.

And as Menzie Chin at Econbrowser clearly shows with lots of arithmetic, there’s nothing anomalous about last week’s estimates.

From his post:

It constantly amazes me that in an era of widely-available spreadsheet programs that can run regressions, and when ostensibly our policymakers and analysts are as educated as they ever have been, that crazy and easily-disprovable assertions of conspiracy have been leveled. It suggests to me that for some, destruction of the trust we have in a professional civil service is considered acceptable in the quest to gain political advantage.

Check out this graph — one of many in the post:

These are various data series tracking employment. The blue line is the total civilian employment as estimated by the household survey. You can see that all the series were tightly clustered back in 2009 but then diverged. Public employment lagged private gains, so the purple and green lines separated, and the household survey has underperformed the other estimates for the last two years.

Given the general noisiness of the household survey and given that the two previous months’ household surveys had been rather weak and did not track with the growth in nonfarm payroll employment, it makes sense that the most recent month would show less divergence.

Here’s the long-term unemployment trend from the BLS last week:

Given the trends of the recovery, there’s nothing bizarre about the unemployment rate being 7.8 percent right now. Of course, it’s quite likely, as the numbers bounce around, that it will tick upward again next month.

But the long-term trends are clear, and it’s a shame that Americans across the political spectrum aren’t willing to embrace the good news in these most recent BLS estimates, in today’s estimate of consumer sentiment, in the clear turn in housing, and in other recent upbeat economic releases.

At the local level, it’s impossible to ignore the 2012 surge in private investment.