The graph is loosely based on an article by Yascha Mounk and Roberto Stefan Foa in the Journal of Democracy. As Ronald Inglehart, Cas Mudde and Jeff Guo have pointed out, the graph’s drama largely flows from a selective presentation of data. Here’s how.
The data for the graph are from the fifth wave of the World Values Survey (WVS), which asked people to place themselves on a 10-point scale where 1 meant that living in a democracy is “not at all important” and 10 “absolutely important.”
So where does this graph go wrong? It plots the percentage of people who answer 10, and it treats everyone else the same. The graph treats the people who place themselves at 1 as having the same commitment to democracy as those who answer 9. In reality, almost no one (less than 1 percent) said that democracy is “not at all important.”
The graph below uses the exact same data, but it plots the average scores rather than the percentages who place themselves at the top end of the scale (see my earlier tweeted version here).
Younger people answer this question differently than older people — although the small sample size of people who were born in the 1990s requires some caution here. For instance, there were only eight people in that category in the Britain sample. Nevertheless, it’s clear that younger people do, on average, assign slightly lower values on the 10-point scale.
But that doesn’t mean that young people are done with democracy. In most countries, people born in the 1980s or 1990s still assign a great deal of importance to living in a democracy.
There’s another issue: We don’t know whether millennials are actually a distinctive generation. Perhaps young people have always been more diverse in their assessments of democracy. If so, then it’s hard to conclude that some new danger is afoot.
Unfortunately, the WVS did not ask the exact same question in previous versions of the survey. In the mid-1990s, it asked whether people agreed (on a 5-point scale) that “democracy may have problems but is better than any other form of government.” The graph below is what the New York Times graph would have looked like in 1995. It shows the percentage of people who “strongly agree” that democracy is the best.
Only about 40 percent of Americans born in the 70s strongly agree that democracy is the best form of government! Apparently, the country’s democratic tradition has been under duress for two decades! Australia and New Zealand are doomed altogether. (There is no data for the Netherlands and Britain.)
Below is the graph based on averages. The vast majority of people who didn’t answer “strongly agree” instead said “agree.” This graph has a lot less drama and is a lot more accurate.
The article by Mounk and Foa does document some small shifts in opinion on related issues. But these aren’t nearly as dramatic as the New York Times graph suggests.
Vast majorities of younger people in the West still attach great importance to living in a democracy.