The primary data source for this analysis is the one-year sample of the American Community Survey (ACS) analyzed by Public Use Microdata Survey Area (PUMA). This is the only source that matches a worker’s residence to their place of work, and provides detail about worker characteristics from the ACS annual survey. However, the dramatic increase in remote work since 2019 skews the data in ways that make straightforward comparisons between pre- and post-pandemic years highly misleading. Because survey respondents are asked to report the physical location where they worked at the time of the survey, people who reported that they worked primarily from home are recorded as having worked at their place of residence, even if their employer is located in a different city, county, or state. This resulted in a sharp drop in the reported number and aggregate wages for people who worked in New York City but resided outside of the five boroughs, as shown in ACS data. This data is important for understanding the transportation impacts of remote and hybrid work, since the number of people physically commuting into the city each day did indeed cause travel to work to decline substantially. However, it significantly understates how much city employers continued to depend on workers who lived outside the city, and how much non-New York City local economies relied on income earned from city employers.

To develop more realistic comparisons of pre- and post-pandemic flows, RPA estimated 2022 workers and wages for all who work in New York City and reside elsewhere using a combination of ACS data and current employment data from the Bureau of Labor Statistics (BLS). The results were compared to earnings data from the Bureau of Economic Analysis (BEA), which provides a more accurate measure of the total inflow and outflow of earnings from individual counties using tax records and other data to supplement Census journey-to-work data, but does not show flows between pairs of counties.

Finally, local jobs and wages created by the spending of commuters and remote workers were estimated using RIMS II multipliers developed by the BEA.

Pre-pandemic Commutation and Wages

2019 was the last full-year prior to the start of the pandemic in early 2020. The tables below show the flow of workers, wages and salaries from ACS PUMA data that matches residents with their place of work.

The ACS data used for these computations likely understates both the number and total wages of New York City workers residing elsewhere in the region for two reasons. Two percent of the total number of people who worked within the region were not assigned a place of work. Some portion of these are presumably commuters to New York City. In addition, 5% of workers living outside of New York City reported that they were working from home. Some share of these workers likely worked for employers located in New York City.

Post-Pandemic Estimates

Comparable data for the current year is unavailable. The most recent year for ACS PUMA data is 2021. Using 2021 as a comparison to 2019 would miss the bulk of New York City’s job recovery. By 2022, however, the city had regained all but 112,000 jobs on an annual basis.

More importantly, the ACS is only able to track where people report working physically, not the location of their employer. As a result of the dramatic increase in working from home, ACS data shows a much larger decline in New York City jobs between 2019 and 2021 (665,000) than BLS (406,000). It indicates the number of New York City workers residing outside the city declined from over 1 million to 653,000, while the number of non-New York City residents working from home surged from 370,000 in 2019 to 1,530,000 in 2021. Many of these remote and hybrid workers worked for employers in the same county or subregion, but a large portion worked for employers in New York City.

Clearly, a large share of the 1.5 million working from home were employed by businesses located in New York City. Other indicators point to an increase in the share of New York City workers who reside in the suburbs, whether they travel into the office each day, are fully remote, or work a hybrid schedule. Industries with a high share of commuters (professional services, information, finance) have grown past their 2019 peak while many industries with low commuter shares (retail, accommodation, personal services) are still well below their pre-pandemic peak. Population and housing data indicate net migration from the city to the suburbs since the pandemic, particularly for office and other workers with the ability to work from home.

Most convincingly, BEA personal income data shows an increase in the net outflow of earnings from Manhattan to those living outside the county from $225 billion in 2019 to $253 billion in 2021. For the five boroughs of New York City, the net outflow increased from $162 billion to $185 billion. This would be consistent with either a small decline or a modest increase in the share of NYC’s workforce residing outside the city, but not with the large decline shown in the ACS data. While BEA data does not show the origin and destination of workers, it uses IRS and other data to supplement Census data to obtain a more accurate picture of earnings flows than the ACS.

To estimate how the flow of workers and earnings within the region changed between 2019 and 2022, RPA used the following assumptions:

1. 2019 ACS employment by place of work was changed by the same 2022/2019 percentage as the BLS Current Employment Series (CES) for major industries.

2. The residence share of each industry sector was held constant between 2019 and 2022. For example, 20.7% of New York City workers in finance and insurance lived in New Jersey in 2019, so the same 20.7% share was applied to finance and insurance in 2022. While there has inevitably been some change in these shares since 2019, they are unlikely to have shifted dramatically. In most cases, the commuter share is more likely to have increased than decreased because of population shifts to the suburbs.

3. The average wage for each industry was changed by the same 2022/2019 percentage as the BLS Quarterly Census of Employment and Wages (QCEW) series, adjusted for inflation using the CPI for all urban consumers for the New York-Newark-Jersey City MSA.

While these estimates have some range of error, they can be considered conservative estimates. Their biggest weakness is the lack of up-to-date commuter shares by industry. For the reasons stated above, the assumptions that were used are likely to err on the conservative side. The 2019 ACS data used in the analysis likely understates the number and aggregate wages of commuters since 7% of the base data could not be used either because a place of work was not available, or because the employer location for respondents who reported that they work from home is unknown. While the estimates assume that the distribution of New York City workers by place of residence within industries remained the same in 2022 as it was in 2019, it is more likely that the share of workers residing outside of New York City increased slightly over those three years.

Local Impacts of Commuter and Remote Worker Wages

To estimate the job and earnings impacts of residents who earned their salaries from employers located in NYC, RIMS II multipliers for detailed industries were obtained from the Bureau of Economic Analysis for the five subregions used in the analysis. To calculate the induced impacts (i.e., those supported by the direct earnings of workers), the Type I impacts for each industry within each subregion were subtracted from the Type II impacts. The results are shown in Section 4.

It is worth noting that the BEA multipliers in this analysis may be less accurate than they were in pre-pandemic years. Measuring the consumption patterns of commuters, particularly defining the area in which workers spend most of their earnings, was always challenging. With increasing remote and hybrid work, these patterns are changing. Presumably, the more time a person works from home, the more of their earnings are spent near their place of residence. The basket of goods that they purchase is also changing, with more spent on residential space and home office equipment and less on commuting costs and business attire. In addition, the profile of a fully remote worker is likely to be different from that of a hybrid worker who still commutes to the office 2-3 days per week.