1. Comments on: “Deep integration an Economic Growth: Counterfactual Evidence from Europe” (Nauro F. Campos and Fabrizio Coricelli and Luigi Moretti) By Neven Mates Standard disclaimer applies, represents only my views Dubrovnik June 13, 2016

2. Counterfactual What if… We cannot know, this is counterfactual… Example: What would be with X had Y not hapend? But this paper suggests that answering this question is easy. It uses “Synthetic counterfactuals”.

3. Synthetic counterfactuals (1)

4. Synthetic counterfactuals (2)

5. Synthetic counterfactuals (3)

6. Levine (1992): How robust are various variables in projecting GDP growth over longer time? Almost all results are fragile… …except for investment to GDP ratio. And to a lesser extent, the initial income level. Projecting GDP growth rates over longer periods is not easy (1)

7. Take a similar country… Which hardly exist… …or construct a new one. Projecting GDP growth rates over longer periods is not easy (2)

8. 1.The pre-treatment (year by year) GDP per capita (PPP Converted at 2005 constant prices, rgdpch) 2.The pre-treatment average of the investment share of per capita GDP (PPP Converted at 2005 constant prices, ki) 3.Population growth (popgr), all from Penn World Tables 7.0; 4.Share of agriculture in value added (agr) 5.Share of industry in value added (ind) 6.Secondary gross school enrolment (sec) 7.Tertiary gross school enrolment (ter), all from World Bank. Take 7 growth factors (1)

9. Try to find a combination of weights and select a set of “donor”, non-EU countries, by which you can via linear combination approximate “characteristics” (explanatory variables, and the GDP per capita) of countries, in pre-EU period. Take 7 growth factors that would give you a similar country (2)

10. 1973 2004

11. 1995 2004