China will move increasingly toward political freedom if it continues its successful movement for economic freedom. — Milton Friedman in 2003.
Milton Friedman was a giant of economics, but, I think we can say he was mistaken about China, see this 2005 article from Foreign Affairs by Bruce Bueno de Mesquita and George W. Downs, from the article:
RICHER BUT NOT FREER
“Ever since Deng Xiaoping opened up China’s economy more than 25 years ago, inaugurating an era of blistering growth, many in the West have assumed that political reform would follow. Economic liberalization, it was predicted, would lead to political liberalization and, eventually, democracy.”
I just finished the reading of recently accepted research in the Journal of Comparative Economics:
Colin O’Reilly, Ryan H. Murphy, Freedom from unit roots? The time series properties of democracy and economic freedom, Journal of Comparative Economics,
2025, https://doi.org/10.1016/j.jce.2025.02.003.
In this research, O’Reilly and Murphy (hereafter OM25) challenges Sobel and Coyne (2011) who find that political institutions and economic freedom are non-stationary and cointegrated.
The central results in OM25 is that: “for many countries, institutions should be thought of as stationary or mean reverting”.
Let me recall that the non-rejection of the unit root hypothesis in a unit root test indicates that the series have not a finite variance and the shocks have permanent effects. In macroeconomics and monetary economics, the question of determining the nature of the time series (stationary fluctuations around a deterministic trend or as non-stationary processes that have no tendency to return to a deterministic path) is very old and goes back to Nelson and Plosser (1982):
Nelson, C. R., & Plosser, C. R. (1982). Trends and random walks in macroeconmic time series: some evidence and implications. Journal of monetary economics, 10(2), 139-162.
In the context of economic freedom and democracy, the question is a bit difficult to tackle because long times series data are not available from economic freedom.
So, OM25 use panel datasets for a large cross-section of countries coming from the following sources:
- V-dem: for the Electoral Democracy, Deliberative Democracy;
- Democracy and autocracy: Polity 5 from the Integrated Network for Societal Conflict Research (INSCR);
- Economic Freedom: Fraser Institute;
- Globalisation Index: KOF Swiss Economic Institute;
- Political and Civil Liberties: Freedom House.
In the following, I will reproduce a few selected abstract. The main contribution and methodologicam contribution:
On the theoretical side:
Here, the literature implies some form of relation between economic freedom and democracy
Some data and methodology considerations:
How to detect unit roots in short T panels?
A conclusion:
While it is very true that institutions are sticky, one big limitation of this study is that they do not capture the transition of institutions because of the length of the time series. Indeed, as shown by Paldam (2021), the political institutions undergo transitions and for the group of more advanced economies, we observe the end of the transition.
Paldam, M. (2021). The grand pattern of development and the transition of institutions. Cambridge University Press.
See the following blog to know how to estimate the transition of the Polity and Polyarchy scores:
So, it is not so suprising to find that most of these countries have mean reversions (rejecting the null of unit root), because we observe a short span of data where some countries have achieve their political transition and some other are still in the authoritarian steady state. That being said, the countries that do not have mean reversion in their political institutions have to be carefully observed in future studies and the question of how economic freedom could help to achieve the democratic transition is very fertile field.