David Brooks doesn't like the stimulus, as readers of his columns know. Today he engages in a bit of magical thinking in putting out his case for deficit reduction.

His first invention is telling us: "deficit spending in the middle of a debt crisis has different psychological effects than deficit spending at other times." This is very interesting, what "debt crisis" is Brooks referring to? We can point to a debt crisis in Greece, and arguably Portugal and Spain, but it is not clear what that has to do with the argument for stimulus in the United States. There were debt crises in Latin America in the 80s, no one ever raised these in the context of the Reagan era budget deficits.

In the real world we would look to things like the ratio of debt to GDP in the United States (@60 percent) and compare it to the ratios in other countries and to the U.S. at other points in time. There are several countries with debt to GDP ratios of far more than 100 percent who are able to borrow money with no difficulty. For example, Japan has a debt to GDP ratio of more than 110 percent yet it pays less than 1.5 percent interest on its long-term debt. Right after World War II the debt to GDP ratio in the United States was also over 110 percent, yet interest rates were low and the economy had decades of solid growth.

The next thing we would do in the real world is look at the interest rates that the United States is currently paying. At present the interest rate on 10-year Treasury bonds is about 3.2 percent, near a post-World War II low. In short, the debt crisis is magic -- an invention of David Brooks -- not something that exists in the world.

Brooks then decides that the stimulus did not work because the economy is not creating many jobs. Again, in the real world we would be interested in the size of the stimulus in order to determine how much impact we would expect it to have. The stimulus was scored at $787 billion. However, roughly $80 billion was a fix to the alternative minimum tax. This is done every year. Since no one ever expected to pay this tax, making this fix could not provide any stimulus to the economy. Roughly $100 billion was projected to be spent in 2011 or later, leaving $600 billion for 2009 and 2010, or $300 billion a year.

If we want to see the net stimulus from the government sector we also have to factor in the cutbacks from the state and local governments. These came to around $150 billion a year, leaving a net stimulus from the government sector of about $150 billion annually, or a bit more than 1 percent of GDP.

This boost from the government sector must counteract the impact of a falloff in annual construction spending (residential and non-residential) of more than $500 billion a year and a comparably sized reduction in annual consumption. In the real world, we are not surprised that $150 billion in net government stimulus cannot offset a decline of more than $1 trillion in private sector demand, but in the magical thinking of David Brooks this is a big disappointment. He notes that Harvard economist Edward Glaeser found no relationship between job creation and stimulus spending by state.

Actually, what Glaeser found was that there was no relationship between the portion of the stimulus that took the form of specific contracts or grants (about 20 percent of the stimulus) and the change in state unemployment rates. It would have been a remarkable coincidence if such a relationship did exist. The money awarded thus far in this category was $61.4 billion. In some cases this money has been handed out and spent, in other cases it has been banked for projects that may now just be beginning.

In any case, the total size of spending is small relative to the economy. If we assume that the money is mostly spent over a 2-year period, then it is roughly equal to 0.2 percent of GDP. Against this we have enormous variation in the economic impact of state and local budget cuts, the performance of housing markets, the performance of major industries (e.g. autos and oil), and the impact of trade. It is also worth noting that Glaeser examines changes in unemployment rates, which are also affected by demographics and state policies on unemployment insurance and other benefits, rather than job creation, which would be the more direct measure.

In short, any serious analyst would expect the impact of the spending examined by Glaeser to be swamped by other factors. We would predict that there would be no correlation between the spending in the category considered by Glaeser and the change in unemployment rates. In other words, Glaeser was testing nothing, his exercise tells us zero about the effectiveness of stimulus, except in the magical thinking of David Brooks.

Finally, Brooks recounts the stories of countries that grew successfully even as they pared back their budget deficits. Let see, we have Ireland and Denmark. These are both fine countries. But there is a very important difference between Ireland and the United States. Both are relatively small countries with very open economies. Trade is more than 70 percent of GDP in Ireland and more than 60 percent in Denmark. By comparison, it is just over 25 percent in the United States. If the governments in Ireland and Denmark cut back their spending it can be offset by increased net exports, especially if their currencies decline relative to the currencies of their trading partners. Does Brooks think this will happen in the U.S.?

Also, the central banks of these countries could respond to fiscal contraction by lowering interest rates. This is not an option in the U.S. at present with the interest rate already at zero.

In short, this is more magical thinking on Brooks' part. The experience of these countries do not provide relevant examples to the United States.

Of course, there is nothing wrong with magical thinking, except when it is likely to affect real world policy. The deficit hawks in public policy positions are running wild. They are throwing people out of work in the real world. Magical thinking like that displayed in Brooks' column will help their case.