The DRC and the Data: Economic Failure is Not a Necessary Result of State Failure

There are a host of mechanisms for why state failure should result in a corresponding economic decline. These mechanisms also can be backed up by a large number of real world examples, including the Roman Empire and many contemporary African and Middle Eastern countries. However, in the course of this essay I wish to defy some of this conventional wisdom, and argue that economic collapse is not a necessary result of state failure. I will do this by presenting a rudimentary large-n regression that shows a statistically significant positive correlation of state fragility with economic growth. After expanding on this, I will explore the test case of the Democratic Republic of the Congo. Through this exploration, I will prove that a reduced state capacity may not have a negative effect on economic growth.

To bring context to this essay, I ran an exploratory large-N regression with economic growth as the dependent variable and state fragility[1] as the independent variable[2]. I controlled for resource extraction and relative wealth by including per capita GDP and percent of GDP dedicated to resource rents[3] within the regression model. I did this because more fragile states tend to have more resource wealth, and this wealth can allow brief periods of excessive growth based mainly on extraction. There are 148 countries in my analysis, with large variation over the independent variables. Using OLS, the regression coefficients are as follows:


Variable Constant State Fragility Proportion of GDP from Resource Rents GDP per Capita
Coefficient in OLS 3.704*** 0.092 0.006 -0.016
Standard Error 0.7791 0.0675 0.0221 0.0191


R-squared: 0.0559

Main conclusion: A 1 unit rise in the State fragility index is associated with a 0.092% rise in GDP growth on average.

This is obviously a very rough model and no strong indication of any causal relationship; however it is evidence that aggregate production in a country is not necessarily negatively correlated with state fragility, as many seem to think, including Robert Rotberg. In all of the auxiliary regressions I ran, none showed a statistically significant negative coefficient of state fragility. In fact, if the GDP per capita control is left out of the regression, the State Fragility coefficient is positive and statistically significant to a 1% level of confidence. The conclusion I draw from this brief analysis is that the effect of State Fragility on production is ambiguous, even when one controls for resource rents and production level.

It is also interesting to note that auxiliary regressions reveal that state fragility is strongly positively correlated with a high level of resource rents. GDP per capita is also strongly negatively correlated with percent GDP growth. These findings are in line with the Soloh Growth model, which predicts diminishing returns for development, and the resource curse hypothesis, which predicts greater conflict and state failure as a result of natural resource wealth.

To get a better sense of process and how economic growth interacts with state fragility, I turn to the Democratic Republic of the Congo. The DRC meets the qualitative requirements of being a failed state that Robert Rotberg lays out in his article, including “weak or flawed institutions…deteriorating or destroyed infrastructures…regular food scarcities[4].” It also receives an abysmal score on the state fragility index of 23, making it the second most fragile state in the index[5]. Despite this, it is notable that it is currently experiencing its tenth year of positive GDP growth, averaging around 7% per year[6].

Now that it is established that the DRC is both improving economically and also remaining an unstable state, it is necessary to unpack the mechanisms behind such a situation. One possible driver of this growth is the natural resource wealth in the form of diamonds and minerals that the DRC contains. Yet, while these surely plays a role, analysis has shown that for Africa as a region, natural resources only account for 32% of growth[7]. So there is something deeper than resource rents at work. Another way to approach this question is to see if the DRC has overcome some of the frequently mentioned mechanisms for why state failure hurts growth. These mechanisms mainly posit that states provide stability, which allows for longer term decision making, clearer information, less transaction costs, and a greater incentive to invest instead of rent seek. While it is probably very true that the DRC, in its current state of weakness, is suffering unobservable economic losses from its instability, these losses are somehow being offset, as evidenced by the positive economic growth discussed earlier.

A key observation is that while originally resource industries drove growth, this trend has transitioned since 2006. Now, retail and agriculture lead the way in terms of economic growth. This happens despite major border control issues, which are augmented by overlapping agencies and frequent “harassment” by government officials. It also occurs despite little government investment in agriculture and large scale deterioration of infrastructure[8]. This interacts with the idea that perhaps state capacity to provide public goods is not the mechanism through which states most effectively can help economic growth. It may be that states are most effective at promoting growth by focusing on not interfering. Freedom House highlights the interference factor, when it discusses the excessive taxation in the DRC, with statutory rates as high as 40%. Such high taxation drives people into the informal sector, and it is in the informal sector that subsistence living and poverty traps develop. If this logic is extended further, it could be that at very low levels of economic development, where activity is mainly subsistence farming and mining, government becomes mainly a kleptocracy, and in that type of a situation reduced government capacity may translate into reduced government ability to rent seek.

The most important responsibilities of a state may then be those that involve providing a point of coordination, and then letting incentives play out. This may be why the DRC has recorded strong banking growth since the stabilization of monetary policy[9]. In a way, stable monetary policy provides coordination equilibrium. A medium of exchange allows people to better calculate value and translate preferences at low cost between unconnected individuals. This coordination function of government also requires low government investment, and as a result coordination activities like monetary policy can be undertaken even in fragile or low capacity states.

Beyond this speculation, it is important to realize that the DRC, while it shows signs of contradicting conventional wisdom, does still suffer endemic poverty on a massive scale, largely as a result of failed government. The GDP measure in my OLS regression is also a weak way to predict development; such a macroeconomic indicator does not measure the distributional effects of growth or whether any of that growth led to real increases in population welfare. At the end of the day, the DRC and this discussion are interesting because they bring a perspective that refines the model of state weakness leading to economic turmoil. They force an examination of what functions are truly critical to economic functionality. The success of the DRC challenges the idea that weak states would do best to try and imitate their stronger peers. When nations like the DRC are faced with a choice of possible undertakings, and they face tight budgets, they may want to focus more on coordination facilitation instead of intensive infrastructure projects, especially when they are prone to kleptocratic behavior.



Works Cited

Rotberg R., 2003, ‘The Failure and Collapse of Nation-States: Breakdown, Prevention, and Repair’, in ‘When States Fail: Causes and Consequences’, Princeton University Press, ch. 1.


[1] The State Fragility Index 2013 dataset was used.

[2] Data derived from Knoema, who compiled it from “IMF World Economic Outlook.” Data is for 2014-2015, so the data used are essentially a two part time series, with the difference being used to present economic growth.

[3] The dataset for resource rents is from Index Mundi: The dataset for GDP per capita is from the World Bank:

[4] For evidence of this, see:

[5] Table 3 of the State Fragility Index 2013.

[6] Index Mundi.

[7] Mckinsey & Company, “What’s driving Africa’s Growth,”, June 2010.

[8] Ferf, Adriaan. Wageningen University. Page 9.