Alternative view:
Most studies that find negative effects from television compare groups of children who watch television to those who do not, even though the economic situations of the two groups are in all likelihood very different, Mr. Gentzkow said. The new study, however, was based on what the authors call a "natural experiment" that resulted from the way television was introduced in the United States in the late 1940's and early 1950's, when some cities got TV service five years ahead of others.
Data from cities where preschoolers were exposed to the new technology, and data from cities where they were not, was correlated with test scores from about 300,000 students nationwide in 1965, as collected in the Coleman Report, a survey done under the Civil Rights Act. The study also looked at test scores from pre- and post-TV age groups within cities.
The result showed "very little difference and if anything, a slight positive advantage" in test scores for children who grew up watching TV early on, compared to those who did not, said Mr. Shapiro. In nonwhite households and those where English was a second language or the mother had less than a high school education, TV's positive effect was more marked.
There is very little that television has not been blamed for when it comes to children, whether it be shortened attention spans, a predilection to violence, earlier sexual activity or a general decline in values. The American Academy of Pediatrics became so concerned about its effects that it suggested in 2001 that children under 2 watch no TV, and that preschoolers older than 2 be limited to one to two hours a day of "quality" programming.
But many parents routinely ignore that warning. Average TV viewing among 2- to 5-year-olds ? the youngest viewers tracked by Nielsen Media Research ? crept up to 3 hours and 40 minutes a day in the 2004-5 TV season. A host of cable channels have are dedicated to the tiniest viewers.
Elizabeth A. Vandewater, associate professor of human development at the University of Texas and director of the Center for Research on Interactive Technology, Television and Children, praised the new study for adding "more evidence that television is not uniformly evil or bad," but said that it ignored "a host of evidence that shows that content matters a lot."
She said that "there is a huge body of evidence that educational television" can be good for children, as well as strong evidence that "violent content is related to antisocial aggressive behavior."
Mr. Gentzkow said the work had nothing to say about how television affects a child's focus, aggression or other behaviors, and that it merely looked at academic outcomes.
The authors also said the study was not meant to evaluate television content in the 1950's versus that of today. But Mr. Shapiro noted: "If you look at the top five children's programs in the 1950's and the equivalent list from 2003, the content is not as different as you might have thought."
The study was released this month by the National Bureau of Economic Research, a nonprofit, nonpartisan group of academic researchers based in Cambridge, Mass. The group is better known for putting out statistical analyses on the state of the economy. But it also serves as a forum for disseminating working papers of new economic research that has yet to be peer-reviewed, including the television study. Mr. Gentzkow said the study had been submitted for peer review to The American Economic Review.