Though I am in fact a Dan Ariely fan.

I happened upon the original paper that was recently reported on in a number of locations, including here.

After reading the survey method, the declared results seem even more dubious. The participants were given three choices of wealth distribution: Perfectly equitable, Sweden's, and the USA's. These weren't marked with these titles, but merely indicated the percentage of wealth controlled by each quintile. People overwhelmingly chose the Swedish distribution, without knowing it.

More reasonable than "everyone wants Swedish wealth distribution" is "people were given two extremes and a moderate option, and they chose the moderate option."

The latter experiment as a little better, in that people could freely offer numbers. However (as previously mentioned) people may not be very good at thinking about the economy in terms of percent of wealth controlled by quintile. It's also not clear if the experiment was done on the same subjects in the same survey, because seeing Sweden's wealth distribution numbers in a context as the moderate distribution could inform the subject's later responses when given a blank to fill in.

Again, I'm not arguing for an inequitable distribution of wealth, just sensible survey methods. Wealth distribution should be more equitable not because everyone agrees to want it, but because the economy and society will be healthier if it is.