Empirical the other way round; results that are

Empirical Evidence.Using panel data from 125 countries from 1960-1985 in early research, Helliwell (1994) addressed the topic from a two-way causality point of view.  While the effect of various democracy indicators on growth is largely negative and non-significant, the effects of income growth on democracy are positive, statistically significant and robust. Since democracy encourages education and investment, this suggests that the effect of democracy on economic performance is indirect. The author, however, could not find any systematic evidence of the effects of democracy on subsequent economic growth.  Under the granger causality approach, Burkhart & Lewis-Beck (1994), found that democracy is caused by economic development and not the other way round; results that are similar to Helliwell (1994).The nonlinear relationship in which democracy has a positive effect on growth during low political freedom levels was found by Barro (1996). Once the country has attained a moderate level of freedom, the power of this effect is lost. The economic effect of democracy is temporary and conditional on certain political features, a suggestion based on panel data about a hundred countries during the 1960-1990 period. Evidence that there is a likelihood of growth in political freedoms and subsequent increase in democracy in coming years when a country advances its economic development path (i.e the life expectancy, higher GDP, and increased education levels) was provided in his paper.A full system of equations to study the economic growth and channel variables through which democracy is believed to affect economic performance was estimated by Tavares & Wacziarg (2001). Since democratic institutions attend to the demands of the poor by lowering income inequality and expanding access to education, done at the cost of limiting physical capital accumulation, the overall effect of democracy in growth is relatively negative. In short, there are two sides of the same coin: by providing human capital accumulation and lowering inequality, democracy fosters growth, but it also brings a hindrance to it by raising government expenditure / GDP ratio and lowering the rate of physical capital accumulation.Persson & Tabellini (2007) estimated the growth effects of political transitions (both from and to democratic institutions) following a non-parametric econometric approach. According to their findings, the cost of leaving a democracy is on the order of -2% in GDP growth, and about  -45% in GDP per capita. Those effects had been underestimated in previous research due to poor specifications assumptions.  Papaioannou & Siourounis (2008) in related research, examined the effect of democratization in countries transitioning to representative institutions from autocratic institutions, and on average, their estimates implied that democratizations are associated with 1% increase in annual GDP per capita growth. This effect, though low during the transition stabilizes at a higher level in the long term.Using panel data econometric methods along with non-parametric methods, Acemoglu et al (2014) provided empirical evidence confirming that democracy has a statistically significant and robust effect on the GDP per capita, and included in their work was the development of a novel indicator of democratic institutions.  Democratizations increase GDP per capita by 20% in the long term, and such results are robust in the changes in the variables and specifications (robustness checks) as per baseline results suggestions. The authors concluded that democracy results to an increase in output by a variety of channels such as schooling increase, investment encouragement, economic reforms promotion, social unrest reduction and public goods provision improvement. Further, there is no support for the view that democracy constraints growth in emerging and less developed countries.At the core of his research, Ruiz-Pozuelo et al (2016) placed endogeneity problems in papers addressing this topic. The novel identification strategy that the authors used relied on the survey conducted by 165 country-specific democracy experts. It yielded that democracy does not cause growth, and the common belief of a positive relationship between both concepts is driven by weak identification schemes. They conclude by remarking that democracy is not the key to unlocking economic growth.Discussion.Due to the greater data cross-country availability and easier, more powerful computational tools, the democracy and growth literature has grown to a larger extent. The vast variety of results, however, suggests that the relationship between both variables is ambiguous. Each author claims to have a point to win the argument, although when literature is analyzed without any prior expectation, the linkage between democratic institutions and growth is conditional on the type of democratic indicator, period, identification strategy, empirical method etc.Even though poor countries seem to be less democratic than their richer peers, such fact is not sufficient to conclude that there is a causal relationship. Rich countries are generally democratic, as pointed out by Persson & Tabellini (2006), but this could reflect reverse causation or omitted variable bias. The reason it appears difficult to identify the effects of democracy from the within-country variation is that the concept of ‘democracy’ is blunt: political regimes come in various forms and are reformed in different circumstances.The truth is that democratic institutions promote human (education) and physical (investment) capital accumulation, and therefore foster a favourable business environment. But even if there was a positive effect of democracy on growth, it would be spurious to neglect the relevance of other elements that could have been proved to cause growth, including human capital, diffusion of knowledge and institutions (inclusive vs extractive ones). Democracy is, therefore, not the ultimate cause of development but rather a small piece of the big puzzle that has captured the interest of economists for decades: the wealth of nations.That said, the evidence for democracy being harmful to growth is also dubious; the effect of democracy might be relatively neutral at the worst. Rodrick & Wacziarg (2005) particularly proved that democratic transitions are not followed by poor economic performance. Such views are part of particular political agendas and commentators speech, supported by the assumption that autocratic institutions deliver better performance than democratic ones.As a final remark, although attaining democracy has been a tough, long process for the vast majority of countries (especially for emerging and developing countries), substantial efforts to foster economic stability are being made by governments worldwide. As much as this academic debate touches political grounds, almost nobody in the public policy arena depends on democracy as a driver for economic growth and development. Most of the policy work nowadays stresses the relevance of macroeconomic stability, private sector development, skills, and innovation.