Experimental Evidence on the Effect of Employer Outreach on Work Sharing Participation

How to cite

Motghare, Swapnil; Wachter, Till von; Wilson, Eric; Segal, Kara. 2026. "Experimental Evidence on the Effect of Employer Outreach on Work Sharing Participation." AEA RCT Registry. AEARCTR-0017066. https://doi.org/10.1257/rct.17066-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Swapnil Motghare* , Till von Wachter , Eric Wilson , Kara Segal

* Primary investigator

Registry ID
AEARCTR-0017066
DOI
https://doi.org/10.1257/rct.17066-1.0
First registered
Last updated
Status
In development
Trial dates
2026-09-01 – 2029-08-31
Countries
Private
JEL codes
J65, C93
Registry record
http://www.socialscienceregistry.org/trials/17066

Abstract

The Work Sharing (WS) program is designed to help employers avoid layoffs by allowing them to reduce hours across their workforce during temporary downturns rather than laying off a subset of workers. Despite its potential to stabilize employment, take‑up of WS remains low in the United States: only 30 states operate a WS program, and Short‑Time Compensation claims rarely exceed 1 percent of all Unemployment Insurance claims. This study evaluates whether outreach to employers increases awareness and take‑up of the WS program, and whether increased program participation affects employer and worker labor‑market outcomes. We implement an Encouragement Design Randomized Controlled Trial, in which a random sample of WS‑eligible firms in California (the “treatment group”) will receive outreach about the program, while the remaining firms (the “control group”) will not. The outreach will be delivered using physical mails and emails. We will estimate the causal effect of outreach on WS awareness and take‑up by comparing employers and their workers across treatment and control groups. To estimate the effect of WS take‑up on employer and worker labor‑market outcomes, we use an instrumental‑variables strategy in which randomized outreach serves as an instrument for WS participation. Outcomes are measured using an employer follow‑up survey and linked administrative data on employer‑employee earnings and program participation.

Intervention

The intervention will consist of physical mail (stand-alone and inserts in existing EDD notices) and emails, designed to inform employers in the treatment group about the WS program.

Experimental design

We will implement an Encouragement Design Randomized Controlled Trial, in which a random sample of WS‑eligible firms in California (the “treatment group”) will receive targeted outreach about the program, while the remaining firms (the “control group”) will not. The outreach will be delivered using physical mails and emails.

Design details

Randomization method
Within each stratum, we apply the three‑variable hierarchic serpentine ordering procedure using UI Benefit charge and employment as implicit stratifiers. The stratification and assignment of firms to treatment and control group will be done by a computer program.
Randomization unit
firm
Clusters
We stratify the sample of firms on firm size, industry, and geographic region. Specifically, we define strata based on: A dedicated stratum for firms with 1000+ employees, and 15 Regional Planning Units (RPUs) x 20 NAICS 2‑digit industries, and a separate stratum for firms with missing industry or RPU, which yields 302 strata in total. To avoid strata with few firms, we combine some strata eventually ending with 198 strata.
Observations
Approx 520,000
Arms
Approx. 260,000 firms in treatment and 260,000 firms in control group.
Primary outcomes
A. Firm outcomes: WS take-up, WS awareness, Firm survival, Likelihood of mass layoff, Employee turnover (Number of Employee Separations / Total Employment) B. Worker outcomes: Probability of layoff, Quarterly earnings, Employment
Secondary outcomes
A. Firm outcomes: the proportion with WS claims, the number of workers with WS benefits, the number of Full-Time Equivalent workers on WS, WS benefits paid, UI benefit charges, WS benefits/ UI Benefit charges, Number of employee separations, and Total employment B. Worker outcomes: Job tenure, Probability of WS, Employment with same employer, and Employment in same industry

Cited by

No citation data available for this registration yet.

Keywords