In the face of Jeff Sachs’ challenge, Bob Lawson proposes a bet on economic freedom:

“Dear Tyler,

I read with obvious interest your post (and the paper itself) about
the endogeneity of institutions. Leaving aside my issues with the IV
literature, I decided to take the bait regarding Jeff Sachs’ challenge
to, “Go back to 1960 and choose any measure of institutional quality you
want. Then see how well it predicts cross-national growth since then.”

Ok, I will.

The Economic Freedom of the World (EFW) index was first published in
the mid 1990s, and the first year of data is 1970. So I’ll have to start
in 1970 instead of 1960.

Here is a regression with growth from 1970-2010 on the lhs, and EFW and GDP per capita in 1970 on the rhs.


A one-unit higher EFW score in 1970 correlates to 0.84 percentage
points in higher annual growth over the next 40 years. A one unit EFW
score improvement during the first decade, 1970 to 1980, correlates to a
1.00 percentage point higher annual growth rate over the 40 years.

I don’t know if that satisfies Jeff Sachs’ challenge, but it works for me.

Lawson’s bet:

Looking forward, I’ve constructed a back-of-the-envelope indicator
that combines each country’s EFW rating in 2000 and with its change from
2000-2010. The top 20 (combined highest level & most positive
change) versus the bottom 20 (combine lowest level & most negative
change) countries are:

Top 20 – Bottom 20
Hong Kong – Haiti
Romania – Cameroon
Rwanda – Senegal
Singapore – Guinea-Bissau
Bulgaria – Mali
Cyprus – Bolivia
Unit. Arab Em. – Algeria
Chile – Guyana
Mauritius – Gabon
Lithuania – Ecuador
Slovak Rep – Burundi
Albania – Cote d’Ivoire
Jordan – Chad
Switzerland – Togo
Bahamas – Congo, Rep. Of
Malta – Central Afr. Rep.
Taiwan – Argentina
Korea, South – Myanmar
Finland – Zimbabwe
Estonia – Venezuela

I’m willing to bet anyone $100 (up to 10 people) that the Top 20
group will outgrow the Bottom 20 group by at least 1 full percentage
point per year (on average) over the the next 20 year period

Who will accept the challenge?