John V. Kane, Clinical Associate Professor at the NYU SPS Center for Global Affairs, joined Verasight to talk about "The Effects of Inflation Information: Rates, Prices, and Politics in the Public Mind," his new working paper with Christopher Wlezien of the University of Texas at Austin. Three of the paper's four studies were fielded on the Verasight panel. In this article, we highlight the key points from the interview and discuss next steps for this area of research.
Inflation conflation
"The idea of inflation conflation, which we talk about in the paper, is you are conflating inflation rates as though they represent levels of prices, rather than changes in prices... The first [approach] was just asking citizens directly using a lot of Verasight data... If inflation is 3% one year and then 3% the next year, how would you describe this?... Over half of the sample answered that that would mean that prices stayed the same, which is not correct. Prices increased... In another one, we asked, what if the rate decreased? What if it went from, say, 4% to 3%?... The correct answer, again, is prices increased. The inflation rate is still positive. Prices are still increasing, just at a slower rate than before. Yet again, over half the sample said that that would mean that prices decreased. Again, conflating rates with prices."
An inflation rate measures how fast prices are changing. A steady or lower rate that is still positive means prices are still going up. In the paper's second study, respondents were told the rate was 3% one year and stayed at 3% the next. A majority, 53.9%, said that meant prices had stayed the same, and only 41% answered correctly that prices had increased. In the first study, respondents were told the rate had decreased from 5% last month to 4% this month, and 50.4% of the weighted sample said that meant prices had gone down. The 4% to 3% example Kane uses in the interview describes the same kind of question.
The same information, shown two ways
"We're showing people the same exact information presented two ways: one as a trend in rates, and another, if you just mathematically take those rates and change it to prices, what that trend would look like... The change in effect, when you change from rates to prices, was just enormous. I think they may have been the biggest effects I've ever seen... The staggering size of some of those effects really, really impressed me and my co-author both."
In the third study, 2,032 Verasight respondents saw one economic trend twice, once as inflation rates and once converted into prices. When the rate held flat at 2%, over 90% said prices were increasing when they saw the trend as prices, compared with 23% when they saw it as rates. The format also changed how people judged the economy and the president. A slowing rate of inflation was associated with better evaluations when it was shown as rates, while the same trend shown as prices went with noticeably worse ones.
Prices almost never go down
"We asked respondents... out of 50 years, how many years do you think saw a price decrease? The right answer is one. One year out of 50 saw a price decrease, and even then it was a pretty incredible year. It was 2009, the Great Recession. And even then the inflation rate was just negative 0.4%... which basically means every year prices are going to go up. So just talking about prices going up is not very informative. We need to be thinking about rates also."
According to the U.S. Bureau of Labor Statistics, 2009 is the only year in the past 50 with a negative annual inflation rate. About 45% of respondents chose the correct range of 1 to 5 years, 33% said prices had not fallen in any year, and 22% thought falling prices were more common than they are. Because prices almost always rise, a report that prices went up says little on its own. The rate is what shows how fast they are rising.
Next steps
Asked what comes next in his research, Kane pointed to whether this confusion can be corrected.
"If we show citizens how rates relate to prices, will that get rid of some of these huge disparities and effects that we're seeing? Can this be taught? Can information be given that would actually help citizens navigate this and not get tricked when an elite chooses to mention one over the other?"
He also noted that people respond to national conditions, not only to what they pay at the checkout line or the gas pump. The experiments showed respondents information about the country as a whole and still moved their evaluations. Kane compared this to the unemployment rate, which voters weigh even when they are comfortably employed or retired.
