The Obama administration continues to pretend that the U.S. economy is making a meaningful recovery.

Its establishment press acolytes are predictably and dutifully playing along. Government data releases with objectively mediocre results are treated as objects of wonder, while tangible signs of decay get waved off. Tuesday’s report from the Census Bureau on new-home sales showed a seasonally adjusted 1.5% increase in March, a level which was still 6% below January. At the Associated Press, aka the Administration’s Press, that was seen as “evidence of a sustained housing recovery.” Meanwhile, the self-described essential global news network “cautioned against reading too much into” Wednesday’s steep 5.7% drop in durable goods orders.

No amount of whitewashing will change the fact that this has been and continues to be the worst economic and employment recovery from a downturn since Franklin Delano Roosevelt extended the Great Depression by more than eight years with his statist New Deal policies and hyper-regulatory, business-intimidating agenda.

What has occurred in the job market hasn’t been a “recovery” by any reasonable definition. Forty-five months after the recession officially ended in June 2009, and 37 months after the February 2010 trough in seasonally adjusted employment, the number of Americans working has yet to return to its pre-recession high. Since World War II, full recoveries from employment lows have never taken longer than 21 months. As of March, a stunning gap of 2.86 million jobs remained from the January 2008 peak of 138.06 million — and that’s before considering workforce growth. An interactive chart available at The Hamilton Project informs us that if the economy continues to add the average of 159,000 jobs per month seen since employment growth finally began going positive, the true jobs gap won’t be filledĀ until late 2024.

It gets worse if you take a broader look at the government’s available data on job additions.

Uncle Sam’s Bureau of Labor Statistics uses two surveys to compile its monthly reports on jobs and unemployment. One of those surveys asks employers how many people are working at their establishments, and is the basis for the bureau’s official monthly estimates of total employment and jobs added or lost. At the state level, the establishment survey tells us how many people are working at employers located there.

The other survey is of households. The bureau identifies the number of working-age household members who are and aren’t working, and uses those results to estimate the various reported unemployment rates.

In the process of compiling the household survey, the government comes up with a different estimate of total employment which rarely receives much attention. The household survey’s total employment estimate is higher than the establishment survey’s, primarily because it includes the self-employed. (Secondarily, but to an unfortunately unknown extent, it also includes illegal immigrants who tell the bureau they are employed while working at establishments on a contract or off-the-books basis.)

While it can be very volatile from month to month, long-term changes in the household survey’s reported total employment figures deserve greater scrutiny and visibility than they have been receiving.

Here is what I found in comparing the two surveys’ employment growth results from March 2010 through March 2013 for each of the 50 states and the District of Columbia:

  • First, while the establishment survey’s total, arrived at by adding all individual state changes, shows that 5.92 million jobs were added during that time, the household survey arrived at similarly shows an increase of only 4.22 million — a stunning 1.7 million job difference.
  • Second, job growth per the household survey in states which voted for Obama in 2012, at 2.56%, was over one-third lower than the 3.89% job growth seen in states which gave their electoral votes to GOP challenger Mitt Romney. That variance is significantly greater than the 1.04-point or less than 20% difference (+4.21% Obama, +5.25% Romney) found in the establishment survey during the same time period

For those who believe that the just-mentioned difference is insignificant, if Obama-supporting states had added household survey jobs at the same rate as those that backed Romney, just over 1.2 million additional Americans would be employed today.