The New Economic Geography of WFH

Matthew E Kahn

Over the last three years, companies from all over the world have learned valuable information about how their firm’s productivity and worker satisfaction is affected when workers can engage in Work from Home (WFH) on at least a part-time basis. Each firm faces fundamental tradeoffs in not requiring workers to return full time to the office.

On the one hand, WFH accommodates worker lifestyles and responsibilities at home.

WFH workers gain from commuting less often and some may choose to live further from their place of work because they commute in on fewer days. Firms with happier workers are less likely to face retention challenges and will pay less retraining workers. Given that WFH is a non-taxed perk, firms that offer this job amenity can be less generous in giving workers pay raises. Firms located in expensive areas can economize on expensive commercial real estate. If a firm requires that workers be in the office just half the week, then a firm can reduce its space requirements by 50%!

On the other hand, firms benefit from face to face communication. Younger workers gain more from such mentoring. Building up the firm’s culture and monitoring worker effort is easier to do if workers are on site.  

Each company has a strong incentive to experiment with different workplace practices to learn what works best for their organization. The pursuit of profit actually accelerates this learning process. Economists are always looking for new ways to study firm dynamics. The rise of the WFH economy opens up many questions related to what types of firms will encourage their workers to engage in WFH. Famous executives such as Elon Musk at Tesla and Jamie Dimon at JP Morgan have been encouraging workers to return to the office. Why?  

Up until now, a lack of data has hindered the systematic analysis of what types of firms prioritize allowing workers to engage in WFH on at least a part-time basis. The newly released Flex Index provides a unique opportunity for researchers.

As an urban economist, I am especially interested in the geographic determinants of why different firms engage in hybrid-WFH. The Flex Index provides data on each firm’s headquarters’ location. I focus on those that are based in the United States. I use data for 2528 of them. For this set of firms, I merged in three geographic variables.

The first variable is the Zillow Median Home Price Index. Zillow uses a repeat home price index methodology to create a standardized “apples to apples” measure of home prices that can be compared across cities at a point in time or within a given city over time. In the analysis I present below I use Zillow data from December 2022.

The second variable I merge in to the Flex Index is the percent of voters in the Headquarters State that voted for Joe Biden in the 2020 Presidential Election.

The third variable I merge in from the Zillow data base is the city’s size. So, New York City is the first ranked city in the nation because it is the largest city. I use these data to test two hypotheses:

**Hypothesis #1: All else equal, firms are more likely to engage in WFH work if home prices are higher in the HQ City.

‍**

**Hypothesis #2: All else equal, firms are more likely to engage in WFH in more Progressive states.

‍**

To test these hypotheses, I use a linear multivariate regression framework. To standardize the firms, I include industry fixed effects (this variable is in the Flex Index database). The dependent variable is a dummy variable that equals one if the firm’s workers all work on site. In the database, 39% of the firms report having the workers work fully on site.

Table One reports four regressions with each column reporting a separate regression. I use the regression output to test the hypotheses presented above.

Main Results

All else equal, firms are more likely to engage WFH work if home prices are higher in the Headquarter (HQ) City.

Based on the results in columns (1) and (3), I find that for the full set of industries and for the Healthcare and Biotechnology industry that workers are more likely to be engaging in WFH if home prices are higher in the HQ’s city. Based on the results in column (2), a doubling of home prices is associated with a log(2)*-.088 = -.06 or a six percentage point reduction in workers working solely on site. This effect is even larger for the Healthcare industry. For this subset of firms, a doubling of local home prices is associated with a 12.7 percentage point increase in WFH. For the other two industries, I fail to reject the hypothesis that firms are not more likely to allow workers to engage in WFH as a function of local home prices.

All else equal, firms are more likely to engage in WFH in more Progressive states.
‍

The test of this hypothesis is based on the coefficient on the 2020 Joe Biden vote share. Across all four specifications presented in Table One, I find consistent evidence that firms are more likely to allow their workers to engage in WFH when the HQ is located in a more liberal state. A ten percentage point increase in Joe Biden’s vote share is associated with a 3 percentage point increase in the probability that a firm allows its workers to engage in WFH on at least a part-time basis.

Conclusion

The Flex Index is a terrific new data source that allows empirical researchers to study how different firms are incorporating WFH into their work routines. In my 2022 Going Remote book, I hypothesized that the rise of WFH would benefit workers in high home price areas. The evidence presented here supports this hypothesis. Here, I also find a Red State/Blue State divide such that firms located in Blue States are more likely to allow their workers to engage in WFH. Future research should explain why we observe this fact. Are such firms in progressive states more concerned about work/life balance and gender equality issues? There are many open questions that improved data access will allow for us to provide credible answers.