The collapse of the housing market has been an object lesson for America. Households and banks borrowed too much on expectations of continuing appreciation in real-estate prices. This extra borrowing inflated a bubble until it burst. By discounting the future too optimistically, we let the good times roll away.

The collapse of Chrysler and GM has been another object lesson. Management and unions pledged too many worker and retiree benefits on expectations of continuing demand for their gas-guzzlers. By discounting the future too optimistically, they let the good times roll away.

The private sector is now rethinking its unrealistic optimism. It has to, since its asset valuations have tumbled. Payrolls are shrinking. Benefits are being cut back. Both management and workers are accepting that they have to work harder for less. Government is upping the pressure by hiking taxes, requiring banks to raise more capital, and demanding more objective risk reporting.

But the government is not applying these lessons to itself. On the contrary, the public sector’s tendency to discount the future too optimistically is growing. It is pledging ever more payoffs to its employees and wards and concealing more than ever their true costs, even as private-sector incomes fall.

The public sector employs about one-sixth of the U.S. work force. Apart from the expansions and contractions of the military, that share hasn’t changed much since World War II. About one in five public-sector jobs is federal, with civilian government and the military each employing roughly 2 million. Nearly another million work in federally owned enterprises, most of them in the Postal Service.

Average federal pay is distinctly higher than private-sector pay. After adjusting for part-time work and the cash value of payment in kind, the Commerce Department’s Bureau of Economic Analysis (BEA) reports a wage and salary premium in 2008 of 33 percent for the military and 58 percent for civilian government. It wasn’t always that way. All of the military premium and nearly half the civilian premium were created between 2000 and 2005.

These premiums were needed to improve recruitment and retention. In 1997, the Congressional Budget Office concluded that the government paid 22 percent less than the private sector for similar jobs. Bolstering the military after 9/11 was also a priority.

For reasons of efficiency and fairness, the extra pay might have been coupled with reducing the job security and trimming the benefits that federal employees traditionally enjoy. It was not. On the contrary, the federal benefit edge has widened. It is not easy to measure how much, for the government refuses to publish direct comparisons. However, a few months ago, Chris Edwards of the Cato Institute deduced from BEA statistics that from 2000 to 2008, benefits grew a whopping $16,000 per full-time federal civilian employee, versus $3,000 per private employee. Federal benefits are now more than four times private benefits.

What was intended as catch-up, then, is now in overdrive. In 2008, federal employees were relatively well insulated from the financial crisis. Their benefits were guaranteed, they bore little risk of layoff, and the thriving business of government buoyed housing valuations for D.C.-area residents. A sense of shared national burden would have called for public-sector restraint in 2009. Instead, as millions of private workers lost their jobs and real incomes declined, civilian federal salaries were boosted 3.9 percent. Not surprisingly, resignation rates for federal employees have sunk to less than a third of private-sector averages.