Empir Econ DOI 10.1007/s00181-016-1089-1
Does democracy dampen the effect of finance on economic growth? Kevin Williams1
Received: 18 August 2014 / Accepted: 26 February 2016 © Springer-Verlag Berlin Heidelberg 2016
Abstract Although a large empirical literature seeks to explain the effect of financial development in promoting economic growth, there is surprisingly little evidence for the impact of political institutions on the growth–finance relationship. This paper finds that political institutions condition the effect of financial development on economic growth. Using a dynamic panel estimator and a sample of 78 developing and emerging economies for the years 1982–2011, the paper investigates the impact of democracy on the relationship between financial development and economic growth. The paper finds that democracy does not enhance the effect of financial development on economic growth. This finding is consistent with a view that democracy can be captured by political elites or other special interests in developing and emerging economies, where institutions are relatively weak. Keywords Financial development · Economic growth · Democracy · Panel data JEL Classification G20 · 043
1 Introduction There are several reasons why financial development contributes positively to economic growth. A sound financial system will facilitate creative destruction, so that the most efficient firms have access to finance at low cost, which in turn will improve growth (Beck 2013; Levine 2005). With this interpretation of the financial system and
B 1
Kevin Williams
[email protected] Department of Economics, The University of the West Indies, St. Augustine, Trinidad, Trinidad and Tobago
123
K. Williams
the hope of achieving high growth, economists have advocated and policy makers have removed restrictions on financial institutions. Bagehot (1873) and Schumpeter (1911) were the first researchers to draw attention to the role of financial institutions in promoting economic development. Since then, there is a large empirical literature that investigates the effect of financial development on economic development. Some of these empirical findings suggest that financial development exerts a positive influence on growth (Beck and Levine 2004; Zhang et al. 2012). Other studies find that the positive relationship between financial development and growth is not robust (Cecchetti and Kharroubi 2013; Rousseau and Wachtel 2011). Still, there is a small literature that examines the political institutions determinants of financial development (Bhattacharyya 2013; Huang 2010). To date, however, there is little evidence on the role of political institutions in the finance–growth relationship. Political institutions play an important role in shaping the rules and the laws that determine how financial institutions function in society. Financial institutions do not exist independent of political institutions; they interact to determine economic outcomes. In countries with extractive political institutions, where political powers are highly concentrated hence weak political institutions, political elites have an incentive to introduce policies that direct financial resources from the financial system to buy political support and provide patronage in order to consolidate their power. In these societies, financial institutions are likely to have negative effects on economic outcomes, in particular economic growth, as projects with high return are not likely to be funded. These extractive political institutions serve the interests of political elites at the expense of the welfare of society. On the other hand, in countries where political powers are widely distributed (hence strong political institutions), more constraints are placed on political elites; hence, they have less incentive to influence the allocation of financial resources to low-return projects. In this case, competitive pressures are likely to be generated in the economy, and thus, financial institutions will be able to allocate resources to the most efficient projects, which in turn will likely boost economic growth. Understanding the extent to which political institutions impact the growth–finance relationship is necessary for forecasting how growth is likely to evolve in developing and emerging economies. In particular, if political institutions influence the growth effect of financial development, countries can improve their growth rates by strengthening their political institutions. This paper empirically investigates the impact of democracy on the effect of financial development on growth. Using a dynamic panel estimator with a sample of developing and emerging economies in a standard growth framework, the empirical results show that democracy mitigates the positive effect of financial development on growth. The finding is consistent with the interpretation that political elites have an incentive to retain political power, and they intervene in the financial system to further their political interests by, for example, ensuring that credit is allocated to their support base. This is possible in democracy with weak underlying institutional support. Democracy has a populist appeal, and political elites who are interested in retaining political power will respond to pressures from voters or other special interests through easy credit policies. This suggests that democracy can be captured by different groups in society to serve their interests, with adverse effect on economic growth.
123
Does democracy dampen the effect of finance on economic. . .
Other studies have also investigated the link between political institutions and growth and report mixed evidence. Tavares and Wacziarg (2001) find that democracy does not contribute to growth. In contrast, the evidence reported by Acemoglu et al. (2013) and Papaioannou and Siourounis (2008) suggests that democracy does improve growth. Khawaja and Mian (2005) show that under weak political institutions firms with political connections benefit disproportionately from the distribution of political rents, which adversely affects growth. In a related analysis, Choong (2012) finds that the effect of international capital flows, in particular foreign direct investment, on economic growth is enhanced by financial development. The present paper contributes to the literature in several ways. First, a major innovation of this paper relative to the above studies is that we provide evidence on how within-country variations in political institutions influence the effect that financial development has on growth. Second, by controlling for country and year fixed effects, we focus on the effect of within-country variations of the interaction between democracy and financial development, thus accounting for omitted fixed factors.1 Third, by using the system GMM dynamic panel estimator, we are able to identify the exogenous variations in financial development and democracy and control for bias arising from the lagged dependent variable. The rest of the paper is organized as follows. Section 2 reviews the literature on the link between financial development and growth and the relationship between democratic political institutions and growth. Section 3 outlines the econometric model and describes the data. Section 4 presents the results. Section 5 further estimates the effect of the different dimensions of democratic institutions on the growth–finance nexus. Section 6 addresses endogeneity issues. Section 7 draws the conclusion.
2 Related literature An important paper by King and Levine (1993) is probably the first to systemically investigate the effect of financial development on economic growth. They use crosscountry data and find evidence that financial development has a positive effect on economic growth and other measures of economic development. Their paper generated a large literature on the link between finance and growth. Demetriades and Hussein (1996), however, caution against the evidence in King and Levine (1993), and instead, they argue that the relationship between financial development and growth, if anything, is bidirectional. Arestis and Demetriades (1997) and De Gregorio and Guidotti (1995) also find evidence using time series methods that caution against the positive relationship in cross-country data. Levine et al. (2000) and Beck and Levine (2004) use a dynamic panel to support the findings in King and Levine (1993). In particular, these authors improved on the methods in King and Levine (1993) and find that the exogenous components of financial development are positively related to growth. 1 These potentially omitted fixed factors are cross-country differences in geography, colonial heritage, and other historical factors that jointly determine the interaction between financial development and democracy and growth.
123
K. Williams
In a related paper, Rioja and Valev (2004a) examine the differential effect of financial development on growth and the channel through which financial development affects growth in a panel of developed and developing countries. They find that financial deepening contributes to growth in low-income countries through capital accumulation, while financial deepening affects growth in middle- and high-income countries through higher productivity. They thus conclude that for financial development to improve growth, countries must reach a threshold level of income. Rioja and Valev (2004b) also find that the level of financial development matters, that is, low levels of financial development do not contribute to growth in developing countries. This finding suggests that the positive contribution of financial development to economic growth depends on the level of financial development. Another strand of the literature investigates the effect of political institutions on growth. The evidence here is also inconclusive. One important feature of democracy is the tendency to redistribute resources in support of the median voter. Alesina and Rodrik (1994) and Persson and Tabellini (1994) argued that this redistribution is likely to be distortionary with negative effect on growth. Other researcher also argued that democracy can be captured by powerful elites, thus discourage economic growth (Olson 1982). To counter the above finding, using exogenous variations in democracy, Acemoglu et al. (2013) and Papaioannou and Siourounis (2008) find that democratic political institutions promoted economic growth. Unfortunately, none of these studies examine the important issue of the impact of political institutions on the effect of financial development on growth. In particular, this paper builds on and complements the extant literature by investigating whether the effect of within-country variations in financial development on growth varies with the effect of within-country variations in democracy.
3 Econometric specification and data The paper uses an empirical model similar to Beck and Levine (2004) and Aghion et al. (2009) to examine the impact of democracy on the effect of financial development on growth in an unbalanced panel of 78 developing and emerging economies, for which data are available, for the years 1982–2011. The dependent variable is the growth rate of per capita GDP (constant 2000 USD). The equation estimated is: ∗ Yi,t = αY i,t−1 + β1 FDi,t + β2 FDi,t POL2i,t + β3 POL2i,t + γ z i,t + μt + ηi + εi,t
(1)
where Yi,t is the natural logarithm of GDP per capita in country i at time t; FDi,t is one of four measures of financial development indicators, bank private credit to GDP ratio, liquid liabilities to GDP ratio, bank deposit to GDP ratio, and private credit by deposit money bank and other financial institutions to GDP ratio; POL2i,t is the Polity2 measure of democracy from Polity IV Project (Marshall et al. 2011); z i,t is a set of standard growth control variables, government consumption to GDP ratio, inflation, the sum of exports and imports to GDP ratio, net FDI inflows to GDP ratio,
123
Does democracy dampen the effect of finance on economic. . .
and debt to GNI2 ratio; μt is year fixed effect, capturing any shocks to growth affecting all countries similarly; ηi is country fixed effect; and εi,t is an error term, containing ∗ POL2 , an interaction all other omitted factors. The main variable of interest is FDi,t i,t between democracy and financial development. We note that the econometric analysis is based on the premise that the effect of financial development on growth may differ across countries with different levels of democratic political institutions. The interaction term identifies whether and how democratic political institutions affect the effect of changes in financial development on growth. POL2i,t , the main measure of democracy, is a composite measure of democracy, which is the difference between democracy and autocracy indices and ranges from 10 to −10 (10 being the most democratic and −10 being the most autocratic). In the Polity IV database, democracy is coded as constraints on the executive, the competiveness of political participation, and the openness and competitiveness of executive recruitment. Autocracy is also coded in a similar way: competiveness of political participation, the regulation of participation, the openness and competiveness of executive recruitment, and constraints on the executive. The paper uses two other widely used measures of democracy, the Polity2 subindex of democracy and the Freedom House Political Rights Index. The Freedom House Political Rights Index is based on a 1–7 scale, where one is the highest Political Rights Index score. Countries rate high if they enjoy free and fair elections, if candidates who are elected actually rule, if political parties are competitive, if opposition parties are allowed to function, and if minority groups have reasonable self-government or can participate in the government through informal consensus. Following Acemoglu et al. (2008), all three measures of democracy are transformed so that they lie between zero and one, with one being the most democratic. This transformation improves interpretation of estimates. The financial development indicators ´ are from the new dataset constructed by Cihák et al. (2013). All the control variables are from the World Bank, World Development Indicators. Other variables are discussed when they are introduced in the analysis. The paper uses the system GMM dynamic panel estimator due to Blundell and Bond (1998) to estimate Eq. (1). In all the regressions, the Windmeijer (2005) small sample correction is applied to compute robust two-step standard errors. The system GMM dynamic panel estimator controls for joint endogeneity of all the right-hand side variables and biases caused by country fixed effects (which are country specific characteristics that are not observed and may be correlated with both growth and the right-hand side variables, in particular democracy and financial development). The system GMM dynamic panel estimator also controls for another key source of bias in the estimation due to the lagged dependent variable. Nickell (1981)3 shows that in finite sample, the lagged dependent variable makes the estimator inconsistent because Yi,t−1 is correlated with εi,t . This econometric problem is less likely with the system GMM dynamic panel estimator. 2 The data on debt to GDP ratio have extensive gaps for most of the countries in the sample: To maintain the dataset, the paper uses debt to GNI ratio. But the main focus of this paper is the impact of democracy on the growth effect of finance. 3 As T > 15 the bias becomes negligible and converges to 0.
123
K. Williams
In all specifications, the paper uses a 5-year panel, where an observation is taken every fifth year from 1982 to 2011, though 5-year averages are used to check the main result. Acemoglu et al. (2008) find that 5-year averages introduced additional serial correlation in the estimation, which makes inferences difficult. The paper follows Acemoglu et al. (2008, 2013) and constructs the panel data using the fifth year in order to prevent serial correlation arising from 5-year averages.4
4 Main result: the impact of democracy on the effect of financial development on growth Table 1 reports estimates from Eq. (1) with the fifth year of per capita GDP growth as the dependent variable. All specifications are estimated with the system GMM dynamic panel estimator. In column 1, the financial development indicator (private credit) is credit to the private sector to GDP ratio by deposit bank. In this column, democracy and financial development have a positive and statistically significant impact on economic growth. For example, the estimate for private credit (column 1) suggests that increasing the ratio of private credit to GDP for a country from 1.0 to 1.50 would raise the growth rate by about 0.09 (0.50 × 0.17) percentage points. The positive association between financial development and growth is consistent with some studies in the financegrowth literature (Beck and Levine 2004; Levine et al. 2000), but does not support others (Cecchetti and Kharroubi 2013). The interaction between democracy and financial development is negative and statistically significant (estimate −0.18 and standard error = 0.07), implying that the marginal effect of private credit is negatively associated with economic growth in democracy. The estimate implies that a country with maximum democratic score can mitigate the positive effect of financial development on growth. For instance, if China and Costa Rica have the same private credit to GDP ratio of, say, 0.5 but China has a low Polity2 score of 0.15 and Costa Rica has maximum Polity2 score of 1, the relative growth rate in China will improve by 0.07 percentage points.5 Though the estimate on the interaction term appears economically small, it is important to note that credit to the private sector does not improve economic growth in democracy. Columns 2–46 use different measures of financial development (liquid liabilities, private credit by deposit banks and other financial institution, and bank deposits), which supports the negative and statistically significant impact of democracy on the marginal effect of financial development on growth. The Hansen J test and the AR(2) test support the estimated coefficients in all specifications. 4 Studies on the finance–growth relationship use 5-year averaged data, and do not consider this additional source of bias. 5 This number is computed: 0.5(0.17) + (−0.18 × 0.15 × 0.50) for China; Rica. 6 With 5-year averaged data, the estimates are very similar to columns 1 and 3. The estimates for liquid
liabilities, Pol2, and the interaction term in column 2 are qualitatively similar, though they are statistically insignificant at conventional levels with averaged data. Column 4 estimates are different with averaged data, but this is not surprising, as Aghion et al. (2005) and Levine et al. (2000) find that private credit by deposit banks is the most robust predictor of growth.
123
Does democracy dampen the effect of finance on economic. . . Table 1 Impact of Polity2 democracy on the effect of finance on growth: Dependent variable is the fifth year of per capita growth
(1) Pol2 Private credit
(2)
(3)
(4)
5.25
3.01
5.80
3.27
(2.47)**
(1.82)
(2.80)**
(4.98)
0.17 (0.06)***
Private credit * Pol2
−0.18 (0.07)**
Liquid liabilities
0.05 (0.02)** −0.06
Liquid liabilities * Pol2
(0.03)* Pcdbfin
0.18 (0.08)** −0.19
Pcdbfin * Pol2
(0.09)** Bankdeposit
0.10 (0.13) −0.09
Bankdeposit * Pol2 Specifications have full year and country fixed effects. Private credit is credit to the private sector by deposit bank. Private credit * Pol2 is interaction between private credit and democracy. Liquid liabilities * Pol2 is interaction between liquid liabilities and democracy. Pcdbfin is credit to the private sector by deposit bank and other financial institutions. Pcdbfin * Pol2 is interaction between private credit by deposit bank and other financial institutions and democracy. Bankdeposit is bank deposit. Bankdeposit * Pol2 is interaction between bank deposit and democracy. ***, **, * significant at 1, 5, 10 %. Robust standard errors in parenthesis. The Windmeijer (2005) small sample correction is used to compute two-step robust standard errors
(0.15) Lagged (log GDP pc) Govconsumption
−0.50
−0.09
−0.55
(0.31)
(0.29)
(0.35)
−0.20 (0.43)
−0.02
−0.02
−0.04
−0.01
(0.06)
(0.05)
(0.07)
(0.09)
Log (1 + inflation)
0.16
−0.42
0.11
−0.32
(0.33)
(0.43)
(0.35)
(0.42)
Trade openness
0.01
−0.01
0.01
−0.00
(0.01)
(0.01)
(0.01)
(0.01)
−0.03
−0.09
−0.09
−0.07
Debt
(0.07)
(0.05)*
(0.08)
(0.06)
FDI (net inflows)
0.06
0.24
0.07
0.12
(0.17)
(0.05)***
(0.14)
(0.22)
Constant
1.20
1.87
1.50
2.02
(2.41)
(2.83)
(2.47)
(3.01)
Observations
313
314
313
314
Countries
78
78
78
78
No. of instruments
28
50
30
22
Hansen p value
0.64
0.34
0.58
0.18
AR2 p value
0.88
0.40
0.84
0.80
The main result in Table 1 suggests that democracy conditions the effect of financial development on growth. If political elites in democracy use their political power to influence the allocation of credit in the banking sector to politically connected projects,
123
K. Williams
it is likely that private credit will reduce growth through favored low-productivity firms. This interpretation is consistent with the captured democracy theoretical model developed in Acemoglu and Robinson (2008) and the private interest view of financial development (Becker and Stigler 1974).7 The empirical evidence is also consistent with one interpretation of the 1997 Asian financial crisis: Inefficient firms with political influence had disproportionate access to credit from the banking sector (Mankiw 2010). Though the banking sector in democracy is motivated by allocating credit to its most efficient use, it is also possible for inefficient firms with political connections to access credit from the banking sector. This is likely to be the case if the banking sector can influence government policies that remove restrictions from the banking sector. An alternative explanation consistent with the empirical finding is that, other things being equal, the political fate of governments in democracy is determined by the largest constituents of voters; therefore, one would expect governments to adopt credit policies that favor these voters as opposed to allocation through market institutions. Debt and FDI inflows are the only two controls that are statistically significant in Table 1, column 2. 4.1 Alternative measures of democracy To further identify whether democracy conditions the effect of financial development on growth, the paper uses two alternative measures of democracy: the Polity2 subindex of democracy and the Freedom House Political Rights Index.8 Table 2 reports the results. Table 2 has similar structure to Table 1, but uses the Freedom House Political Rights Index as the democracy measure. Though the impact of democracy on the growth effect of financial development is slightly smaller in Table 2, maybe because of different measure of democracy, the coefficients are precisely estimated. The estimates in columns 1–4 show that the interaction between democracy and financial development is negative and statistically significant in columns 1–3. In column 1, for example, the coefficient on the interaction term is −0.15 (standard error= 0.04), which is statistically significant at the 1 % level. The coefficients on democracy and financial development are broadly similar to Table 1.9 All the other controls are statistically insignificant, except for FDI inflows. Table 3 uses the Polity2 subindex of democracy and again the main result in Table 1 holds. Columns 1–3 show that democracy does not enhance the growth effect of financial development. 7 According to the private interest view of financial deepening, ‘politicians do not intervene into the financial system to further public welfare but to divert the flow of credit to politically connected firms’ (Beck 2013: 19). 8 Acemoglu et al. (2008) use the Freedom House Political Rights Index as their main measure of democratic political institutions to investigate the modernization hypothesis. Cavallo and Cavallo (2010) use the Polity2 subindex of democracy to examine the impact of democracy on the growth effect of crises. 9 Five-year averaged results are similar to fifth year data for the main estimates, and the results for the control variables are stronger with 5-year averaged data. But, again, we prefer our data to 5-year averages, because of the additional serial correlation it introduces in the regressions.
123
Does democracy dampen the effect of finance on economic. . . Table 2 Impact of Freedom House Political Rights Index on the effect of finance on growth: Dependent variable is the fifth year of per capita growth
FH
(1)
(2)
4.75
3.25
(1.73)*** (1.64)** Private credit
(3)
(4)
4.49
1.39
(2.16)** (1.93)
0.13 (0.03)***
Private credit * FH
−0.15 (0.04)***
Liquid liabilities
0.05 (0.02)*** −0.06
Liquid liabilities * FH
(0.03)** Pcdbfin
0.12 (0.06)** −0.14
Pcdbfin * FH
(0.06)** Bankdeposit
0.02 (0.05) −0.01
Bankdeposit * FH
(0.06) Lagged (log GDP pc) Govconsumption Log (1 + inflation) Trade openness
Private credit * FH is interaction between private credit and Political Rights. Liquid liabilities * FH is interaction between liquid liabilities and Political Rights. Pcdbfin * FH is interaction between private credit by deposit bank and other financial institutions and Political Rights. Bankdeposit is bank deposit. Bankdeposit * FH is interaction between bank deposit and Political Rights. See notes to Table 1
Debt
−0.37
−0.08
−0.44
(0.26)
(0.28)
(0.29)
−0.13 (0.28)
−0.20
−0.03
−0.04
−0.04
(0.06)
(0.06)
(0.07)
(0.06)
0.10
−0.42
0.08
−0.55
(0.30)
(0.41)
(0.30)
(0.48)
−0.01
−0.01
−0.00
−0.01
(0.01)
(0.01)
(0.01)
(0.01)
−0.06
−0.09
−0.07
−0.08
(0.07)
(0.05)
(0.07)
(0.06)
FDI (net inflows)
0.18
0.24
0.17
0.23
(0.13)
(0.05)*** (0.11)
(0.05)***
Constant
0.88
1.99
1.66
3.96
(2.10)
(2.60)
(2.19)
(2.60)
Observations
314
315
314
315
Countries
78
78
78
78
No. of instruments
28
50
30
50
Hansen p value
0.79
0.37
0.63
0.26
AR2 p value
0.86
0.56
0.79
0.43
4.2 Additional controls Table 4 probes the robustness of the main result in Table 1, focusing on the Polity2 Index of democracy and private credit by deposit bank. Aghion et al. (2005) and
123
K. Williams Table 3 Impact of Polity2 subindex of democracy on the effect of finance on growth: Dependent variable is the fifth year of per capita growth
(1) Dem2 Private credit
(2)
(3)
(4)
3.04
2.11
3.42
1.58
(1.36)**
(1.36)
(1.63)**
(3.24)
0.11 (0.03)***
Private credit * Dem2
−0.11 (0.03)***
Liquid liabilities
0.04 (0.02)** −0.04
Liquid liabilities * Dem2
(0.02)* Pcdbfin
0.12 (0.04)*** −0.12
Pcdbfin * Dem2
(0.05)** Bankdeposit
0.05 (0.09) −0.04
Bankdeposit * Dem2
(0.10) Lagged (log GDP pc) Govconsumption Log (1 + inflation) Trade openness
Private credit * Dem2 is interaction between private credit and Polity2 subindex. Liquid liabilities * Dem2 is interaction between liquid liabilities and Polity2 subindex. Pcdbfin * Dem2 is interaction between private credit by deposit bank and other financial institutions and Polity2 subindex. Bankdeposit * Dem2 is interaction between bank deposit and Polity2 subindex. See notes to Table 1
Debt
−0.40
−0.06
−0.52
(0.29)
(0.29)
(0.33)
−0.18 (0.40)
−0.02
0.03
−0.03
−0.03
(0.06)
(0.06)
(0.07)
(0.07)
0.22
−0.36
0.20
−0.30
(0.31)
(0.43)
(0.32)
(0.44)
−0.00
−0.01
−0.00
−0.00
(0.01)
(0.01)
(0.01)
(0.01)
−0.07
−0.09
−0.08
−0.06
(0.07)
(0.05)
(0.08)
(0.07)
FDI (net inflows)
0.12
0.24
0.13
0.18
(0.14)
(0.05)*** (0.11)
Constant
2.57
2.30
2.51
3.49
(2.22)
(2.66)
(2.14)
(2.43)
(0.12)
Observations
314
315
314
315
Countries
78
78
78
78
No. of instruments
28
50
30
22
Hansen p value
0.72
0.37
0.68
0.09
AR2 p value
0.70
0.40
0.60
0.57
Levine et al. (2000) find that private credit is the most robust predictor of growth; the paper adopts their approach and focuses on private credit in the rest of the empirical analysis. Because political elites have an incentive to influence the allocation of credit,
123
FH squared
Pol2 squared
Private credit squared
Remittance
Log(population)
Private credit * Dem2
Private credit * FH
Private credit * Pol2
Private credit
Dem2
FH
Pol2
(0.04) −0.00 (0.00)
−0.05 (0.04) −0.00 (0.00)
(0.04)
4.8 1 (3.88)
(4.37)
(0.00)
−0.00
(0.04)***
−0.12
(0.30)
−0.15
(0.03)***
−0.11
(0.10)*
0.21
5.58
−0.05
(0.29)
−0.05
0.01
(0.04)***
(0.30)
0.00
(0.04)***
−0.11
(0.31)
(0.05)**
(0.06)**
−0.12
(0.09)*
(0.36)
−0.13
−0.15
(0.09)*
0.15
(5.07)
−1.95
−3.18 (5.52)
(5)
(4)
−0.14
(0.05)***
(0.06)**
0.17
(1.73)
3.11
(3)
−0.19
0.13
0.15
3.76
(2.00)*
4.00
(2)
(2.11)*
(1)
(4.64)
6.33
(0.00)
−0.00
(0.02)***
−0.09
(0.25)
0.16
(0.02)**
−0.06
(0.12)
0.18
(5.71)
−5.29
(6)
(2.76)**
5.95
(0.00)
−0.00
(0.02)***
−0.09
(0.25)
0.19
(0.02)**
−0.06
(0.12)
0.20
(3.36)
−3.64
(7)
(0.00)
−0.00
(0.03)***
−0.11
(0.27)
0.02
(0.02)***
−0.08
(0.13)*
0.23
(6.06)
2.78
(8)
Table 4 Impact of Polity2 democracy on the effect of private credit on growth: Dependent variable is the fifth year of per capita growth (robustness: additional covariates)
Does democracy dampen the effect of finance on economic. . .
123
123
78
Countries
78
284
(6.67)
(6.58)
313
4.82
4.26
(0.11)
(0.19)
(0.07) 0.08
(0.08)
0.06
(0.01) −0.01
−0.02
78
284
(6.87)
2.23
(0.13)
0.00
(0.07)
−0.05
(0.01)
0.01
−0.00
0.00
(0.01)
(0.35)
0.25
(0.06)
0.25
(0.05)
(0.06)
0.02
(0.40)
(0.33)
0.03
−0.02
0.17
(0.36)
(0.37)
−0.54
(3)
(0.35)
−0.55
(2)
−0.46
(1)
Observations
Constant
FDI (net inflows)
Debt
Trade openness
Log(1 + inflation)
Govconsumption
Lagged (log GDP pc)
Private credit * log (1 + inflation)
Investment
Dem2 squared
Table 4 continued
78
284
(7.08)
4.44
(0.13)
0.02
(0.07)
−0.03
(0.01)
0.01
(0.35)
0.22
(0.06)
0.02
(0.45)
−0.67
(4)
78
283
(6.42)
4.57
(0.11)
0.14
(0.07)
−0.09
(0.01)
−0.00
(0.34)
0.02
(0.06)
−0.07
(0.44)
−0.73 (0.36)
(0.32)
78
283
78
284
(5.85)
−3.52
−3.52 (6.15)
(0.10)*
0.18
(0.06)
−0.07
(0.01)
−0.00
(0.72)
0.81
(0.06)
−0.05
(0.11)*
0.20
(0.07)
−0.08
(0.01)
−0.00
(0.75)
0.76
(0.07)
−0.06
−0.33
(0.03)
−0.26
−0.05
(0.04)***
(0.03)
(0.04)***
(0.03)***
0.15
(7)
−0.04
0.14
(6)
0.12
(5)
78
284
(6.96)
−1.48
(0.10)
0.16
(0.07)
−0.08
(0.01)
−0.00
(0.73)
0.74
(0.06)
−0.07
(0.47)
−0.34
(0.03)
−0.05
(0.04)***
0.16
(6.40)
0.06
(8)
K. Williams
0.89
AR2 p value
0.43
0.46
30
(2)
0.33
0.26
31
(3)
0.40
0.23
32
(4)
0.17
0.16
33
(5)
0.44
0.06
34
(6)
0.53
0.14
34
(7)
0.39
0.14
34
(8)
Specifications have full year and country fixed effects. Private credit * Pol2 is interaction between private credit and Polity2 Index of democracy. Private credit * FH is interaction between private credit and Political Rights Index. Private credit * Dem2 is interaction between private credit and Polity2 subindex for democracy. ***, **, * significant at 1, 5, 10 %. The Windmeijer (2005) small sample correction is used to compute two-step robust standard errors. Robust standard errors in parenthesis. Columns 1–6 use additional controls and Polity2 Index of democracy. Columns 7–8 use Freedom House Political Rights Index and Polity2 subindex of democracy and additional controls
29
0.51
Hansen p value
(1)
No. of instruments
Table 4 continued
Does democracy dampen the effect of finance on economic. . .
123
K. Williams
private credit by deposit bank is more likely to capture the impact of democracy on the growth effect of financial development. The major concern with the main result in Table 1, however, is that the correlation between the interaction of democracy and the financial development indicators and growth is driven by a third variable. The paper includes a set of covariates to investigate this hypothesis. This is important in order to isolate the impact of democracy on the growth effect of finance. Table 4 reports the results. Columns 1–6 show the regressions corresponding to column 1 of Table 1, with additional controls. Columns 7 and 8 use the Freedom House Political Rights Index and the Polity2 subindex of democracy as further checks on columns 1–6. Column 1 includes the log of total population. This is useful since the empirical literature finds that population size can explain growth rates across countries (Mankiw et al. 1992). The size of the estimates for the Polity2 measure of democracy and private credit are broadly similar to Table 1, though they are not precisely estimated. The main result holds when total population is included as an additional regressor: Democracy reduces the growth effect of private credit. The estimate of −0.15 (standard error = 0.06) is slightly smaller than Table 1, column 1 in absolute value (estimate of −0.18 and standard error= 0.07). The coefficient on total population10 is statistically insignificant. Column 2 controls for the level of remittances11 to GDP ratio. The evidence on the relationship between remittances and growth is inconclusive. There is evidence that remittances are negatively associated with growth (Chami et al. 2005), but other studies find a positive association (Catrinescu et al. 2009). More recently, Abdih et al. (2012) find that remittances can also lead to deterioration in democratic institutions. The paper controls for this source of influence in column 2. The effect of remittances on growth is negative and statistically insignificant. Although the level of remittances reduces the impact of the interaction between democracy and private credit on growth, the estimate (−0.13 and standard error = 0.05) is negative and statistically significant at the 5 % level. Column 3 controls for the potential nonlinear relationship between growth and financial development. Arcand et al. (2012) and Cecchetti and Kharroubi (2012) find evidence of a nonlinear relationship between private credit and growth. There is no statistically significant effect, in column 3, of a nonlinear relationship between private credit and growth. The interaction term again remains negative and is now statistically significant at the 1 % level. In column 4, the paper follows Barro (2012) and others and controls for a nonlinear effect of democracy on growth. This coefficient is positive and statistically insignificant. The interaction effect of democracy and private credit remains negative and statistically significant at the 1 % level, and the estimate (−0.11 and standard error = 0.04) is reduced by about 39 % compared with Table 1, column 1.
10 Data on total population are from the World Bank, World Development Indicators. 11 Data on remittances are from the IMF and include: current transfers by migrant workers and wages and
salary earned by nonresident workers.
123
Does democracy dampen the effect of finance on economic. . .
Column 5 includes investment12 to GDP ratio, which is gross capital formation. The investment ratio is a robust determinant of growth (Levine and Renelt 1992). The coefficient on this variable is positive and statistically significant at conventional levels, as expected, and the main finding does not budge. Rousseau and Wachte (2002) find that the effect of financial development on growth varies with the level of inflation. The paper tests this hypothesis with an interaction term between private credit and inflation in column 6. In contrast to this hypothesis, there is no statistically significant evidence in the data that the effect of private credit varies with the level of inflation. The impact of democracy on the growth effect of financial development is still negative and statistically significant when the interaction between private credit and inflation is included and all the other additional controls together, though the estimate (−0.06) is about 33 % the size of the estimate in Table 1, column 1. Finally, the paper re-estimates the regression in column 6 with the Freedom House Political Rights Index (column 7) and the Polity2 subindex of democracy (column 8). The evidence with these two different measures of democracy supports the result in columns 1–6: Democracy dampens the effect of private credit on growth. 4.3 Different subsamples Thus far, the paper uses the fifth year panel with the full sample to investigate the impact of the interaction between democracy and financial development on growth and finds this relationship to be negative. Table 5 presents further robustness checks on the baseline sample by investigating different subsamples. The different subsamples ensure that the main result is not influenced by any particular group of countries or regions. In columns 1–4, the paper excludes, in turn, the Middle East and North Africa (MENA), Asia, Sub-Saharan Africa (SSA), and Latin America and the Caribbean (LAC) from the baseline sample. Based on the CIA World Factbook, column 5 omits all oil-based economies from the main sample to ensure that the result is not driven by oil-rich economies. These countries are: Algeria, Bahrain, Gabon, Iran, Kuwait, Nigeria, and Trinidad and Tobago. This check is inspired by Bhattacharyya and Holder (2014) who find that natural resource rents retard financial development in countries with weak democratic political institutions, which reduces economic growth. According to the World Bank classification, in columns 6–8, countries with different income levels are dropped. This will again provide evidence whether the result is unique to countries at different income levels. The main result is negative in all specifications and is statistically significant at conventional levels in six of the eight specifications. The statistically insignificant coefficient in columns 8 and 3 may be explained by the substantial reduction in the sample size. Based on the results in Table 5, the main result is not driven by any region, oil-rich countries, or countries at different income levels. The main result is also supported with the Freedom House Political Rights Index and the Polity2 subindex of democracy.13 12 Gross capital formation is from the World Bank, World Development Indicators. 13 These results are available upon request.
123
123
Constant
FDI (net inflows)
Debt
Trade openness
Log(1 + inflation)
Govconsumption
Lagged (log GDP pc)
Private credit * Pol2
Private credit
Pol2
(0.07)
(0.04)
6.99
−2.43
(4.26)
(2.47)
(3.60)*
(0.09)**
(0.19)
(0.13)
0.62
0.24
(0.08)
(0.01)
0.20
(0.05)
0.09
−0.05
−0.16
0.02
(2.06)
−0.00
(0.17)
0.13
(0.10)
−0.16
(0.01)
−0.01
(0.01)
(0.01)
−0.00
0.00
(0.01)
(0.36)
0.32
(0.05)
0.04
(0.31)
−0.27
(0.03)***
−0.12
(0.02)***
0.13
(1.51)
2.98
(4)
(0.35)
−0.42
(0.11)
−0.05
(0.47)
−0.48
(0.06)
−0.09
(0.04)
0.08
(2.47)
1.63
(3)
0.00
(0.40)
−0.06
−0.00
(0.31)
(0.36)
(0.32)
0.33
−0.29
−0.54
0.20
−0.33
(0.13)**
−0.19
(0.11)***
(0.05)***
(0.06)***
0.30
0.18
7.31
(3.58)
5.55
(2)
(2.23)**
(1)
(2.52)
1.40
(0.19)
0.03
(0.07)
−0.00
(0.00)
−0.01
(0.34)
0.15
(0.05)
−0.01
(0.31)
−0.58
(0.07)**
−0.18
(0.06)***
0.17
(2.55)**
5.29
(5)
(4.24)
2.30
(0.17)
0.10
(0.07)
−0.01
(0.00)
0.00
(0.38)
0.09
(0.06)
−0.02
(0.40)
−0.57
(0.05)**
−0.14
(0.04)***
0.13
(2.35)
3.78
(6)
(3.37)
1.14
(0.17)
0.23
(0.09)**
−0.21
(0.01)
−0.01
(0.42)
0.05
(0.12)
−0.11
(0.42)
0.04
(0.06)*
−0.12
(0.05)**
0.12
(3.18)
4.38
(7)
Table 5 Impact of Polity2 democracy on the effect of private credit on growth: Dependent variable is the fifth year of per capita growth (alternative samples)
(4.63)
−0.07
(0.16)
0.17
(0.09)
−0.05
(0.01)
−0.00
(0.40)
0.43
(0.08)
0.03
(0.36)
−0.52
(0.18)
−0.27
(0.13)*
0.27
(3.49)
4.11
(8)
K. Williams
28
0.83
0.76
Countries
No. of instruments
Hansen p value
AR2 p value
0.88
0.15
28
63
249
ASIA
0.38
0.31
28
47
187
SSA
0.75
0.55
28
58
232
LAC
0.91
0.61
28
71
303
Oil-based economies
0.41
0.51
28
57
231
Low income
0.78
0.29
28
50
191
Low middle income
0.71
0.50
28
54
209
Upper middle income
Specifications have full year and country fixed effects. MENA is Middle East and North Africa. SSA is Sub-Saharan Africa. LAC is Latin America and the Caribbean. Windmeijer (2005) small sample correction is used to compute two-step robust standard errors. Robust standard errors in parenthesis. ***, **, * significant at the 1, 5, 10 %
271
66
Observations
MENA
Omitted observations
Table 5 continued
Does democracy dampen the effect of finance on economic. . .
123
K. Williams
5 The effects of dimensions of democratic institutions This section extends the baseline result in Table 1 by providing estimates of the impact of the different dimensions of democratic institutions on the effect of financial development on economic growth. This extension will provide important details about which dimension of democratic institutions is more important in explaining the effect of financial development on growth. It is possible that different components of democratic institutions have different effects, and estimating the effect of the composite measure of democratic institutions may mask this heterogeneity. In the Polity IV dataset, democracy is coded based on three components: constraints on the executive, competitiveness of political participation, and the openness and competitiveness of executive recruitment. We also explore the effect of civil liberties from the Freedom House dataset. Table 6 presents within-country estimates of the effect of the different components of democratic institutions on the finance–growth relationship. Column 1 reproduces the estimates from Table 1, column 1 where the composite measure of democratic institutions is used as the main measure of democracy, for comparison with columns 2– 5. The main message from Table 6 is that when components of democracy are estimated separately, the effect of financial development on growth is negative. This effect is statistically significant at conventional levels. In column 2, we estimate an interaction model between private credit and constraints on the executive. This estimate (−0.03 and standard error = 0.01) is negative and significant at the 5 % level. In column 3, the interaction term comprises of private credit and competitiveness of executive recruitment. In this case, we can reject the null at the 1 % level (estimate of −0.04 and standard error = 0.02) that the effect of financial development on growth does not vary with within-country variations of competitiveness of executive recruitment. In column 4, the interaction term between private credit and competitiveness of political participation is negative and significant at the 1 % level. In column 5, we use the Freedom House component of civil liberties interacted with private credit. In the Freedom House database, civil liberties are associated with the rule of law, individual freedom to organize, freedom of expression, and freedom to one’s belief. The estimate on the interaction term is again negative and significant at the 5 % level. We conclude from Table 6, columns 2–5 that the impact of the components of democratic institutions does not enhance the effect that financial development has on growth. This evidence echoes the finding of the composite measure from the Polity IV database and the Freedom House Political Rights Index.
6 Endogeneity concerns Another potential concern with the main result is the issue of endogeneity. The system GMM dynamic panel estimator, however, treats all explanatory variables as endogenous and uses lags as internal instruments to control for the possibility of endogeneity and biases induced by the lagged dependent variable. Lagged values of the explanatory variables provide an alternative to external instruments; as is often the case, it is difficult to find external instruments that are correlated with the regressors and uncorrelated with the error term. Under the assumption of weak exogeneity of the explanatory
123
Does democracy dampen the effect of finance on economic. . . Table 6 Effect of dimensions of democracy on the finance–growth relationship (1) Pol2
(2)
(3)
(4)
(5)
5.25 (2.47)**
Private credit Private credit * Pol2
0.17
0.17
0.14
0.25
0.23
(0.06)***
(0.07)***
(0.04)***
(0.07)***
(0.08)***
−0.18 (0.07)**
Constraints on executive
0.47 (0.33) −0.03
Private credit * constraints on executive
(0.01)** Comp of executive recruitment
1.11 (0.49)** −0.04
Private credit * Comp of executive recruitment
(0.02)*** Comp of political participation
1.54
Private credit * comp of political participation
−0.06
(0.59)**
(0.02)*** Civil liberties
10.24 (4.10)** −0.34
Private credit * civil liberties
(0.13)*** Lagged (log GDP pc) Govconsumption Log (1 + inflation) Trade openness Debt
−0.50
−0.36
−0.55
−0.42
−0.36
(0.31)
(0.29)
(0.30)*
(0.30)
(0.28)
−0.02
−0.01
−0.00
−0.02
−0.05
(0.06)
(0.06)
(0.05)
(0.06)
(0.07)
0.16
0.22
0.17
0.34
0.28
(0.33)
(0.33)
(0.35)
(0.35)
(0.29)
0.01
−0.00
0.00
−0.00
−0.01
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
−0.03
−0.06
−0.04
−0.03
−0.08
(0.07)
(0.07)
(0.07)
(0.07)
(0.07)
FDI (net inflows)
0.06
0.09
0.09
0.14
0.32*
(0.17)
(0.17)
(0.15)
(0.15)
(0.17)
Constant
1.20
0.95
3.08
−1.70
−1.58
(2.41)
(2.48)
(2.33)
(2.77)
(3.40)
123
K. Williams Table 6 continued (1)
(2)
(3)
(4)
(5)
Observations
313
305
305
305
314
Countries
78
76
76
76
76
No. of instruments
28
28
28
28
28
Hansen p value
0.64
0.35
0.41
0.58
0.72
AR2 p value
0.88
0.11
0.13
0.39
0.27
Model specifications have full year and country fixed effects. Columns 2–5 are interaction models between private credit and constraints on the executive, competitiveness of executive recruitment, competitiveness of political participation, and civil liberties. **, *** significant at 5 and 1 %. The Windmeijer (2005) small sample correction is used to compute two-step robust standard errors. Robust standard errors in parenthesis
variables, the regressors are uncorrelated with future realization of the error term; the Hansen test evaluates the validity of this assumption. In all cases, Tables 1, 2, 3, 4, 5 and 6, the Hansen test does not reject the validity of the internal instruments, and the AR (2) test supports the assumption of no second-order serial correlation in the difference error term. In the context of the main result, Acemoglu et al. (2013) find that democracy as an explanatory variable is unlikely to be endogenous when lagged GDP per capita is also included as a regressor. Furthermore, the interaction variable is less likely to be endogenous compared with a single variable, because even if one variable in the interaction term is correlated with the error term, it requires the other variable to be simultaneously correlated with the error term for the interaction term to be endogenous (Aghion et al. 2009). Finally, though it is difficult, however, to completely eliminate endogeneity concerns, the paper performs a number of robustness checks and the main result remains stable. And the fifth year data reduce serial correlation bias.
7 Conclusion and policy implications The recent financial crisis has renewed interest in the relationship between financial development and economic growth. This paper contributes to the literature on this subject. Researchers are divided over the contribution of financial development to economic growth (Arcand et al. 2012; Beck and Levine 2004; Levine et al. 2000; Lucas 1988; Rousseau and Wachtel 2011). The inconclusive finding in the literature may be because researchers do not consider the role of political institutions in the relationship between financial development and economic growth. The key insight of this paper is that political institutions are important in explaining the growth effect of financial development. The paper uses panel data estimator for the period 1982–2011 and 78 developing and emerging economies and finds that democracy mitigates the growth effect of financial development. This continues to be the case whether we use the Polity2 composite measure or the dimensions of democracy. This finding may be explained by weak democratic institutions in developing and emerging economies, which makes it likely for policies to be captured by competing interests, with negative effects on economic outcomes. It is useful to point out, however, that the finding does not imply that democratic institutions are bad for the growth effect of
123
Does democracy dampen the effect of finance on economic. . .
financial development. Acemoglu et al. (2013) and Jaunky (2013) find that democracy improves growth. In fact, the paper’s reading of the evidence is that there are economic gains from improving democratic institutions in developing and emerging economies. The main finding has important policy implications for developing and emerging economies. Weak democratic political institutions across developing and emerging economies make it more likely for development policies and economic outcomes to reflect the preferences of special interest groups. This special interest politics is common across Africa, but exists in other developing and emerging economies. In general, these special interest groups do not support growth, because growth leads to decentralization of political power, hence the pattern of weak democratic political institutions and thus inefficient allocation of resources across developing and emerging economies (Acemoglu and Robinson 2012). Public investment in education is one channel through which a society can promote pluralistic political institutions where the median voter can hold political elites accountable. And because democracy promotes growth (Jaunky 2013), growth is likely to be higher where strong political institutions exist. A further implication is that under weak democratic political institutions, some international capital will flow to countries in order to bribe public officials to extract favorable investment terms and conditions. As they move between countries to gain preferential treatments, these international capital flows can generate economic crisis. Sound regulatory structure of the financial system supported by strong democratic institutions will shield recipient countries from the potentially adverse effects of ‘hot international capital flows.’ With international capital flowing to developing and emerging economies in search of high return (in light of low interest rates in developed countries), a robust regulatory structure of the financial system is especially important. Efforts to improve democratic institutions, and the support they provide for the financial system, in developing and emerging economies should therefore be a main priority for policy makers as there are gains to be made in terms of economic development. Acknowledgments I thank Sophia Preston, Zachary Williams, an anonymous referee, and the editor, Robert Kunst, for helpful suggestions and comments. All errors are, of course, my own.
Appendix Sample Algeria, Argentina, Bahrain, Bangladesh, Bhutan, Bolivia, Botswana, Brazil, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, China, Colombia, Congo Democratic Republic, Costa Rica, Cote d’Ivoire, Dominica Republic, Ecuador, Egypt, El Salvador, Ethiopia, Fiji, Gabon, The Gambia, Georgia, Ghana, Guatemala, Guinea Bissau, Guyana, Haiti, Honduras, India, Indonesia, Iran, Jamaica, Jordan, Kenya, South Korea, Kuwait, Lesotho, Madagascar, Malawi, Malaysia, Mali, Mauritius, Mexico, Mongolia, Morocco, Mozambique, Nepal, Nigeria, Pakistan, Panama, Papa New Guinea, Paraguay, Peru, Philippines, Rwanda, Senegal, South Africa, Sri Lanka, Sudan, Surinam, Swaziland, Syria, Tanzania, Thailand, Togo, Trinidad and Tobago, Tunisia, Turkey, Uganda, Uruguay, Zambia, Zimbabwe.
123
K. Williams
Definition of variables Financial development variables • Bank private credit to GDP (private credit): The financial resources provided to ´ the private sector by domestic money bank as a share of GDP. Source: Cihák et al. (2013). ´ • Liquid liabilities to GDP (liquid liabilities): Broad money or M3. Source: Cihák et al. (2013). • Private credit by deposit banks and other financial institutions to GDP (Pcdbfin). ´ Source: Cihák et al. (2013). • Bank deposit to GDP (Bankdeposit): The total value of demand, time and saving ´ deposits at domestic deposit money banks as a share of GDP. Source: Cihák et al. (2013). Democracy variables • Polity2 Index (Pol2): A composite index of autocracy and democracy, where the autocracy scores are subtracted from the democracy scores. The index ranges from -10 (strongly autocratic) to 10 (strongly democratic). This index is transformed to range from 1 to 0, where 1 is most democratic. Source: Marshall et al. (2011). • Polity2 subindex (Dem2): Institutionalized democracy. The democracy scores range from 0 to 10 and is transformed to 1–0, where 1 is the most democratic. Source: Marshall et al. (2011). • Freedom House Political Rights Index (FH): Ranges from 1 to 7 and transformed to 1–0, where 1 is the highest freedom. Source: Freedom House. Other variables • GDP per capita (Lagged[logGDPpc]): GDP per capita in constant 2000USD lagged one period. Source: World Bank, WDI. • General government final expenditure to GDP (Govconsumption). Source: World Bank, WDI. • Consumer price index (Log[1+inflation]). Source: World Bank, WDI. • Imports and exports to GDP (Trade openness). Source: World Bank, WDI. • Foreign direct investment to GDP (FDI, net inflows). Source: World Bank, WDI • Remittance inflows to GDP (Remittance). Transfers by migrant workers and wages and salaries earned by nonresident workers. Source: IMF. • Total population (Log[population]). Source: World Bank, WDI. • Total debt service to GNI (Debt). Source: World Bank, WDI. • Gross capital formation to GDP (Investment). Source: World Bank, WDI. Summary statistics See Table A1.
123
Does democracy dampen the effect of finance on economic. . . Table A1 This table contains summary statistics Variables
Mean
SD
Min
Max
Polity2 Index
0.58
0.32
0
1
Freedom House Index
0.52
0.31
0
1
0
1
0.41
0.36
Private credit
Polity2 subindex
26.57
23.57
0.37
137.45
Bank deposit
30.69
22.34
1.39
122.19
Liquid liabilities
39.61
26.68
2.95
170.61
Private credit by deposit bank and other financial institutions
28.42
26.01
0.37
141.40
Government consumption
14.23
6.41
3.21
76.22
2.26
1.06
1.21
7.67
56.73
33.20
2.96
193.62 42.64
Log (1 + inflation) Trade openness Debt
5.04
4.72
0.03
Net FDI inflows
2.51
4.68
−8.58
53.81
Log population
16.21
1.70
12.71
21.01 79.03
Remittances
4.30
8.17
0.00029
Investment
22.48
9.25
1.57
74.82
6.88
1.19
4.41
10.12
Log GDP per capita
References Abdih Y, Chami R, Dagher J (2012) Remittances and institutions: are remittances a curse? World Dev 40(4):657–666 Acemoglu D, Johnson S, Robinson J, Yared P (2008) Income and democracy. Am Econ Rev 98(3):808–842 Acemoglu D, Naidu S, Restrepo P, Robinson J (2013) Democracy does cause growth. NBER working paper series no. 20004 Acemoglu D, Robinson J (2008) Persistence of power, elites, and institutions. Am Econ Rev 98(1):267–293 Acemoglu D, Robinson J (2012) Why nations fail: the origins of power, prosperity, and poverty. Crown Business, New York Aghion P, Howitt P, Mayer-Foulkes D (2005) The effect of financial development on convergence: theory and evidence. Q J Econ 120(1):173–222 Aghion P, Bacchetta P, Ranciere R, Rogoff K (2009) Exchange rate volatility and productivity growth: the role of financial development. J Monet Econ 56(4):494–513 Alesina A, Rodrik D (1994) Distributive politics and economic growth. Q J Econ 109(2):465–490 Arcand J, Berkes E, Panizza U (2012) Too much finance? IMF working paper, WP/12/161 Arestis P, Demetriades P (1997) Financial development and economic growth: assessing the evidence. Econ J 107(442):783–799 Bagehot W (1873) Lombard street: a description of the money market. McMaster University Archive for the History of Economic Thought Barro RJ (2012) Convergence and modernization revisited. NBER working paper series no. 18295 Beck T (2013) Finance, growth and fragility: the role of government. CEPR discussion paper series no. 9597 Beck T, Levine R (2004) Stock markets, banks, and growth: panel evidence. J Bank Financ 28(3):423–442 Becker G, Stigler G (1974) Law enforcement, malfeasance, and compensation of enforcers. J Leg Stud 3(1):1–18 Bhattacharyya S (2013) Political origins of financial structure. J Comp Econ 41(4):979–994
123
K. Williams Bhattacharyya S, Holder R (2014) Do natural resource revenue hinder financial development? The role of political institutions. World Dev 57:101–113 Blundell R, Bond S (1998) Initial conditions and moment restrictions in dynamic panel data models. J Econom 87(1):115–143 Catrinescu N, Leon-Ledesma M, Piracha M, Quillin B (2009) Remittances, institutions, and economic growth. World Dev 37(1):81–92 Cavallo AF, Cavallo EA (2009) Are crises good for long-term growth? The role of political institutions. J Macroecon 32(3):838–857 Cecchetti SG, Kharroubi E (2013) Why does financial sector growth crowd out real economic growth? Bank for International Settlements Cecchetti SG, Kharroubi E (2012) Reassessing the impact of finance on growth. Bank for International Settlements Chami R, Fullenkamp C, Jahjah S (2005) Are remittance flows a source of capital for development? IMF Staff Pap 52:55–81 ´ Cihák M, Demirgüˇc-Kunt S, Feyen E, Levine R (2013) Financial development in 205 economies, 1960 to 2010. NBER working paper series no. 18946 Choong C (2012) Does domestic financial development enhance the linkages between foreign direct investment and economic growth? Empir Econ 42(3):819–834 De Gregorio J, Guidotti PE (1995) Financial development and economic growth. World Dev 23(3):433–448 Demetriades P, Hussein KA (1996) Does financial development cause economic growth? Time-series evidence from 16 countries. J Dev Econ 51(2):387–411 Huang Y (2010) Political institutions and financial development: an empirical study. World Dev 38(12):1667–1677 Jaunky VC (2013) Democracy and economic growth in sub-Saharan Africa: a panel data approach. Empir Econ 45(2):987–1008 Khawaja AI, Mian A (2005) Do lenders favor politically connected firms? Rent provision in an emerging financial market. Q J Econ 120(4):1371–1411 King RG, Levine R (1993) Finance and growth: Schumpeter might be right. Q J Econ 108(3):717–738 Levine R (2005) Finance and growth: theory and evidence. In: Aghion P, Durlauf S, Durlauf N (eds) Handbook of economic growth. Elsevier Science, Amsterdam, pp 865–934 Levine R, Loayza N, Beck T (2000) Financial intermediation and growth: causality and causes. J Monet Econ 46(1):31–77 Levine R, Renelt D (1992) A sensitivity analysis of cross-country growth regressions. Am Econ Rev 82(4):942–963 Lucas R (1988) On the mechanics of economic development. J Monet Econ 22(1):3–42 Mankiw GN (2010) Macroeconomics. Worth Publisher, New York Mankiw GN, Romer D, Weil DN (1992) A contribution to the empirics of economic growth. Q J Econ 107(2):407–437 Marshall MG, Jagger K, Gurr TR (2011) Polity IV project. University of Maryland Nickell S (1981) Biases in dynamic models with fixed effects. Econometrica 49(6):1417–1426 Olson M (1982) The rise and decline of nations. Yale University Press, New Heaven Papaioannou E, Siourounis G (2008) Democratisation and growth. Econ J 118(532):1520–1551 Persson T, Tabellini G (1994) Is inequality harmful to growth? Am Econ Rev 84(3):600–621 Rioja F, Valev N (2004a) Finance and the sources of growth at various stages of economic development. Econ Inq 42(1):127–140 Rioja F, Valev N (2004b) Does one size fit all?: A reexamination of the finance and growth relationship. J Dev Econ 74(2):429–447 Rousseau PL, Wachte P (2002) Inflation thresholds and the financial-growth nexus. J Int Money Financ 21:777–793 Rousseau PL, Wachtel P (2011) What is happening to the impact of financial deepening on economic growth. Econ Inq 49(1):276–288 Schumpeter JA (1911) A theory of economic development. Harvard University Press, Cambridge Tavares J, Wacziarg R (2001) How democracy affects growth. Eur Econ Rev 45(8):1341–1378 Windmeijer F (2005) A finite sample correction for the variance of linear efficient two-step GMM estimators. J Econom 126(1):25–51 Zhang J, Wang L, Wang S (2012) Financial development and economic growth: recent evidence from China. J Comp Econ 20(3):393–412
123