When recessions end, as they always do, economies recover. Yet today’s bipartisan gloom does not welcome improving economic data.

Republicans believe that acknowledging any good economic news must somehow give credit to President Obama’s penchant for deficit spending. Democrats believe that disregarding or downplaying favorable economic news provides an argument for even more deficit spending — an economic theory that could be labeled the Grecian Formula.

For dismal politicians and dreary pundits, recent economic news has been terribly embarrassing. Retail sales rose by 7.6% over the past year, and by 1.6% in March alone. Housing starts were also up 1.6% in March, and 20.2% from a year ago. Industrial production increased at a 7.8% annual rate in the first quarter, with the March figure up 6% from last June. The Conference Board’s leading economic indicators rose 1.4% in March and 5.6% over the past six months.

When facts are so uncooperative, some seek refuge in theory. Former Labor Secretary Robert Reich, in a recent Wall Street Journal op-ed, argued that the recent good news is flatly impossible:

“The approaching recovery,” he wrote, “will be tepid because so many people will lack the money needed to buy all the goods and services the economy can produce. … Given how many Americans are unemployed or underemployed, it’s hard to see where we get sufficient demand to support a vigorous recovery.”

That demand-side logic has a big problem: If high unemployment prevented vigorous recovery, then no economy in world history could ever have experienced a vigorous rebound from a deep recession. But that’s not true. Aside from numerous “economic miracles” abroad, there were at least two here at home.

In 1982, the unemployment rate hit 10.8% and averaged 9.7% all year. Yet the economy nonetheless grew by 4.5% in 1983, 7.2% in 1984 and continued growing at a 4.3% pace for seven years.

In 1933, the unemployment rate was 25.2%. Yet the economy grew by 10.9% in 1934, 8.9% in 1935, and 13% in 1936. According to Reich’s Keynesian musings, what happened in 1983–89 and 1934–36 was inconceivable.

Reich says: “Some economic cheerleaders say rising stock prices are making consumers feel wealthier and therefore readier to spend. But most Americans’ biggest asset is their homes. The ‘wealth effect’ is felt mainly by the richest 10%, whose net worth is largely stocks and bonds. The top 10% accounted for about half of total national income in 2007.

“But they were only about 40% of total spending. A vigorous jobs recovery can’t be based on 40% of what was spent before the economy collapsed.”

The stock market did not rise for no reason, but because first-quarter earnings of the S&P 500 companies that have reported so far were up about 75% from a year ago, with sales up 6.2% (or 7.8% for nonfinancial firms). And household wealth gains have not just been in stocks, as Reich imagines, but in widely held bonds, pension funds and, yes, homeowner equity.

From the first to the last quarter of last year, the Federal Reserve estimates that homeowner equity rose by $1 trillion, or 20%. But home equity accounted for only $6.3 trillion of $54.2 trillion in total household wealth, which rose 11.7% in the same period (and probably 20% by now).

Adding about $10 trillion to wealth over the past year may not matter in Reichian theory, but it does in fact. Prospering retailers like Wal-Mart and Costco do not depend on the top 10%, although the net worth of higher-income families is critically important for sales of new cars and homes.

Reich claims, “Most households rely on two wage-earners, of whom at least one is now likely to be unemployed, underemployed or in danger of losing a job.” In reality, only 31.3% of households had two wage-earners or job-seekers in 2009. Reich may have mistakenly defined households to include only working couples, not singles or retirees.

Yet the Census Bureau reports that among two-earner married couples, just 10.2% had one spouse unemployed last year, and fewer than 1% had both spouses unemployed. In short, the danger of one spouse losing a job was one in 10 — not what most people consider “likely.”

If not unemployed, perhaps many were underemployed? “Among those with jobs,” says Reich, “more and more have accepted lower pay and benefits as a condition for keeping them. Or they have lost higher-paying jobs and are now in new ones that pay less.”

If falling wages and benefits were nearly as likely as Reich suggests, then average hourly compensation (wages and benefits) would have been falling. Yet hourly compensation among nonfarm businesses rose by 3% in 2008 and 2% in 2009.

For comparison, nonfarm compensation also rose by just 2.5% a year from 1993 to 1997, when Reich was Bill Clinton’s labor secretary. Adjusted for inflation, real hourly compensation fell every year from 1993 to 1995 and barely increased in 1996 and 1997. Measured by real compensation, the 1993–97 expansion was worse than the 2008-09 recession.

“Since the start of the Great Recession in December 2007, the economy has shed 8.4 million jobs,” says Reich, “and failed to create another 2.7 million required by an ever-larger pool of potential workers. … This means even if we enjoy a vigorous recovery that produces, say, 300,000 net new jobs a month, we could be looking at five to eight years before catching up to where we were before the recession began.”

Actually, it would take 28 months to add back 8.4 million jobs at the rate of 300,000 a month. The labor force has not grown at all since December 2007, so Reich’s 2.7 million “potential workers” are imaginary.

If Reich merely meant to say that it will take 5–8 years to get the unemployment rate back below 5%, that is true but trivial: It was not until late in 2007 that unemployment got back to the pre-recession levels of early 2000.

Reich dismisses facts in favor of offhand impressions and his circular theory that only more consumer spending can create more jobs, and only more jobs can create more consumer spending. As the economic facts look better and better, Robert Reich’s theories looks worse and worse.