What's the truth about the unemployment numbers?

The recent sudden drops in both initial unemployment insurance claims and unemployment rates have generated a slew of positive news stories and lifted White House spirits. Fox News contributor and former Clinton pollster Doug Schoen even writes about the “high-fiving in the White House.”

But other numbers show a much weaker job market and economy; the average unemployment duration remains near its all-time high, hiring is stuck near record lows, and there are almost 3 million workers in part-time rather than full-time jobs. Further, GDP grew just 1.7% last year and few new companies are being started. So which scenario is right?

First, let us set one thing straight. The apparent contradiction between a falling unemployment rate and other indicators isn’t because the Obama administration “manipulates numbers to get [the] unemployment headline under 9%” as Rush Limbaugh claims. The current "recovery" is so unusual that the normal rules for relating the unemployment rate to other labor market indicators don’t hold; indeed those rules can be quite misleading.

Take initial unemployment insurance claims. As the Wall Street Journal reported the end of last month: “Analysts generally believe the economy is adding jobs when jobless claims are consistently below 400,000.” Seasonally adjusted initial unemployment insurance claims held at around 353,797 on Thursday, and they have been consistently below 400,000 since the beginning of December.

But the “below 400,000” rule only makes sense if new hiring parallels growth during past recoveries.

Forecasting the number of people employed by focusing on workers filing for unemployment insurance is like guessing a pool’s water level by measuring how much flows out but ignoring the rate at which water is being added.

Few journalists appear aware that the Bureau of Labor Statistics (BLS) has collected data on new hires since December 2000. From that date to the recent recession, 5.2 million Americans were hired on average each month. The recession dropped this number substantially, with the average falling to 4.3 million hires per month. In December 2008, just before Obama took office, only 4.1 million Americans were hired.

But hiring has fallen even further during the “recovery” -- falling to an average of just 3.9 million per month. The number of hires was still stuck at 3.9 million during the last month for which data are available, December 2011.

The BLS’s seasonal adjustments to the job numbers also make the labor market seem stronger than it actually is.

The big news this month was that “seasonally adjusted” payroll employment increased by 243,000 in January. Yet, the “unadjusted” or actual raw totals showed a loss of 2.7 million jobs.

It’s useful to try accounting for the well-known layoffs that occur at different times of the year, such as right after Christmas, but how you make the adjustment is important.

Although the Bureau of Labor Statistics does not disclose its exact adjustment method, the problem is that their adjustments put most weight on recent years.

Thus, the greater the drop in job numbers a few years ago, the greater the seasonally adjusted increase in jobs now.

It is easy to see the huge difference this makes. The average December to January job loss over the preceding three years was 3.1 million. So the most recent 2.7 million drop in jobs was indeed relatively small. In the five years before the recession, December to January job losses averaged 2.7 million. If pre-recession monthly changes had been used as the base, the most recent relative jobs “gain” would have been much smaller.

The very weak recovery means the official unemployment rate is an unreliable barometer of the labor market. People are only counted as unemployed if they have been actively looking for work in the past four weeks. It is good news when the number of unemployed falls due to more hires. It is not so good if the number falls due to people giving up looking for work.

Historically people give up looking for work during recessions and resume looking for work in recoveries. Not this time.

While 1.66 million net jobs have been added during the Obama "recovery,” over that same time the number of working age Americans not in the labor force rose by 7.14 million. There is no comparable post-World War II "recovery" where this type of exodus has occurred.

Some have attempted to explain this anomaly by an aging workforce, but labor force participation rates have fallen for all age groups except workers age 65 and over.

Rules of thumb work pretty well most of the time. But you have to know how the numbers are put together to tell when those rules break down.

Record-setting drops in new hires and people leaving the work force are just such cases. The “high-fiving in the White House” is far too premature.

John R. Lott Jr. is a FOXNews.com contributor. He is an economist and co-author of "Debacle: Obama's War on Jobs and Growth and What We Can Do Now to Regain Our Future.".