More Evidence of Lottery?s Regressivity
August 11, 2006
We have argued (here and here) that the lottery is a regressive form of taxation. Recently we discussed the misunderstanding—or intentional misuse–of the word “regressive” as it applies to studies on lottery tax burdens and zip code income data.
A dramatic illustration of the lottery’s regressivity can be found in a Newark Star-Ledger article on an analysis the paper conducted of five years of lottery data and Census income data by zip code. The authors found that “lottery revenues (ticket sales) rise as income falls,” especially for 3- and 4-digit numbers games and instant games. The study concluded that the lottery “is a regressive form of taxation.” Among the findings:
Per-capita ticket sales were much higher in lower-income ZIP codes. In communities with average household income below $52,000, the lottery sold an average of $250 of tickets per person annually. That was more than double the amount for ZIP codes with $100,000 households.
The study’s authors took into account a common problem with zip code/income studies: areas with large “daytime populations”—that is, people who work in an area but do not reside there—will show high ticket sales that do not accurately reflect residents’ purchases. The authors used Census Bureau data to account for the extra daytime purchases and found that the results remained the same: in zip codes with lower per capita incomes, more tickets were purchased:
Take the towns of Hillsborough and Belleville.
The two have roughly the same population, about 36,000, and similar daytime populations that include people who work but don’t live in town, about 42,000.
When it comes to sales of lottery tickets, however, there’s no comparison. Average annual sales at stores in Hillsborough, a wealthy Somerset County town, are about $1.8 million. In Belleville, a blue-collar town outside Newark in Essex County, it’s $13.3 million.
A lottery spokesperson responding to the study clearly did not understand the concept of regressivity:
[A] spokesman for the state Treasury Department, which oversees the lottery, pointed to one question from a 1999 Star-Ledger/ Rutgers-Eagleton Poll that showed lottery participation was about the same regardless of income. But the overall conclusion of that poll was that “the lottery is in fact regressive,” meaning it takes a greater percentage of income from the poor than the wealthy.
A leading lottery expert summed up the study concisely:
“It walks like a tax and talks like a tax,” said Charles Clotfelter, a Duke University economist …. “It just needs to be recognized that the state is putting a greater burden on low-income people.”