NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Sports Betting in Uganda: Causes and Consequences
Last registered on March 21, 2018


Trial Information
General Information
Sports Betting in Uganda: Causes and Consequences
Initial registration date
February 05, 2016
Last updated
March 21, 2018 11:37 AM EDT
Primary Investigator
University of California, Berkeley
Other Primary Investigator(s)
Additional Trial Information
Start date
End date
Secondary IDs
Sports betting has become very popular across many countries in sub-Saharan Africa over the last decade. It is a multi-million dollar industry in many developing countries and continues to expand. This study looks at sports betting behavior among bettors in Kampala, Uganda and seeks to identify causal factors contributing to high betting intensity and prevalence. We will employ randomized treatments to identify possible behavioral biases contributing to high betting intensity. We will also employ quasi-experimental methods to analyze additional causes as well as the impacts of betting participation that cannot be explored experimentally.
External Link(s)
Registration Citation
Herskowitz, Sylvan. 2018. "Sports Betting in Uganda: Causes and Consequences." AEA RCT Registry. March 21. https://doi.org/10.1257/rct.1025-4.0.
Former Citation
Herskowitz, Sylvan. 2018. "Sports Betting in Uganda: Causes and Consequences." AEA RCT Registry. March 21. http://www.socialscienceregistry.org/trials/1025/history/26980.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Two randomized interventions will be implemented. The first will target ex-ante mental accounting of sports betting expenditures. The second will provide information on betting performance to randomly selected study participants.

In addition, two randomized primes will be embedded in the endline survey followed by a measurement of betting demand. The first prime will help to identify if betting demand is driven by demand for liquidity in pursuit of lumpy goods. The second will isolate whether betting demand is driven by hot states tied to sports excitement.

In the final round of data collection, an additional randomized prime was included that centered on budgeting for saving. Randomized budgeting exercises were conducted so that the effect of positive or negative updated perceptions of saving ability on betting demand could be measured.

Quasi-experimental methods will be utilized to examine what expenses betting crowds out among bettors as well as how wins are used. These, methods will exploit naturally occurring variation in the international football schedule as well as baseline dimensions of heterogeneity among the study population.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Betting expenditures, targeted betting payouts, household and personal consumption, investment, savings, and transfers.
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
This study will conduct a listing of young men between 18-40 working with small enterprises in randomly selected communities around Kampala. From this listing 800 participants will be invited into the study. The study will include five rounds of in-person visits and four phone check-ins on weeks between visits. Field team members will implement the randomized treatments during the course of their visits.
Experimental Design Details
See attached document.
Randomization Method
In office by computer.
Randomization Unit
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Budget improvements lead this to increase to a total of 2,774

Sample size: planned number of observations
An increased budget allowed this number to increase to 1,004 for the full study. An additional 1,770 were included in an add-on single visit study.
Sample size (or number of clusters) by treatment arms
For full treatments there will be 400 treatment and 400 control.

For each of the experimental primes there will be approximately 200 treatment and 200 control.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The primary outcome of interest for the two randomized is reported expenditure on betting. Data from earlier piloting showed that the mean weekly expenditure on betting among respondents was 2 USD while the observed standard deviation was 3USD. For these treatments 50% of subjects are assigned to treatment and 50% assigned to control. With power of 0.9 and a false detection rate of 5% (alpha = 0.05), the minimum detectable effect (MDE) size will depend on how much other covariates or fixed effects are able to explain variation of betting intensity. If, conservatively, they only explain 10% of variation in reported betting expenditures the MDE will equal 0.235 standard deviations or $0.70. If we are able to explain up to 50% of variation in betting, the MDE will fall to 0.175 standard deviations or $0.53 in weekly betting expenditures. For the prime experiments the primary outcome variable of interest will be the revealed measure of demand for betting recorded during the endline. Participants will be allowed to “purchase” up to four betting tickets, or can keep cash, or any mix of cash and betting. These primes will divide the full study population into smaller groups with approximately 400 respondents for treatment and control (200 each). If mean number of desired betting tickets is 2 and standard deviation is 2 then the range of MDE standard deviations will be 0.247-0.332 or mean difference of 0.495-0.664 betting tickets demanded.
Supporting Documents and Materials

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
IRB Name
UC Berkeley Institutional Review Board
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Intervention Completion Date
July 10, 2016, 12:00 AM +00:00
Is data collection complete?
Data Collection Completion Date
July 14, 2016, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
2,774 respondents
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
1,004 In full sample
1,770 In Single-visit study
Final Sample Size (or Number of Clusters) by Treatment Arms
Note: Not all respondents are included in all treatment samples Savings Interventions: Saving Box = 370, Bank Account = 120, Controls = 514 Wallet Intervention: Wallet = 400, Controls = 604 Betting Info: Info = 576, Controls = 382 Lumpy Good Prime: Prime = 658, Control = 625, Both = 520 Budgeting Exercise Prime: Treat= 203, Control = 495
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)