Research groups
Keep televised prize formats separate
Triss winners grouped by draw year and prize plan
- Lump-sum participants
- 3,260 participants grouped into 18 comparison cells.
- Monthly annuity participants
- 569 participants grouped into 19 cells; value payments at a 2% annual discount rate.
Illustrative summary of the source-described TV-show matching. Exclude shared tickets and multiple prizes under the same plan in one year. Prize amounts are random conditional on qualifying and the prize plan; formats are not sequential stages.
See source →How
When to use
Researchers need to isolate randomized wealth shocks among scratch-ticket lottery winners who qualify for televised prize draws. Selecting players who qualify for a TV show introduces selection bias relative to the general population. Grouping show participants into cells defined by draw year and prize plan isolates exogenous variation in the final ticket draw.
Group televised scratch-ticket lottery winners into matching cells based on draw year, payment format, and prize structure.
- Identify all participants in televised prize draws during a specific calendar year.
- Separate participants by payment format into lump-sum prize draws or monthly annuity installment draws.
- Convert monthly annuity installments into net present value using a two percent annual discount rate.
- Exclude participants who share ticket ownership or win multiple prizes under the same prize plan in one year.
- Assign participants who win one prize under the same prize plan and year to a single matching cell.
- Control for cell fixed effects in panel regressions to estimate labor supply adjustments across prize amounts.
Example
In the Swedish Triss lottery between 1994 and 2010, researchers grouped 3,260 lump-sum show participants into 18 cells and 569 monthly annuity participants into 19 cells based on draw year and prize plan.
About the study
Participants
3,829 Swedish TV show scratch-ticket participants aged 21 to 64 tracked between 1994 and 2010.
Study Design
Panel regression controlling for year-by-prize-plan cell fixed effects to estimate labor supply responses across lump-sum and annuity payout formats.
Results
Dynamic labor earnings responses over ten years are statistically indistinguishable between lump-sum winners and monthly annuity winners, with pooled pre-tax earnings falling by 1.066 SEK per 100 SEK won.