Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Back to History Current Version
Testing for Labor Rationing: Revealed Preference Estimates from Demand Shocks
Last registered on April 16, 2018


Trial Information
General Information
Testing for Labor Rationing: Revealed Preference Estimates from Demand Shocks
Initial registration date
April 13, 2018
Last updated
April 16, 2018 2:28 PM EDT
Primary Investigator
University of Pittsburgh
Other Primary Investigator(s)
PI Affiliation
University of California, Berkeley
PI Affiliation
Harvard University
Additional Trial Information
On going
Start date
End date
Secondary IDs
In this project, we develop a novel revealed preference approach to test for and quantify labor rationing. In addition, we provide evidence on the two components of rationing: involuntary unemployment and disguised unemployment (also referred to as "forced entrepreneurship"). The setting for our test is rural labor markets in Odisha, India, which mirrors other poor rural settings: seasonality from subsistence agriculture leading to low employment levels for much of the year.
External Link(s)
Registration Citation
Breza, Emily, Supreet Kaur and Yogita Shamdasani. 2018. "Testing for Labor Rationing: Revealed Preference Estimates from Demand Shocks." AEA RCT Registry. April 16. https://doi.org/10.1257/rct.2743-1.0.
Former Citation
Breza, Emily et al. 2018. "Testing for Labor Rationing: Revealed Preference Estimates from Demand Shocks." AEA RCT Registry. April 16. http://www.socialscienceregistry.org/trials/2743/history/28337.
Experimental Details
We exploit an opportunity to recruit workers for month-long employment in factories. We use this to generate transitory aggregate demand shocks in (random) villages-absorbing up to 30% of the labor force of casual male workers. To test for rationing, we examine how this external demand shock affects wages and employment in the local village labor market.

If the amount of labor rationing is weakly greater than the size of the external demand shock, then we predict the demand shock will have:
i. no effect on the local village wage
ii. no effect on the level of local employment in the village
iii. positive employment spillovers-higher individual employment and lower reported involuntary unemployment-on the laborers who remain in the village (i.e. who are not removed)

Examining who benefits from employment spillovers can be used to bound the level of involuntary unemployment and disguised unemployment. Further, we will compare our revealed preference estimates with a series of traditional self-reported survey-based measures of unemployment.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Wages (cash and in-kind), employment (activity, location, hours worked), self-reported measures of involuntary unemployment
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Our study design conceptualizes the village as a distinct local labor market. We will select villages into our sample in which between 40 and 100 households participate in the casual daily labor market. A randomly selected 50% of villages will be treatment villages, from which we will hire up to 30% of prime age, male casual daily laborers. This will constitute a large demand shock to the village that will last for one month. The remaining 50% will be control villages, from which we will hire only one or two individuals. This will constitute a negligible shock. To test our predictions, we will examine the impact of the labor demand shocks on village wages and employment.

In each of the control and treatment villages, we will survey a representative sample of all workers who supply labor to the casual daily labor market. We will conduct three waves of surveys: at baseline (immediately before the demand shock), at endline (during the third week of the month-long demand shock), and at post-intervention (two weeks after the end of the demand shock, after all workers are back in the village labor force). In each survey, we will collect detailed daily recall data about wages: cash wages, details of in-kind payments (e.g. tea, whether meat was included in lunch, cash value of in-kind payments), whether the worker was paid on time, etc. In addition, we will collect details of employment (activity, length of breaks, hours worked, location) and self-reports of involuntary unemployment. We will also collect information about baseline characteristics of household enterprises (such as asset use, investment, and profits). Further, we will survey a small sample of agricultural employers in each village at endline (during the third week of the demand shock). This survey will capture data on the hiring of workers from inside and outside the village, employers' search efforts for workers, and farm operations. We will ask employers to rate the quality of a subset of workers in the village.
Experimental Design Details
Not available
Randomization Method
Randomization of villages and workers is done on a computer.
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
60-80 villages
Sample size: planned number of observations
60-80 villages
Sample size (or number of clusters) by treatment arms
30-40 treatment villages, 30-40 control villages
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
Columbia University
IRB Approval Date
IRB Approval Number