Teachers with long commutes are more likely to leave their jobs

No one enjoys sitting in traffic.

That is the motivating principle of our recent study examining whether teachers with longer commute times are more likely to leave their schools.

There is a robust body of research on commuting outside education that shows how longer commutes can lead to employee turnover. Commuting is stressful, especially by car, which can be bad for workers’ health. That stress can spill over into work and home. Employees with longer commutes can feel less connected to their workplaces, perhaps because commuting reduces opportunities to form social connections with colleagues before or after work. Commuting can also take away from time with family or after-work leisure activities.

Perhaps it’s not surprising, then, that teachers account for their commute when choosing a job. But we wondered if teachers with longer commutes might be more likely to turn over, too. Commuting makes work (and life) less enjoyable. And teachers have many options for schools to employ them. Wouldn’t teachers with long commutes tend to turn over more frequently, preferring schools closer to their homes? This movement could have implications for students, since high levels of teacher turnover can disrupt student learning.

Getting data on teachers’ commute times

Turns out, research had not explored the link between commute times and teacher turnover. The main challenge, as you might guess, is knowing teachers’ commute times. With access to teachers’ home addresses, commute times can be calculated, but most administrative data sets researchers use to study teacher turnover don’t have this information.

Fortunately, human resources officials in Metropolitan Nashville Public Schools (MNPS) were willing to share address information as part of deidentified personnel records we used for this research. The district provided address information for all employees from the 2008-09 to 2019-20 school years. We could then use an online application that draws on historical traffic data to calculate the typical driving time between two addresses—in our case, a teacher’s home and work address—at a given time of day (we used 7 a.m. and 3 p.m.).

We could be confident that these calculated driving times equated to commute times because relatively anemic public transportation options mean that nearly everyone in Nashville commutes by car. In fact, according to Census data, among Nashville workers not working at home, roughly 85% drive to work, and 11% carpool. Just 4% use public transportation, walk, or bike. That context means that our study looks almost exclusively at car commuting in an urban district and may not generalize to other settings or commuting types.

How long are Nashville teachers’ commutes?

The average one-way commute time for teachers in our data was 18.8 minutes. But there was a distribution around this mean. About 19% of teachers commuted 10 minutes or fewer each way, while 12% had one-way commutes of more than 30 minutes.

High school teachers’ commutes were a little longer, on average, which makes sense, given that there are fewer high schools than elementary or middle schools. Commutes were also slightly longer for teachers in schools with lower test scores and those serving higher proportions of economically disadvantaged students. This pattern may reflect that middle-class professionals like teachers are less likely to live in the lower-income neighborhoods where those schools are often located.

Longer commutes predict turnover

Accounting for characteristics of educators, their neighborhoods, and the schools in which they work, we find that teachers with longer commutes are more likely to transfer to other schools within MNPS. The association between commute time and transfer probability is almost linear. Each one-minute increase in one-way commute time increases the likelihood of transferring schools in the next year by about 0.3 percentage points. Compared to a teacher with a commute of 10 to 15 minutes, a teacher with a 40- to 45-minute commute is about 10 percentage points more likely to transfer in a given year. The average rate of transfer for all teachers is only about 11%, so this difference is very large.

We also find an association between commute time and the likelihood of leaving the districts’ schools altogether. This tendency is concentrated among teachers with the longest commutes (40 minutes or more each way). Many teachers with these long commute times live outside MNPS. They could be exiting the district to take jobs in their home districts, though we can only observe teachers who stay within the district, so we cannot say for sure.

Interestingly, when we compare before-and-after commute times among teachers who switch schools within MNPS, we find that one-way commutes decrease by about 1.1 minutes, on average. While this may seem like a small change, it is not insubstantial, representing 6% of the 19-minute average pre-transfer commute.

Teachers with longer commutes are absent from work more often, get lower observation ratings

Research outside education has shown that employees with longer commutes are absent from work more often. We looked for such a pattern in MNPS personnel records and, indeed, MNPS teachers with longer commutes miss more days of work per school year. The average teacher with a 10-minute commute misses about 10.5 days per year, while at 40 minutes, teachers average about 12 days absent.

We also tested whether commute time correlated with classroom observation ratings. The intuition here is that commute times may compete against teachers’ time spent focusing on teaching. Again, we found worse outcomes for teachers with long commutes. Among teachers traveling only 10 minutes to work, the average rating was about 3.95 on a 5-point scale. At 40 minutes, the average rating was about 0.10 points lower, accounting for other factors. Among teachers in tested classrooms, value-added scores followed a similar pattern, though smaller samples made results less precise.

What do these patterns mean for teachers and schools?

Teachers with longer commutes leave their schools at higher rates, are absent from work more often, and are somewhat less effective. Although our methodological approach accounts for many factors that might confound the relationship between commute time and these job outcomes, we stop short of saying for sure that our results are causal. People choose their residences and workplaces for many reasons, so we can’t be sure that it’s the commute itself (and not some unobserved factor related to longer commutes) that is negatively affecting turnover and performance.  

Yet, we think that even if our results are descriptive rather than causal, they still have practical relevance. Shorter commutes are associated with better outcomes for teachers. That suggests that there are benefits to teachers working in the communities in which they live. In filling teacher vacancies, schools can expect greater longevity, attendance, and performance from the teacher who lives closer. Relatedly, when hiring teachers who live further away, principals may need to consider the strain the commute puts on the teacher and offer additional supports when they can. For example, it may help teachers with long commutes to have limited duties before school.

Perhaps more importantly, teachers should be aware of the advantages of living near work. Finding a long-term fit may be easier at a school that is close to home. School districts may be able to promote workforce stability by providing teacher applicants with information about the potential tradeoffs of accepting a job that is closer to or further from home, or about alternative transportation options.

Our results also suggest that commute times may contribute to higher turnover rates in schools serving larger proportions of low-income students. As middle-class professionals, teachers probably are less likely to live in lower-income neighborhoods, making travel time to work longer. Compensating for longer teacher commute times joins a long list of reasons districts should aim to direct resources toward making teaching more attractive in such hard-to-staff schools.