If David Cay Johnston is right and total compensation is what counts, Rick Hess’s reply to Diane Ravitch is particularly interesting:

As the University of Arkansas’s Robert Costrell calculated out last week, Wisconsin’s public employees collect 74 cents in benefits for every dollar in salary, more than triple the rate for their private sector counterparts. In Milwaukee, the average ten-month salary for a Milwaukee Public Schools teacher is $56,500 but the total cost to MPS is $100,005. Why? For one thing, the Wisconsin state pension calls for a 6.8% employer contribution and a 6.2% employee contribution, but MPS agreed back in 1996 to pay the employee share as well. For another, the 1982 collective bargaining agreement also grants MPS teachers a second pension, funded entirely by a 4.2% district contribution. As for health care, MPS spends 39% of wages on health insurance, compared to a private-sector norm of 11%. MPS pays the full premium for medical and vision benefits and also grants full health care to retirees, with the district picking up the entire premium in effect at retirement.

DCJ would argue that MPS wasn’t just picking up the employee share. Rather, MPS was increasing total compensation. Remember this factoid about MPS?

Basically, reading and math scores were the same for district schools and for choice schools. But choice schools receive a maximum amount of roughly $7,000 per pupil while the MPS spends over $16,000. I have to assume that these numbers aren’t directly comparable. A crude reading of these numbers would suggest that choice schools are “buying” the same educational outcomes at less than half the cost — an extraordinary result if true. To be charitable, let’s assume that these numbers mask hidden subsidies for choice schools. Parochial schools, for example, have existing infrastructure, and so one assumes that capital expenditures aren’t factored in to per pupil spending in the same way as they are for MPS.
But surely at least some of this gap reflects differences in the cost structure between choice schools and district schools.

Given this extraordinary gap, it’s hard to imagine an employer responsive to the interests of all stakeholders, and not just employees, making these extraordinary concessions.

Back to Rick’s post:

Why steer so much money into benefits that do little to improve schooling or services? It’s simple, really. Benefit costs are long-term and largely invisible to taxpayers, while they satisfy the demands of political, influential union leaders. Like sugar subsidies or pet Pentagon pork, this makes benefit giveaways as tempting as catnip to policymakers in Madison or Milwaukee. They get to say yes to a concerned constituency, and nobody much complains or cares–at least until the bill comes due (which is now). It’s far too appealing for state and local officials to shovel out the goodies, figuring that the public employee union leaders will appreciate it and no one else will really notice.
While politically attractive, these benefits are an awful way to spend scarce money. Retirement benefits and massively subsidized health care aren’t much use in recruiting young educators in the modern labor market (in large part, because these benefits are easy for potential hires to overlook). And these promises create huge obligations that suck dollars from schools and classrooms even as they leave future policymakers to foot the bill.

Very correctly, Rick doesn’t blame teachers for this predicament. It is perfectly natural that teachers will fight for the most generous compensation package they can get. Rather, the blame should be assigned to elected officials who want to curry favor with a large and powerful constituency. He makes another important point:

The problem with collective bargaining by public employees is that these unions are unchecked by competition and wield massive influence as they help to elect their bosses.

There are many confounding variables when we compare union and non-union shops in the private or the public sector, but the key difference in the private sector is that firms can be formed and dissolved with relative ease. We can say that outcomes in, say, Indiana vs. Wisconsin are a meaningful test, but it’s not nearly as good a test as offering the residents of any state the opportunity to choose between public services offered by a union or non-union shop, just as consumers can purchase automobiles from union or non-union shops. And the latter example tells us that union shops can perform pretty well in certain contexts. The really key thing is allowing for competition and a variety of providers.