People carry their sleds on Boston Common on Feb. 9 during a snowstorm in Boston. A new study shows that schools closing for snow days has less of an effect on student achievement than absenteeism when schools are open. (Steven Senne/AP)

It’s become a maxim in education: More learning time leads to greater student achievement. So when schools close for snow — as they did over and over this winter across many states — the assumption is that student achievement will suffer.

Not so, says Joshua Goodman, an assistant professor of public policy at Harvard’s Kennedy School of Government. Goodman examined weather data, student test scores and attendance data in Massachusetts between 2003 and 2010. He found that the number of canceled school days because of snow in a given year had no impact on children’s math and reading test scores.

Instead, it was the number of days that were merely snowy — when schools remained open, but many students were absent — that appeared to hurt achievement, particularly in math.

He found that each one-day increase in bad-weather absenteeism reduces math achievement by up to 5 percent of a standard deviation. Overall, differences in average student attendance could account for as much as 25 percent of the state’s achievement gap between poor children and their more affluent peers, he concluded.

The upshot is not that superintendents should shut down schools when the first snowflake falls, Goodman said, but that student absenteeism is a bigger problem than it usually gets credit for in national education debates.

“Somehow, schools seem to have a way of dealing with the closures, but they haven’t found a way to deal with absences,” said Goodman, whose research is set to appear online Thursday in the journal Education Next, which is published by the conservative Hoover Institution.

Snow days do have a significant effect on student achievement in schools in which more than half of the students qualify for free or reduced-price lunch, a rough proxy for poverty. But even in those schools, the snow-day effect is much smaller than the effect of absenteeism, Goodman said.

Goodman, who is a former high school math teacher, hypothesized that when a bunch of children miss class on a snowy day, their teacher is faced with a dilemma when they return: Take time to help students catch up and therefore slow down the rest of the class, or continue on and risk that some students will be lost?

“There’s not a great answer to that question,” he said.

Previous studies have found that higher snowfall amounts tend to be correlated with lower student test scores, and researchers and policymakers assumed that school closures explained the link. Goodman, who first published a version of his work last year as a National Bureau of Economic Research working paper, is the first to try to parse the difference between the effects of bad-weather closures and bad-weather absenteeism.

Students missed an average of 2.2 days of class a year because of snow-day closures, according to Goodman. They missed far more time — eight days — because of absences for reasons that include weather, illness, vacation and skipping class. That average masks a wide range: Poor children, for example, were absent an average of 10 days per year, three more days than their more affluent peers.

So perhaps the focus on emulating high-performing charter schools that have longer school days or a longer school year is misplaced, Goodman suggested. Perhaps schools should first work on improving attendance, he said.

“There’s all this talk about extending the school day or school year to get in more instructional time, all of which is expensive and unclear what kind of effect that will have,” Goodman said. “But we’ve got what you might think of as low-hanging fruit, which is: Can we improve attendance rates or improve the way schools and teachers deal with student absences?”

Piles of studies have shown that poor student attendance is correlated with weaker academic performance and that chronically absent students — who miss more than 10 percent of the school year, or 18 days out of a traditional 180-day year — are far more likely to struggle and eventually drop out.

Stacy L. Scott, superintendent of the Framingham public school system west of Boston, agreed that teachers find ways to adjust their lessons to deal with closures. The district of 8,500 students averages five snow days a year and had seven this winter. “The teachers do a good job of being able to sustain the learning and keep it going,” Scott said.

Ross Wilson, chief of staff for Boston Public Schools, said that the school system knows that absenteeism can be destructive and has made a big push to improve student attendance. But school system officials also believe it’s important to find ways to extend the school year and the school day, he said.

“I don’t think it’s better attendance or a longer school year,” he said. “I think it should be both.”

Wilson couldn’t quite stomach the idea that snow days don’t affect student test scores, however. During a winter with record snowfall, Boston closed school eight times this year for inclement weather, and students will not make up those lost days until June — long after state testing has concluded.

“Almost two weeks — I can’t imagine how that would not have an impact on student test scores,” Wilson said. “It does conflict with common sense. If kids are having less learning time, you would expect that there would be an impact on their learning.”

Goodman agreed. Though his study showed that schools are good at dealing with an average number of snow days, he said, “this winter was anything but average.”