There are many ideologues who have their own set of truths, who refuse to listen to any evidence that points in a different direction. However George Will is in a class all by himself. He refuses to pay attention to the evidence that he himself puts forward.

Exhibit A is a classic Will rant at the left for being unwilling to consider the cultural factors that affect poverty. The hero of this piece is Daniel Patrick Moynihan:

"In March 1965, Moynihan, then 37 and assistant secretary of labor, wrote that 'the center of the tangle of pathology' in inner cities — this was five months before the Watts riots — was the fact that 23.6 percent of black children were born to single women, compared with just 3.07 percent of white children."

Will then goes on to tell readers:

"Forty-nine years later, 41 percent of all American children are born out of wedlock; almost half of all first births are to unmarried women, as are 54 percent and 72 percent of all Hispanic and black births, respectively."

If we follow Will's link we find that 29 percent of non-Hispanic white children are born out of wedlock today.That's considerably higher than the 23.6 percent of black children born out of wedlock in 1965 that Moynihan thought was the explanation for their poverty. Yet the poverty rate for white children today is less than 10 percent. It was over 40 percent for African American children in the mid-1960s.

If black children in 1965 had a poverty rate that is more than four times the poverty rate for white children in 2012, despite their lower rate of births to unmarried mothers, this would seem pretty solid evidence that being born to an unmarried mother is not the main factor in determining child poverty.

The far more important factor is the earnings potential for the children's parent(s). This is determined by the factors that Will discourages us from considering, such as macroeconomic policy, trade policy, policies toward labor organizing, and other policy choices that will determine the health of the labor market facing parents of young children. Of course their access to health care and quality child care will also be important factors determining the children's well-being which are picked up in the Census Bureau's supplemental poverty measure. 

And it is easy to show that government policy has made poverty worse on this score. The fall in employment rates following the 2001 recession was associated with a rise in poverty. The much sharper fall in employment rates following the 2008 recession was associated with an even larger rise in poverty. The decision of Congress to run high unemployment budgets (i.e. lower deficits) also will predictably result in a higher poverty rate for children.

There is a different and trivially true point that perhaps Will is trying to make in his column. Children with two loving committed parents will fare better in life on average than children who only have one parent.

Many liberals don't find this very useful from a policy perspective because they don't think the government is very good at creating and preserving good marriages. However we do know how to run economic policies that will reduce the unemployment rate.