Abstract: Heterogeneous Effects of Savings on Participation in the Gig Economy (Society for Social Work and Research 23rd Annual Conference - Ending Gender Based, Family and Community Violence)

Schedule:

Friday, January 18, 2019: 3:30 PM

Union Square 19 Tower 3, 4th Floor (Hilton San Francisco)

* noted as presenting author

Stephen Roll, PhD, Research Assistant Professor, Washington University in Saint Louis, St Louis, MO

Sam Bufe, MS, Statistical Data Analyst, Washington University in Saint Louis, St. Louis, MO

Olga Kondratjeva, PhD, Postdoctoral Research Associate, Washington University in St. Louis, St. Louis, MO

Michal Grinstein-Weiss, PhD, Professor, Director, Envolve Center for Health Behavior Change, Associate Director, Center for Social Development, Washington University in Saint Louis, St. Louis, MO

Background:

Nearly one tenth of Americans earn money through the through online and mobile platforms made possible by the gig economy (Smith, 2016). The gig economy offers workers flexible hours and has relatively low barriers to entry and exit (Dokko and Schanzenbach, 2015). As the labor market continues to change, there is a growing need to understand the needs of this new kind of worker.

This paper seeks to identify how LMI households’ access to liquidity affects their participation in the gig economy. On one hand, access to liquidity may assure some workers that their necessities will be covered, reducing the incentive to work in the gig economy. Alternatively, access to liquidity may give some liquidity-constrained workers the ability to cover some up-front costs of participation, allowing them to work in the gig economy. This paper tests the effect of access to liquidity on different segments of the population.

Methods:

This paper uses data from the 2017 Household Financial Survey (HFS), which is offered to users of TurboTax Freedom Edition and asks a variety of questions about respondents’ financial circumstances and characteristics. The first wave was administered immediately after tax-filing, with the second wave administered six months later. This survey data was merged with individual-level administrative tax data.

When filing their taxes, survey takers participated in a large-scale randomized controlled-trial, which tested the effectiveness of four behavioral interventions designed to encourage filers to save their tax refunds. The random assignment of tax filers to these conditions provides a source of exogenous variation in tax refund savings rates, allowing us to use an instrumental variable design that identifies the effect of savings on participation in the gig economy over the six months after filing.

Results:

While we find no general effect of savings on gig economy participation for our sample as whole, we find strong heterogeneous effects of savings on gig participation for different subsets of our sample. For traditional students, refund savings reduced the likelihood of working in the gig economy by 31%. For working-age people, however, saving a tax refund increased the likelihood of working in the gig economy by 21%. Although we find refund savings had no effect on working-age households with access to liquidity, we find it increases the likelihood of working in the gig economy by over 50% for liquidity-constrained households. We also report evidence suggesting that respondents in our sample are using their refunds primarily to cover the fixed costs of entry into the gig economy.

Implications:

These findings point to the importance of liquid assets for LMI workers. Although the gig economy may have relatively low barriers to entry, the paper’s findings suggest that these barriers may still be preventing liquidity-constrained population segments from participation. These findings have important implications for policymakers trying to address these workers’ needs and ensure equal access to jobs in the changing economy.