A new study suggests that in hard economic times, people drink more alcohol. Intuitively, it might make sense, but the findings run counter to most previous research, which shows that alcoholism and other drinking-related problems tend to decline along with the economy.
For the new study, researchers at University of Miami looked at the connections between risky drinking and state unemployment rates in a sample of more than 43,000 people surveyed between 2001 and 2005. That was before the current economic crisis, but because the health of individual states’ economies varied during that time, researchers could make comparisons between them.
The study found that each 1% increase in the state unemployment rate corresponded to a nearly 17% increase in cases of alcoholism or alcohol abuse and a 35% increase in rates of drunk driving.
“We are one of the first to show that, even though incomes decline for most people during an economic downtown, they still increase problematic or risky drinking,” lead author Michael French, director of the Health Economics Research Group at the University of Miami, said in a statement.
MORE: How We Cope: What Do Addiction Rates After 9/11 Tell Us?
Most previous studies have found that when income takes a dive, so does drinking — alcohol is expensive, so people are less able to afford it when they lose their jobs. A large number of studies have also found that increasing the price of alcohol leads to reduced consumption, even among heavy drinkers.
“It is clear just from examining the per-capita consumption trends that during the Great Depression of 1929 and after, in country after country in conditions of a major depression, the alcohol consumption dropped radically,” says Robin Room, chair of social research in alcohol at the University of Melbourne in Australia, a long time researcher in this area.
Room cites a study done in Finland, which suffered a major economic contraction after the breakup of the Soviet Union. During good economic times, deaths related to alcohol increased, but during the recession, they decreased slightly. Interestingly, the poorest people had the greatest increase in drinking-related mortality during good times — and the smallest decline in bad times.
Overall, the existing research suggests a complicated relationship between unemployment and alcohol and other drug problems. For example, the National Household Survey on Drug Use and Health consistently finds that rates of substance misuse and addiction in unemployed people are nearly twice as high as in those who have jobs. Unemployment is also linked with increased odds of relapse for people with addictions, while employment is linked with recovery.
MORE: Bullish or Bearish on Prop 19? Pegging Marijuana Legalization to Market Indicators
However, the direction of cause in these cases is not always clear: problems associated with addiction can cause unemployment, just as the free time that comes with unemployment can increase risk for addictions (about half of people with addictions are employed).
The conflicts in the research may also relate to individual differences between drinkers. Some people surely self-medicate by increasing their drinking when they lose their jobs, while others sober up. Overall, the latter group seems to dominate.
Room calls the new study “interesting and competently done,” but says, “In my view, the results they find are not likely to be in the same direction as what would be found in a severe economic downturn,” which is an unfortunately apt description of what we’re continuing to experience right now.
MORE: Mind Reading: Steven Pinker’s Case for Why the World Is Heading Toward Peace
Maia Szalavitz is a health writer at TIME.com. Find her on Twitter at @maiasz. You can also continue the discussion on TIME Healthland’s Facebook page and on Twitter at @TIMEHealthland.