Minimum wage and social costs

Over at Greg Mankiw's blog there's an interesting debate over minimum wage in the U.S.

Prof Mankiw says:

In the end, there is no good substitute for an appeal to facts. What the facts show is that the minimum wage is poorly targeted as an anti-poverty program. Moreover, while the evidence is controversial, some studies find significant long-term adverse effects. As a result, most economists prefer more efficient and better targeted anti-poverty tools, such as the EITC, which has grown significantly over the past few decades.

If minimum wage fails to keep up with inflation, all else being equal, it's likely that the social cost will lead to a negative net fiscal effect. So, which is more effective to reduce poverty: a social net, or a high primary income distribution? I thought I would explore the effects of minimum wage on net fiscal effects, since I know little on this. Further, it seems that this should be an especially hot topic in Canada given the recent ruling against Wal-Mart in the Saskatchewan labour dispute and the similar labour cases pending against the store elsewhere in Canada.

The following explanation by Jared Bernstein and L. Josh Bivens makes the most sense to me. Their paper was written for the Economic Policy Institute (EPI), a US-based “nonpartisan think tank.” It is specific to wages at Wal-Mart:

Part of this debate comes down to whether policy makers should be more concerned about primary or secondary income distributions. In other words, is it better to intervene directly and require employers to pay a living wage or to allow whatever compensation employers choose to pay but supplement it with government supports? The primary income distribution is that which holds before the effects of taxes and government transfer payments (i.e., Social Security payments, unemployment compensation, disability income, etc.) are factored in.
The secondary distribution includes the effect of all taxes and transfers. When it comes to offsetting the damage to the wage structure we document above, many policy analysts seem to believe that the sole intervention point is the secondary distribution. That is, they are quick to accept the primary distribution as an outcome that cannot and/or should not be altered. In their view, if the market is generating "too much" inequality, the government can offset this through redistributive fiscal policy. Many defenders of Wal-Mart's current business model advocate expanding the Earned Income Tax Credit (EITC) as the "correct" way to help Wal-Mart's workers.
While there is some merit in this view, this strategy makes U.S. workers and their households too reliant on the single instrument of fiscal redistribution—the expansion of transfer programs. EITC expansion, for example, is clearly a viable policy remedy for lost earnings, but there are reasons not to rely solely on this
strategy.
First, it relies on tax increases. We cannot ask American workers to depend exclusively on taxpayers and politicians continually ratcheting up their willingness to offset the degradation of the wage structure induced by Wal-Mart, not to mention globalization, the loss of manufacturing employment, union power, and so on.
Second, the federal budget is already constrained by the current and, more importantly, the projected gap between future federal outlays and revenues under current policies. According to the Congressional Budget Office, under plausible assumptions, that gap is expected to grow much wider in coming decades, largely due to the pressure of health care price increases and current tax policies.3 This does not mean that potentially useful policies like expanding the EITC should be off the table, but it does mean that any spending program outside of defense and homeland security faces a very steep challenge for at least the medium term. Wal-Mart defenders who argue that workers harmed by its practices should rely on government transfers to make ends meet are essentially telling those workers to get taxes raised if they want any help, and to struggle in the meantime.
Recent budget developments indicate that no help is on its way. While Wal-Mart supporters argue that it's fine for Medicaid to pick up the health coverage of uninsured workers, the president has proposed $5 billion in cuts to Medicaid over the next five years and has proposed an additional $5 billion in cuts to other programs for low-income people in his most recent budget. Further, in February of this year, Congress passed a budget reconciliation that included a $27 billion cut in Medicaid over 10 years. In other words, the tide is pushing hard against expanding these redistributive measures, and may be for some time to come.

Another point I found interesting in the article is this:

Essentially, the defenders of Wal-Mart argue that the price-depressing effects of Wal-Mart outrun the wage-depressing effect, leading to rising purchasing power for American workers. However, the prices that are reduced through Wal-Mart's expansion constitute an ever-shrinking share of American families' expenditures.

The paper provides further explanation.

Information on earned income tax from a Canadian perspective can be found here. If anyone can suggest some good up-to-date reading on earned income tax credits and minimum wage (especially commentary) in Canada, I'd love to hear about it.