A New School Myth

Public-school teachers choose public schools for their own children more often than the general public.

In "The History of a Hoax," a wonderful New York Times Magazine article last March, Yale professor Barry O'Neill recounted his search for the origins of two widely quoted lists of top school problems. There was a 1940's list featuring gum-chewing, talking, and not putting paper in wastebaskets, and a 1980's version with drug abuse, suicide, pregnancy, and violence. The contrast between the lists was supposed to show how bad our schools had become. People who cited them had no idea where they came from; everyone just picked them up from someone else. O'Neill never did unearth any original studies because there weren't any; instead, he discovered a man who had created the two lists. The man said he knew what to include because he had gone to school in the 40's and "read the newspapers." In other words, the lists were neither polls nor scientific surveys. They were one man's invention circulated by many as social science.

I cited the lists myself until I learned better. So did columnist George Will and former Education Secretary William Bennett. Will and Bennett have also been going around saying that those who know the schools best -- the people who teach in them -- are more likely than other Americans, to put their own children into private schools. That's a myth, too. But it gets repeated often by people who are promoting voucher plans and for-profit schools.

Now these folks have a study by Denis P. Doyle, "Where Connoisseurs Send Their Children to School," that pretends to be research proving them right. But Doyle's own data actually show the opposite: Public-school "connoisseurs" (teachers) choose public schools for their children more often than the general public and far more often than other families with similar incomes. Even three out of four private-school teachers choose public schools for some of their children and two-thirds have all of their children in public systems.

Nationally, Doyle's data show, only 12.1 percent of public-school teachers enroll any of their children in private schools, compared with 13 .1 percent of the public. About a third of those teachers also have children in public schools. That means that 92 percent of all public-school teacher families send some or all of their children to public schools, compared with 90 percent of the public. Public-school teachers are more likely than the public to use public schools in 40 of the 50 states. And in 6 of the 10 states where teachers are more likely than the public to use private schools, the difference is less than 1 percent.

Teachers are typically in two-income households, so they tend to have higher family incomes than the general population. And higher-income people tend to use private schools more than lower-income people do. So you would think that the higher-income "connoisseurs" would follow that pattern.

Not so. In the $30,000-$70,000 bracket, 11.6 percent of all public-school teachers have a child in private school, according to Doyle's data, compared with 15 .2 percent of other families in that bracket. At an income level of $70,000 or more, only 15.2 percent of teachers use private schools, compared with 24.2 percent of comparable families.

In big cities, the overall proportion of families using private education is above the national average. That's not surprising, since city schools cope with education's highest concentrations of social and behavioral problems. What is surprising is that Doyle's numbers -- though not his words -- show that public-school teachers in a large majority of the 100 largest U.S. cities choose public schools more frequently than other city families, and at about the same rate in all but a handful of the rest.

Examined by income level, the preference of the city-dwelling public-school teachers for public education is even more marked: Teacher families earning $70,000 or more are far more likely than comparable city families to use public schools. Middle-income teachers ($35,000 -$70,000) in many cities are also more likely than other families to choose public schools. And city public-school teachers are at least twice as likely as city private-school teachers to use the public schools. Doyle questions the "allegiance" of teachers to public education, but his report shows strong allegiance.

The families in urban areas who send their children to private schools mostly do so because they fear violence and disruption in public schools. But much of the problem in public schools is caused by a small fraction of students, 2 to 3 percent. Doyle argues, though, in a conclusion as flawed as his analysis, that the "trapped" 97 percent should be given vouchers to attend private schools. Wouldn't it make more sense to send the 3 percent to private schools? Will private schools take and keep them? The toughest students I had as a teacher were those who were kicked out of private school.

The only realistic way to improve education for city families is to make every public school a safe, orderly place where students meet high standards of conduct and academic achievement. That requires hard work and commitment. And that's not a myth.