The sports betting industry has put forth a report challenging the government's assertions that betting contributes to higher household debt and defaults in Brazil. This report, commissioned by the National Association of Games and Lotteries (ANJL), aims to highlight flaws in a study conducted by Brazil's National Trade Confederation (CNC), which linked a decline in household finances to betting activities.
According to details from Radar Econômico, a column in the Brazilian magazine Veja, the main criticism in the ANJL's report is compelling. The CNC's analysis employs a "differences-in-differences" approach to compare periods before and after January 2023, yet it lacks a control group. As a result, it fails to present a credible counterfactual scenario that would illustrate debt levels unaffected by the developments in the betting industry.
Merely observing a timing coincidence does not implicate bookmakers as the cause. Consequently, other economic factors such as interest rates, inflation, credit availability, employment conditions, and income transfer policies may have impacted the observed changes, rather than betting alone. With a limited dataset of just 59 aggregated monthly observations, the study can suggest a temporal correlation but does not convincingly argue that betting has been the catalyst for altered debt levels.
The ANJL-commissioned report effectively dismantles the causality claim made by the CNC. However, it also presents a notable limitation: disproving the CNC's assertion does not confirm that no impact from betting exists. The ANJL document does not calculate the estimates anew or provide an alternative dataset to independently assess the platforms' effects on household budgets.
A review by Radar Econômico further highlighted inconsistencies in the CNC's analysis regarding coefficients and statistical significance indicators. For example, a coefficient of -0.305 with a standard error of 0.348 was inaccurately labeled with three stars, typically indicating a significance level of 1%. Reality suggests that the values published do not support such a classification.
This inconsistency raises questions about the integrity of the CNC's findings while also exposing weaknesses in the ANJL’s report, as it reproduced some statistical markers without proper recalibration. Notably, the ANJL report did not audit the CNC's dataset and lacks econometric replication or a statistical appendix.
While the report includes regulatory recommendations concerning the advertising of authorized betting platforms and warns that stricter regulations might bolster the illegal market, these points arise more from industry studies than from empirical data.
This ongoing dispute remains unresolved. The industry has successfully identified a significant flaw in the CNC's primary study, yet the broader implications drawn from this finding may exceed what the data substantiates. The ANJL report refutes the claim that the CNC has established a direct link between betting and increased indebtedness but stops short of exonerating betting activities from potentially contributing to such impacts.
