Are Commutes Really Getting Longer?

There are only three certainties in life: death, taxes, and constant complaining about traffic on I-95 in Fairfield County.

The public sphere seems to be particularly attuned to that last one, as evidenced by a recent CT Insider article addressing increased post-COVID commute times in Connecticut.

According to that analysis of U.S. Census data, mean commute time in the state increased about half a minute between 2016 and 2024.

The author suggests that’s because drivers in Fairfield County spent “more time on I-95 or the Merritt Parkway,” but is this data truly indicating such a change?

I would argue that, no, the U.S. Census data does not support that, and the suggestion that traffic is the primary culprit masks important changes in the underlying data. 

A Whole New World 

Writing at my desk at home on a Friday is a reminder that nowadays, a significant proportion of workers have no commute at all, which matters hugely when analyzing data.

The American Community Survey is asking folks how long their commute is and the mode (driving, walking, transit) they take to work.

They take that total time commuting and divide by the number of commuters to determine average commute time.

However, if you work from home, you are not included in that figure. And, between the 2016 and 2024 five-year ACS, the number of people recorded as working from home in Connecticut grew by 209,971, or 259%.

Despite a growth in population and number of total workers, there were 39,452 fewer total commuters from Fairfield County in 2024.

If we look at just Fairfield County, the WFH share grew from 5.7% of workers to 19.3%. As a result, despite a growth in population and number of total workers in the county, there were actually 39,452 fewer total commuters from Fairfield County in 2024 compared to 2016.  

We can take the data a step further and look at the “buckets” of commute times to get a sense of how commutes have changed.

Looking at Fairfield County, the biggest changes were in the shortest commute times, with commutes shorter than 25 minutes declining 36,450.

The segment that saw growth was in those commuting between 35-60 minutes, which increased 4,070 people.

Remote Work Trends

Now, if the composition of commutes structurally shifted from people with short commutes to people with long commutes, then we would see the average commute increase, but that doesn’t mean traffic dynamics are the root cause.

It certainly could be! However, it could just as easily be that workers attracted to the area post-pandemic have longer distances to commute overall.

Even still, it is worth noting that the percentage of workers who live in Connecticut but work in another state was flat at 7.7% between the two surveys.  

The percentage of workers who live in Connecticut but work in another state was flat at 7.7% between the two surveys.

The effects of WFH on more local trends can be even more pronounced. For example, Powell shows that Weston was the town that has the longest commute in the state, with commute times up eight minutes between the two surveys.

However, Weston also saw their WFH share increase from 16.7% in the 2016 survey to 44.9% in 2024, while the total number of commuters declined by 30%.

Meanwhile, the share of commuters with commutes between 20-24 minutes declined from 19.9% to 10.4% while the share of commuters who commute more than 90 minutes increased from 14.4% to 22.2%.

So, in that town’s case, the average increase is clearly due to changing commuting dynamics rather than traffic. 

Plenty of Caveats 

This is not to say that WFH can perfectly explain an increase in commutes. For example, if we look at all 169 towns there is almost no correlation between changes in WFH share in a town and its average commute time. 

As I mentioned previously, it is also perfectly possible that individual commute times have increased slightly because of traffic AND an increase in travel distance.

One could imagine a scenario where more people relocating north of Stamford only exacerbates the I-95 bottleneck, even if fewer people in total are commuting overall.

At the same time, the ACS does not capture the complexities of hybrid work. The way the question is phrased simply asks respondents how long it took them to get from home to work last week, not how often they make the trip.

It is perfectly possible that individual commute times have increased because of traffic and an increase in travel distance.

So, a person who commutes 40 minutes from home to work three days a week is counted the same as a person who commutes 40 minutes to work five days a week.

One NBER paper from last year estimates that mean distance between where workers live and their employer’s worksite has increased by 70% from 2019-2024. 

Data from the Metropolitan Transit Authority and Connecticut Department of Transportation, and U.S. Census location data can also inform these discussions and give us a better sense of the true nature of commutes.

Before we start spending big on building bigger highways, we should be sure the data is not leading us astray. 

---

About the author: Dustin Nord is the director of the CBIA Foundation for Economic Growth & Opportunity.