This is a linkpost for

This paper was recently shared with me in relation to some of Giving What We Can's messaging, our How Rich Am I calculator, and our member motivations:

"Your Place in the World: Relative Income and Global Inequality" by Dietmar Fehr, Johanna Mollerstrom & Ricardo Perez-Truglia. Abstract:

Although there is abundant evidence on individual preferences for policies that reduce national inequality, there is very little evidence on preferences for policies addressing global inequality. To investigate the latter, we conduct a two-year, face-to-face survey experiment on a representative sample of Germans. We measure how individuals form perceptions of their ranks in the national and global income distributions, and how those perceptions relate to their national and global policy preferences. We find that Germans systematically underestimate their true place in the world’s income distribution, but that correcting those misperceptions does not affect their support for policies related to global inequality.

Hive mind: I'd be interested in your thoughts this research and it's applicability to EA contexts where we use this messaging.

In terms of GWWC we'll often use this messaging in part to de-bias, and also to reduce scarcity mindset (people not giving because they never seem to "find their enough").

[Table 6 that David R mentioned]

Sorted by Click to highlight new comments since:


I think this is the key table is table 6 (link HERE -- can't seem to  put images in comments -- @Luke maybe add it to your post?) 

Aside: However, I’m also curious about the overall effect of the information intervention in *itself*, without considering whether it is a downward or an upwards update. I don't see that presented anywhere!

In the table above  I guess that the row I care most about is the second row, and then

 Columns 2 — support for global redistribution — with a small and positive mean estimated effect and  

Column 5.   actual choice of allocating one’s own money to the poor Kenyan person — estimated near 0.

These are not ‘statistically significant’ but that doesn’t mean they are zero, we just don’t know.  

We may be able to rule out _very large effects_, but  The standard errors in parentheses seem to be moderately large. (ThinkRhink +/- 1 std error) you cannot rule out a substantial positive effect of ’learning you are ranked higher globally … on support for giving/redistributing globally.

There is still a substantial likelihood that updating beliefs about global ranking upwards increases my support fo these by, maybe 10-25% of a standard deviation, which seems substantial to me.

I suppose  for EA giving orgs target audience tend to be left-of-center politically (in the German context? Not sure) .  For that group — the very important column 4  seems to be somewhat positive… and with a pretty large standard error.

Checking on EA Survey data (noting that these are not the GWWC target audience, but just for reference), about 300 of the 400 Germans who respond to the politics question say they are either 'Left' or "Center left" and under 10% identify as right, center, or center-right.

Context and value

I've given it a moderate skim-read, and here are my thoughts (reprising some of our Slack conversation)

It's a very important question, and seems well done -- great design choices in several important ways (the writing could be a bit clearer though, IMO). I always wanted to do a study like this, it is long overdue.

While "what applies to policy attitudes doesn’t necessarily apply to giving(*), and these things may differ based on context" ... this does seem like a fairly relevant context, and it does involve some outcomes that are approximately direct giving. I might provide some cause to update our beliefs against ‘debiasing’ as a powerful motivator here.

(*) And I suspect that the EA, and global health/poverty communities cares at least as much about support for global aid and pro-poor poverties as we do about giving Also, as @davidmoss notes "it's kind of a trope in debiasing research that changing people's beliefs usually doesn't do anything" ... indeed there has been previous work finding similar 'debiasing doesn't matter' in the field of support for international aid iirc. I guess saw this as 'folk wisdom' needing further confirmation ... and here is further confirmation ... perhaps. (I'll comment on the empirics in another thread)

Curated and popular this week
Relevant opportunities