Stephen Bainbridge makes the case against public sector unions:
A core problem with public sector unionism is that it creates a uniquely powerful interest group. In theory, bureaucrats are supposed to work for and be accountable to the elected representatives of the people. But suppose those bureaucrats organize into large, well-funded, powerful unions that can tip election results. With very few and very unique exceptions, no workplace in which the employees elect the supervisors functions well for long. ...
In effect, public sector unionism thus means that representatives of the union will often be on both sides of the collective bargaining table. On the one side, the de jure union leaders. On the other side, the bought and paid for politicians. No wonder public sector union wages and benefits are breaking the back of state budgets. They are bargaining with themselves rather than with an arms'-length opponent.
This model is perfectly logical. The problem is that it does not reflect actual reality at all. If public sector unions were merely bargaining with themselves, wouldn't they be winning unbelievably fat contracts that paid their workers considerably more than they could get in the private sector? (They aren't.) For that matter, would they be facing contract rollbacks in states across the country? It seems obvious that there's a very powerful adversary at the bargaining table -- namely, the desire by elected officials to minimize taxes on their constituents.
Now, this doesn't mean public employee unions are always harmless. Indeed, one result of the clash between politicians seeking to keep taxes low and public employees seeking higher renumeration is the funneling of compensation into pensions, whose costs are long-term.
The split on the Wisconsin fight on the center-left -- Bainbridge is not center--left, but he typifies the right-wing view -- revolves around whether you view this through a broad or a narrow political economy model. Public union critics like Joe Klein and Chuck Lane are interpreting the role of public sector unions within the narrow realm of the question of public sector compensation. And it is true -- within that narrow realm, the public employee unions are, if not all-powerful or even close, far more powerful than any other interest group. Smashing the unions thus looks like a triumph of public policy "as free of special interest influence as possible," as Chuck puts it.
But, of course, in the broader spectrum, public unions carry weight far beyond narrow questions, as a force for middle-class economic interests generally. That is the point supporters like Paul Krugman are making. In the real world, politics is dominated by the influence of the rich and the business lobby, with unions providing a small countervailing force. Breaking up the public sector unions would have many results, but one of those surely would be to increase the relative power of the business lobby even more.
The burgeoning success of the education reform movement shows that it's possible to fight public employee unions when they take positions contrary to the public interest. Scott Walker's attempt to virtually wipe out the public employee unions -- or, at least, the ones that campaign for his opponents -- is not a way to create a more rational policy process.
Update: Bainbridge protests that I failed to account for research cited in his post. I disagree. The research shows that public unions have influence, a point with which I agree. It doesn't back up his sweeping claims that public unions are a "uniquely powerful interest group" that is "bargaining with themselves rather than with an arms'-length opponent."