Differences in political orientation and foreign aid utilization in Africa

ABSTRACT

This study examines the differences in political orientation and foreign aid effectiveness in Africa. The findings confirm our hypothesis that political parties extend their differences in orientation in utilizing foreign aid in Africa. Precisely, our results show that leftists are more likely to use aid efficaciously on pro-poor growth. On the contrary, rightists are good managers of aid on private sector growth projects. Our results provide a convergence path on foreign aid effectiveness, which has been dominated with contrasting results in the existing literature. The results also give a grasp of how foreign aid can better be allocated to Africa.

KEYWORDS:

• Africa
• foreign aid
• leftist
• political orientation
• rightist

1. Introduction

Existing literature on aid-effectiveness provides evidence on the role of politics in aid allocation from the donor perspective (Clemens, Radelet, and Bhavnani 2004; Dreher, Minasyan, and Nunnenkamp 2015; Fleck and Kilby 2006; Minasyan, Nunnenkamp, and Richert 2017; Smets, Knack, and Molenaers 2013; Therien and Noel 2000; Tingley 2010). The fairly extensive studies that investigated political ideologies/orientation¹ focused on the use of aid by donor countries in establishing allies and global powers. A few studies also provided evidence on the proximity of donor-recipient political ideology on aid effectiveness (Dreher, Minasyan, and Nunnenkamp 2015, Smets, Knack, and Molenaers (2013); Lskavyan 2014). Hence, our current study answers the question: Does the political orientation of an aid recipient government affect the use of foreign aid in Africa?

We argue that our study fills an important gap in the aid-effectiveness literature for the following reasons. First, political orientations affect all government policies including the use of foreign aid (Bjørnskov 2000; Blinder and Krueger 2004; Boone 1996; Facchini and Melki 2014; Lskavyan 2014; Nincic and Ramos 2010; Tawiah, Barnes, and Zakari 2019). Regardless of the conditionality on foreign aid, recipient governments are likely to use distortionary policies to redirect funds towards projects that support their political agenda. Political parties, whether leftist or rightist usually revert to their primary ideology of socialism or capitalism in delivering their promises. They revert because these are promises upon which they are voted for and are invariably linked to how they want society to be organized. As argued by Lesschaeve (2017), leftist-rightist placement is a better predictor of policy positions of political parties.

Second, our studies respond to the repeated calls by scholars such as Quibria (2014) to examine why, how and when foreign aid works in a particular country. Quibria (2014) argues that existing literature on aid effectiveness has not paid particular attention to policies and politics of allocation and use of foreign aid. Thus, the disjunction between aid effectiveness in literature and that of practices can be explained through other factors, including the political orientation of the recipient country.

Third, most African countries depend on foreign aid to finance their infrastructural development. This relationship between foreign aid and Africa’s development has led to growing concerns about the effectiveness of foreign aid in Africa in relation to its impact on the recipient economy (Loxley and Sackey 2008; Wako 2011; Wamboye, Adekola, and Sergi 2013).

To establish the relationship between political orientation and foreign aid utilization, we categorized political parties into leftist and rightist based on Cesi, Keefer, and Scartascini (2017) party orientation classifications, with some modifications. We also used an alternative measure based on the political party’s manifestos and other sources. Our data covers 16 African countries over 21 years. We test the empirical prediction of the impact of foreign aid on different measures, including economic growth, social development indicator (poverty reduction) and market development indicator (private sector investment). The findings confirm our hypothesis that differences in political orientation matter in the utilization of foreign aid in Africa. Precisely we find that left-leaning governments are more likely to use aid effectively on direct pro-poor growth projects. On the contrary, right-leaning governments are good managers of aid on free-market-oriented developments. Our results provide a convergence path on the relationship between aid and political ideologies, which has been dominated with mixed results in the existing literature.

Contrary to prior studies such as Dreher, Minasyan, and Nunnenkamp (2015); Milner and Tingley (2010); Therien and Noel (2000); Tingley (2010), on the political orientation of the government in donor countries, our study focused on the political orientation of the recipient governments. Although our study is close in spirit with Boone (1996) and Islam (2003), it differs by examining political orientation, as compared to government regimes in prior studies.

Our study also contrasts volumetric literature on aid effectiveness as we do not directly examine aid-effectiveness on generic measures of economic growth but precisely on segments of the economy. With the differentiators mentioned above, we update and extend the debate on the utilization of foreign aid in Africa.

Our study gives fair evidence on how political orientation can be used to direct the allocation of foreign aid to the recipient countries for specific objectives. For example, foreign aid for a social project such as poverty elevation will be more effective under a left-leaning government. More so, the results highlight the type of foreign aid that will be of interest to a recipient government. Given their capitalist orientation, rightists are likely to pursue foreign aid that targets private-sector or infrastructure growth.

The rest of the paper is structured into four sections. Review of literature is presented in section 2.0 with subsections on political orientation and foreign aid effectiveness. Research methods are presented in section 3.0. Section 4.0 presents the results and section 5.0 concludes the study.

2. Review of literature

2.1. Political orientation

Political orientation, as defined by Erikson and Tedin (2003), is a set of beliefs about the proper order of society and how it can be achieved. Beliefs and values affect all areas of life just as ideologies affect all fields of human endeavor. Consequently, there are numerous types of ideologies, such as social ideologies, religious ideologies, political ideologies, and economic ideologies. Political orientations have emerged as a significant influence on how different governments pursue their agenda (Jost, Federico, and Napier 2009; Nincic and Ramos 2010). Hence, political orientation has attracted long pedigree of definitions (Gerring 1997). McClosky (1964) defines political ideology as a system of elaborated, integrated and coherent beliefs, which justifies the use of power, identification of good and bad, explanation of the interconnections of politics and other activities in the society. In other words, political orientation is the manner and extent to which political parties cognitively organize their thoughts, and policies according to one or more judgment (Converse 2006; Duckitt 2001; Eagly and Chaiken 1998). Thus, political orientations are fundamentally based on how political parties believe society should be organized, and the best ways to achieve such an ideal society as envisaged by them.

Historical classification of political orientations has been dominated by leftist and rightist categorization. Lesschaeve (2017) argues that leftist-rightist is a prediction of policy positions of parties. At the strategic and policy levels, the distinguishing feature is the party’s level of commitment to rectifying inequality (Bobbio 1997; Therien and Noel 2000; White 2013). Whereas the leftist is inclined to rectify inequality, the rightist is known to be sceptical about it. From the perspective of the ordinary citizen, the definition of leftist and rightist is diverging attitudes towards the significance of government in the economy (Fuchs and Klingemann 1990). With these features, the leftist gives more preference to the government in the development of the country, whereas, the rightist argue for a free market with less government participation. Thus, the leftist attitude supports social equality (socialism), and the rightist supports social hierarchy (capitalism). Although the binary coding of leftist-rightist political orientation has been criticized both theoretically and empirically, (Benoit and Laver 2006; Boone 1996; Carney et al. 2008), it is the most widely used classification in existing literature (Dreher, Minasyan, and Nunnenkamp 2015; Facchini and Melki 2014; Nincic and Ramos 2010).

Empirical evidence has shown how political orientation affects different sectors of an economy. Bjørnskov (2000) suggests that rightist governments have a positive impact on economic growth than the leftist due to their (rightist) preference for less government intervention resulting in high legal quality. In line with this argument, Facchini and Melki (2014) find that the GDP of France increased by 1.10% in 1871–2004 due to a shift from total leftist parliament to rightist parliament. According to Nincic and Ramos (2010), political orientation separate conservatives from liberals on both domestic and foreign policy performance. They argue that conservatives are likely to support punitive foreign policies, while liberals will prefer positive incentives. More so, leftist and rightist attribute different causes to the recent global crisis. For instance, White (2013) concluded that leftists claim that the crisis was caused by injustice and acts of moral and intellectual transgression. On the contrary, the rightists argue that the crisis was caused by disorder and patterns of adhesion to problematic doctrines and practice. Thus, while the leftists attribute the crisis to the free market, the right-wing points to excessive public spending.

In Bangladesh and Philippines, Lewis (2010) finds that governance and management of non-governmental organizations were influenced by the political orientation of the parties in power especially when it came to donations and the use of funds. However, in economic strategies, Gregor (2018), argues that in Bolshevik and Fascist systems, the operational features and intended results of both leftist and rightist strategies are similar, hence, categorization of orientation in practices may be a classificatory as well as an intrinsic and cognitive issue.

Turning to the impact of political orientation on foreign aid, Therien and Noel (2000) report that social democratic parties give out more foreign aid than liberals. Moreover, the positive influence of social-democratic parties on foreign aid works stronger in the long run through government institutions and social spending. Therien and Noel (2000) results affirm Lumsdaine (1993)’s opinion that parties with strong socialist concerns strengthen aid while liberal government retrenches it. Similarly, Watkins (1995) finds that US and UK rightist political parties are inclined to cutting aid, which erodes the fundamental rights of the poor. Contrary to Watkins (1995), however, Goldstein and Moss (2005) find that the rightists provide more aid than the leftist government.

Research into ideological proximity of aid donating and receiving countries indicate that aid is less effective when there are differences in the political orientation of donor and recipient countries (Dreher, Minasyan, and Nunnenkamp 2015). Consistent with this argument, Lskavyan (2014) report that foreign aid from the US to leftist recipients are more under the leftist US administration, whereas the rightist government received more aid from the rightist USA government. There is also evidence that the World Bank, a conservative institution which works towards poverty reduction, provision of social services and gender equality issues aligns more with leftist than rightist governments (Dreher, Minasyan, and Nunnenkamp 2015; Smets, Knack, and Molenaers 2013).

Outside democratic institutions, Boone (1996) shows that authoritarian and dictatorship government uses distortionary policies to sway funds to their supporters. The author also finds that liberal government provide more primary welfare than other regimes—nonetheless, all forms of governments support high-income elite class. However, an early study by Williamson (1994) shows no difference in democratic and authoritarian regimes when it comes to the effective utilization of foreign aid. Within electoral studies, political parties are found to be using foreign aid for vote-buying and pursuing their policies (Kuncic 2011; Tawiah et al. 2019). The party in power often uses foreign aid in areas where they can get more votes in the next elections (Jablonski 2014). What is more, Asongu and Jellal (2013) suggest that government in power are likely to redirect and manipulate aid to align with their political agenda and aspirations with the view of gaining a political advantage over their opponents.

2.2. Foreign aid effectiveness

The debate on foreign aid effectiveness is as old as the first foreign aid to recipient countries, yet, the results remain inconclusive, perhaps due to a disjunction between research on aid effectiveness and practices of recipient countries (Quibria 2014). While some scholars find aid to be effective for stimulating capital formation, others argue that aid has virtually caused what is commonly called the Dutch disease and over-dependency in the recipient countries (Dalgaard and Olsson 2008).

On the positive side, an earlier study by Crouch (1973) finds that foreign aid in the neoclassical growth theory has a transitionary effect on the per capita consumption of the underdeveloped country. Dalgaard and Hansen (2017) show an average of 20% return on aid investments. However, Tanaka and Tsubota (2013) find that aid for roads has limited influence in attracting foreign and domestic companies across communes. Alvi and Senebeta (2012) show that aid has significantly reduced poverty in most developing countries, and the impact of aid on poverty is much higher for multilateral aids than bilateral. Dalgaard and Olsson (2008) indicate that more foreign aid means less corruption, implying a positive impact of foreign aid on the level of corruption in Africa partly due to aid conditionality. However, Charron (2011) submit that only multilateral aid after 1997 reduced corruption. Nevertheless, Kangoye (2013) argues that corruption increases when aid is unpredictable, especially in countries with weak institutional environment.

On the flip side, Easterly (2003) provides compelling evidence on how countries like China and India which received small aid have rapidly developed compared with African countries which are the highest recipients of aid over a long time yet show an insignificant rise in development. Consequently, Rajan and Subramanian (2008) argue that foreign aid does not complement economic growth. Easterly (2007) also contends that large aid that flows to developing countries undermines government incentives for accountability. Further, the beneficiary country sees such inflows as windfall resources, which create corruption. Powerful individuals, such as government officials take advantage of these so-called windfall resources and appropriate a large portion of these inflows to themselves at the expense of the mass citizenry (Svensson 2000). The situation is even worse when there are no punitive measures in the allocation of aid (Kosack and Tobin 2006). Interestingly most aid donating countries are aware of this rent-seeking behavior but are restricted to take any a meaningful action due to the UN general policy of restricted interference of donor countries in the domestic affairs of recipient countries (Svensson 2000).

Other scholars argue that the effectiveness of aid depends on factors such as conditionality, ethnic factions, and the policy environment of the recipient country. In Papua New Guniea, Feeny (2005) provides evidence that aid is more effective under the structural adjustment program but not as the single inflow of funds. Similarly, McCormick (2008) claims that the impact of Chinese and Indian aid to Africa depends on the structural and institutional conditions of the recipient government. Ethnic fractionalization can also affect the effectiveness of foreign aid. For instances, Hodler and Knight (2012) find that foreign aid promotes economic development in ethnically homogeneous countries, but it is ineffective and harmful in heterogonous ethnic countries in many sub-Saharan countries. Although Islam (2003) found a significant inverse relationship between aid and the level of development, the author notes that the returns on aid appear to be constant in tinpot countries while totalitarian countries give a high return on aid. The author further highlights the insignificant impact of aid on the improvement of human rights and development indicators.

3. Research methods

3.1. Sampling

Positioning political parties or government on the left-right bipolar scale is not straightforward as parties do not declare themselves as such during campaigns or in manifestos. Four main approaches have emerged from prior studies; expert surveys, opinion polls, data content analysis of party manifestos (CMA) and database of political institutions (DPI) Dinas and Gemenis (2010). Although each approach has its advantage and disadvantage (Budge 2000; Mair 2001; McDonald, Silvia, and Kim 2007), CMA and DPI are widely used approaches (Dinas and Gemenis 2010; Klingemann et al. 2004; Molder 2013).

DPI by Cesi, Keefer, and Scartascini (2017) formerly Beck et al. (2001) provides yearly indices on governance institutions of countries starting from 1975. One such index is the party orientation concerning economic policy. DPI scores countries under five codes, namely; 1 - rightist, 2 - center, 3 - leftist, 0 - no information and NA - no executive. The data covers 47 African countries. We eliminated countries which have 0 and NA scores for almost all the years. This process yields 24 countries over 21 years (1995–2016). Following from our discussion on the operationalization of political ideologies (below), we reclassify DPI 5 codes into 2 codes; 1-leftist and 0-rightist. We used additional sources of information such as manifestos and newswire to collapse DPI 0 and 2 codes into our binary coding. Due to limited data on some countries, the sample size reduced to 16 countries. Details on sample countries are presented in Table 1.

Table 1. Brief of sample countries.

Country	Brief	Political parties in power	Ideology
Benin	Benin got independence from France in 1960. There have seven presidents in 12 elections.	Action Front for Renewal and Development – FARD 1996–2006	Leftist
		Independent 2006–2016	Rightist
Cote d'Ivoire	Cote d’Ivoire gained independence from France in 1960. There have been five presidents and from 11 elections.	Ivorian Popular Front – FPI (2000–2010)	Leftist
		Rally of the Republicans –RDR 2011–date	Rightist
Comoro Island	Comoros gained independence from France in 1975, and there have been four presidents since it formed a Union of Comoros	Convention for the Renewal of the Comoros – CRC 2002–2006 2016–till date	Rightist
		Baobab Movement – BM 2006–2011	Leftist
Cape Verde	Cape Verde gained independence from Portugal in 1975 and have had six elections since 1991.	African Party of Independence of Cape Verde – PAICV 1975–1991; 2001–2011	Leftist
		Movement of Democracy – MpD 1991–2001: 2011–date	Rightist
Ghana 1992–2017	Ghana gained independence from Britain in 1957 and had seven elections since it returned to democracy in 1992.	National Democratic Congress – NDC. 1992–2000 2009–2016	Leftist
		New Patriotic Party –NPP. 2001–2008 2017–date	Rightist
Kenya 1995–2017	Kenya gained independence from Britain in 1963 and had had 12 elections without any Military rule during the period.	Kenya African National Union. 1969–1997	Leftist
		National Rainbow Coalition 2002–2007	Leftist
		Party of National Unity 2007–2013	Leftist
		The National Alliance 2013–2016	Rightist
		Jubilee Party 2017–date	Rightist
Algeria 1994–2017	Algeria gain independence from France in 1962. There has been ten elections with serious of civil unrest.	National Liberation Front. 1978–1994 1999–2019	Leftist
		National Rally for Democracy 1994–1999, 2019–date	Rightist
Liberia	Liberia gained independence in 1847 and had had 51 elections with 25 presidents.	National Patriotic Party– NPP 1986–1990 1997–2003	Leftist
		Unity Party – UP 2006–2018	Rightist
		Congress for Democratic Change – CDC	Rightist
Lesotho	Lesotho gained independence from Britain in 1966. There have been ten elections and 11 presidents, including military regimes.	Basutoland Congress Party. 1993–1997	Leftist
		Lesotho Congress for Democracy 1997–2011	Leftist
		Democratic Congress 2011–2012 2015–2017.	Leftist
		All Basotho Convention 2012–2015 2017 to date	Rightist
Malawi	Malawi gained independence from the British in 1964 and has had five elections since then under four political parties	United Democratic Front – UDF. 1994–2004	Rightist
		Democratic Progressive Party – DPP 2004–2012 2014 – to date	Rightist
		People’s Party 2012–2014	Leftist
Mauritius 1992–2017	Mauritius gained its independence from Britain in 1968. There have been seven presidents since the country became a republic in seven elections.	Labour Party – LP 1992, 2012–2015.	Leftist
		Mouvement Militant Mauricien – MMM 1992–2002	Rightist
		Militant Socialist Movement – MSM 2002–2003 2003–2012	Leftist
Morocco	Morocco gained its independence from France in 1956, and there have been 16 presidents in ten elections.	Istiqlal Party 2007–2011	Leftist
		Justice and Development Party	Rightist
Nigeria (1997–2017)	Nigeria gained independence from the British in 1960. Nigeria has experienced a series of coups. There have been seven elections under the fourth republic.	Peoples Democratic Party – PDP 1997–2007 2007–2010 2010–2015	Rightist
		All Progressive Party – APC 2015 – Til date	Leftist
Senegal 2000–2017	Senegal gained independence from France in 1960. There have been four presidents from ten elections without any military rule.	The Socialist Party of Senegal – PSS 1980–2000	Leftist
		Senegalese Democratic Party (PDS) 2000–2012	Leftist
		Alliance for the Republic. 2012–date	Rightist
Tunisia	Tunisia gained independence from France in 1956. There have been six presidents and ten elections with civil unrest in some years.	Democratic Constitutional Rally 1999– 2004 2004–2009 2009–2011	Leftist
		Congress for the Republic	Rightist
		NidaaTounes 2011–2014	Rightist
Zambia	Zambia gained independence from Britain in 1964. There have been 13 elections and seven presidents with no military rule.	Multiparty Movement Democracy (MMD). 1991–2011	Leftist
		Patriotic Front (PF). 2011-date	Rightist

3.2. Empirical model

Our empirical model is rooted in the null hypothesis consistent with the findings of Jablonski (2014); Asongu and Jellal (2013) that aid recipient governments use distortionary policies to redirect funds towards a project that supports their (recipient countries government) political agenda (see also Bayart 1992; Boone 1996). Hence, each political party uses aid differently, and this has different impacts on the country. Our baseline model is stated below:$\begin{array}{rl}GDPgrowt{h}_{it}=& a+{\text{β}}_{1}Ai{d}_{it}+{\text{β}}_{2}Aid\ast Pol.orientatio{n}_{it}\\ & +{\text{β}}_{3}Politicalorientatio{n}_{it}\\ & +{\text{β}}_{4}Politicalstabilit{y}_{it}+{\text{β}}_{5}Corruptio{n}_{it}\\ & +{\text{β}}_{6}Tradeopennes{s}_{it}\\ & +{\text{β}}_{7}Grosscapita{l}_{it}+{\text{β}}_{8}Inflatio{n}_{it}\\ & +{\text{β}}_{9}Foreigninvestmen{t}_{it}+{\epsilon }_{it}\end{array}$In the above baseline model, GDP growth is the annual increase in gross domestic product. Aid is the net official development assistance (NODA), a country received at a time t. NODA consists of disbursement of loans made on concessional terms and grants by official agencies of the Development Assistant Committee (DAC), multilateral institutions and non-DAC countries to promote economic development. It also includes loans/grants of at least 25 per cent. Political orientation is a binary variable of 1 if the incumbent government is leftist or 0 if rightist. Aid*Pol.orientation is an interaction term between Aid and Political orientation.

Both corruption and political instability have besieged Africa and political instability, therefore, we used data from Polity IV as a proxy for political stability (Political stability), and the perceived corruption index from transparency international as a proxy for corruption (Corruption). Trade openness is the sum of import and export. Gross capital is the total wealth of the country measured by the total capital formation as a percentage of GDP. Inflation is the annualized consumer price index. Foreign investment is the stock of foreign direct investment inflow to the country. $it$ represents country and time respectively and ${\epsilon }_{\mathrm{i}\mathrm{t}}$ is the associated error.

To align our model with the ideologies of the political parties, we develop specific leftist and rightist development indices that flow from their characteristics. Social development index (SI) is calculated as the sum of government expenses on education, healthcare, and rural development. We argue that investment in education, health and rural communities’ captures government commitment toward social developments. This measure is consistent with Boone (1996), Masud and Yontcheva (2005). The capitalist development index (CI), which represents a commitment to the private sector, is also constructed from infrastructural and urban expenditure as a percentage of GDP. Private sector thrives well on major infrastructures, and development in urban centers attracts private investments.

To check the robustness of our results, we employ additional variables. As a pro-poor growth-focused government, leftist activities are expected to reduce poverty and improve the general welfare of the masses. Following (Boone 1996, Burchi and Gnesi 2016; Kovacevic 2011; UNDP 1990) we use Human Development Index (HDI) as a measure of social intervention, which is central to the leftist government.

For a free-market-oriented government, rightist activities are expected to concentrate on private sector growth, which will yield competition. As such, we use the global competitive index (World Economic Forum 2018) as alternative proxies for development performance of the rightist government. Due to data limitation on these alternative measures, we filter our sample period based on the availability of these indices.$\begin{array}{rl}& Socialdevelopmentinde{x}_{it}=a+{\text{β}}_{1}Ai{d}_{it}\\ & +{\text{β}}_{2}Aid\ast Pol.orientatio{n}_{it}+{\text{β}}_{3}Politicalorientatio{n}_{it}\\ & +{\text{β}}_{4}Politicalstabilit{y}_{it}+{\text{β}}_{5}Corruptio{n}_{it}\\ & +{\text{β}}_{6}Tradeopennes{s}_{it}+{\text{β}}_{7}Grosscapita{l}_{it}+{\text{β}}_{8}Inflatio{n}_{it}\\ & +{\text{β}}_{9}Foreigninvestmen{t}_{it}+{\epsilon }_{it}\end{array}$ $\begin{array}{rl}& Capitaistdevelopmentinde{x}_{it}=a+{\text{β}}_{1}Ai{d}_{it}\\ & +{\text{β}}_{2}Aid\ast Pol.orientatio{n}_{it}+{\text{β}}_{3}Politicalorientatio{n}_{it}\\ & +{\text{β}}_{4}Politicalstabilit{y}_{it}+{\text{β}}_{5}Corruptio{n}_{it}\\ & +{\text{β}}_{6}Tradeopennes{s}_{it}+{\text{β}}_{7}Grosscapita{l}_{it}\\ & +{\text{β}}_{8}Inflatio{n}_{it}+{\text{β}}_{9}Foreigninvestmen{t}_{it}+{\epsilon }_{it}\end{array}$

3.3. Econometric procedure

We have accounted for non-linearity, stationarity, and fixed effects to reduce potential misspecification, and spurious regression, (see Clemens, Radelet, and Bhavnani 2004; Dreher, Minasyan, and Nunnenkamp 2015). The un-tabulated results of IPS² (Im, Pesaran, and Shin 2003) unit root tests show that all the variables of interest are stationary at first difference. After establishing the stationarity, we use Pedroni (2004) co-integration test to ascertain the long-run relationship among the series. The results indicated the existence of a long-term equilibrium relationship between the variables. Other diagnostic tests such as VECM, Jacque Bera test of normality, Breusch Godfry serial correlation and heteroscedasticity were used.

4. Results and discussion

4.1. Main results

Table 2 presents the results of the main regression estimations. We start our analyses with the traditional measure of aid-effect by estimating the impact of foreign aid on economic growth without considering political orientation. The results in column 1 of Table 2 shows a significant positive impact of aid on growth at 10 per cent though the coefficient is small. The results imply that foreign aid contributes to the economic growth of most African countries. Our results are reasonable considering the high dependence of African countries on foreign aid and the continuous flow of aid to the continent. Despite this significant relationship, the small coefficient indirectly highlights that the large flow of foreign aid has not fully manifested in the growth of the recipient country. Perhaps due to corruption and mismanagement as evident in the significant coefficient of Corruption variable. The coefficient and significant level of the other control variables appear to conform to standard assumptions. Our results on the impact of aid on GDP is consistent with Dreher, Minasyan, and Nunnenkamp 2015, Hadjimichael et al. (1995), Hansen and Tarp (2000), but contradict Islam (2003), Burnside and Dollar (2000), Bjørnskov (2000).

Table 2. Main regression results.

Variables	GDP		Social Index		Capital Index
	(1)	(2)	(3)	(4)	(5)	(6)
Aid	0.025* (0.021)	0.021** (1.919)	0.151** (2.191)	0.035*** (5.694)	0.095* (1.67)	0.019** (2.039)
Aid*Pol.orientation		0.015 (0.814)		0.066** (2.127)		−0.043** (−2.106)
Political orientation		0.011 (1.003)		0.023** (2.459)		−0.017* (−1.804)
Political stability	0.010 (0.009)	0.009 (0.011)	0.013 (1.231)	0.006 (0.003)	0.004 (0.27)	0.007 (0.002)
Corruption	−0.223*** (−0.123)	−0.213*** (−0.112)	−0.281** (−2.013)	−0.233*** (−2.117)	−0.215** (−2.09)	−0.257** (2.139)
Trade Openness	0.033** (0.017)	0.031** (0.017)	0.009 (0.870)	0.0314 (0.018)	0.053* (1.89)	0.0426* (1.922)
Gross capital	0.057** (0.029)	0.0633** (0.0401)	0.032 (1.503)	0.0616 0.0431	0.157*** (7.33)	0.123*** (8.820)
Inflation	−0.513*** (−5.367)	−0.562*** (−3.450)	−0.195* (−1.66)	−0.231*** (−4.102)	−0.147* (−1.90)	−0.163** (−2.263)
Foreign investment	0.037*** (3.023)	0.019** (2.047)	0.017 (0.48)	−0.014 (0.032)	0.033** (1.99)	0.036** (2.015)
Intercept	0.044*** (4.135)	0.0342*** (6.015)	0.065*** (3.09)	0.073*** (3.490)	0.129*** (9.03)	0.029*** (8.104)
Net effect	NA	NA	NA	0.0659	NA	−0.0450
Adjusted R^2	0.312	0.29	0.34	0.30	0.39	0.33
Observations	317	317	317	317	317	317
Number of Countries	16	16	16	16	16	27
First Diff.	Yes	Yes	Yes	Yes	Yes	Yes
Aid lagged	Yes	Yes	Yes	Yes	Yes	Yes

Notes: Table 2 presents the main results. Aid is the net official development assistance and aid. Political orientation is a dummy variable of 1 if the government in power is leftist or 0 otherwise. Aid*Pol.Orientation is an interaction term for Aid and Political orientation. Political stability is a measure of the absence of violence and political instability. Corruption is the measure of perceived corruption from Transparency International. Trade Openness is measure as a sum of import and export. Gross capital is gross capital formation as a percentage of GDP. Inflation is the annualized consumer price index. Foreign investment is foreign direct investment inflow to the country. t-statistics in parenthesis. *, ** and *** represent 10%, 5% and 1% significant level respectively.

Next, we introduce, political orientation in the aid-effect on GDP model and results are presented in column 2 of Table 2. As stated earlier, Political orientation is a dummy variable of 1 for the leftist government and 0 for the rightist government. The results of both political Political orientation and the moderating term (Aid*Pol.orientation) are positive but insignificant. However, Aid is positive and significant. These results suggest that the impact of foreign aid on the overall growth of the recipient country is similar regardless of the ideological orientation of government in power. Our results align with Gregor (2018), who argues that even if strategies of leftist and rightist may differ; the intended results on the overall economy will be the same. We argue that the GDP growth rate is a sum of all sectors and hence does not actually reflect the performance measure of the leftist and rightist classification. After all, rightists may do well in infrastructure and not so well on social interventions. The problem with GDP as a measure of aid usage is that the negative of one sector offsets the positive of other sectors.

Following the arguments above, we employ two additional measures of a country’s performance which are likely to capture the intended results of each political orientation. That is; social development indicator (Social development index SI) and free-market-oriented development indicator (Capitalist development index CI) investment. We regressed each indicator separately on the variable of interest and results are presented in columns 3–6 of Table 2. As suggested by Bjørnskov (2000), 5% significant level of political orientation coefficient is deemed more than satisfactory given the crudeness of the measure.

In columns 3 and 5, we first tested whether foreign aid has a relationship with socialist and capitalist indicators without including political orientation. Our results confirm the expectation that foreign aid has a significant positive impact on socialist development (SI) and capitalist development indicators (CI).

Our results in column 4 and 6 on both Social development index and Capitalist development index highlight some striking differences compared with results in column 2 on GDP. First foreign aid has a higher significant impact on SI than CI. Moreover, this is expected given that, the World Bank is considered as a social organization with much interest in poverty reduction Dreher, Minasyan, and Nunnenkamp (2015). Second, in comparison to political orientation, the significant positive coefficient of Political orientation and Aid*Pol.orientation shows that leftist government is more likely to use aid for direct social intervention and advancement of pro-poor projects. On the contrary, the significant inverse relationship between Aid*Pol.orientation and Capitalist development index imply that rightist government is most likely to use aid in the development of the favorable free-market environment which is consistent with Watkins (1995) argument that aid is an element aimed at integrating southern states into the capitalist world economy.

Third, regarding the control variables, whereas, Trade openness, Gross capital and Foreign investment are positively and significantly associated with Capitalist development index, they have an insignificant relationship with Social development index. Both Corruption and Inflation are strong disincentives for development in any segment of the economy.

Also, the coefficient and significant level of Aid*Pol.orientation are larger and stronger than Political orientation in column 3 and 4 of Table 1. This implies that the impact of foreign aid on the development indicators is much better when conditioned on the political orientation of the government compared with the linear form of Political orientation. Our results provide evidence to support the hypothesis that the political-ideological orientation of aid recipient government influences the use of aid in different segments of the country.

To put these results into economic perspective, the leftist government is likely to increase spending on direct social intervention and pro-poor growth initiatives including basic education health care by 6.6% when Aid is increased by 1%. Rightists, on the other hand, are more likely to spend and improve private sector environment by 4.3% when Aid increases by 1%. Overall, our result is a reflection on the development agenda and practices of both political orientations. The spin-0ff of the results indicate that governments in aid-dependent countries will seek for aid that resonates well with their political orientation.

Another dimension for estimating the effect of political orientation is the computation of net effect. The net effect shows the additional influence the interaction term has on the dependent variable (Asongu and Nwachukwu 2018a; Asongu, Le Roux, and Biekpe 2017) suggest that the net effect of the interaction term provide additional information. Therefore, following Asongu and Nwachukwu (2018b), Tchamyou and Asongu (2017) Tchamyou (2019) we calculate the net of Political orientation as$\begin{array}{rl}Neteffect& =(MeanofPoliticalconnection\ast \\ & Cofficientofinteractionterm)\\ & +CofficientofPoliticalconnection.\end{array}$The results of the net effect included in Table 2 also confirm how the difference in political orientation affects the use of foreign aid. For example, in column 4 of Table 2, the net effect of leftist government on social development index is 0.0659 [computed as (0.65*0.066) +0.023; where 0.65 is the mean of political connection]. Similarly, the net effect of capital development net is −0.0450 [(0.65*−0.043)−0.017]. The net effect results³ imply that the overall leftist political orientation increases social development with an increase in foreign aid. In contrast, a rightist government will increase capital project by 0.0450⁴ given an increase in foreign aid.

4.2. Robustness check

To test the Robustness of Social development index (SI) and Capitalist development index (CI), we employ the Human Development Index (HDI) and the Global Competitiveness Index (GCI) as an alternative measure for SI and CI, respectively. We repeated all analysis on 16 sample countries, and results are presented in columns 1 and 2 of Table 3. These results confirm that leftists are more likely to use aid in the pursuit of their social intervention agenda, which increases HDI. Rightist also follows their political orientation by spending more funds in improving the competitiveness of the environment of the country as evidenced in the significant inverse coefficient of Aid*Pol.orientation under GCI.

Table 3. Sensitivity and robustness check regression results.

Variables	16 countries sample (alternative measure of dependent variables)		7 countries sample (sensitivity analysis.)
	(1)	(2)	(3)	(4)
	HDI	GCI	SI	CI
Aid	0.012** (2.080)	0.018** (2.134)	0.024** (2.29)	0.021** (2.36)
Aid*Pol.orientation	0.067*** (5.371)	−0.041*** (−3.019)	0.157** (2.016)	−0.081*** (2.039)
Political orientation	0.042** (2.328)	−0.034** (−1.97)	0.031* (2.108)	−0.014** (−1.99)
Political stability	0.014 (0.916)	0.004 (0.806)	0.011 (1.103)	0.008 (0.704)
Corruption	−0.314*** (−5.023)	−0.335** (−2.216)	−0.121** (−2.003)	−0.135** (−1.97)
Trade Openness	0.029 (1.021)	0.031*** (0.624)	0.017 (0.413)	0.022*** (7.015)
Gross capital	0.049 (1.027)	0.098** (1.98)	0.032* (1.73)	0.042** (2.128)
Inflation	−0.019** (−2.114)	−0.193 (−0.513)	−0.015** (−2.021)	−0.131 (−0.703)
Foreign investment	0.012 (0.015)	0.058** (2.437)	−0.005* (1.68)	0.038** (2.173)
Intercept	0.162*** (9.091)	0.192*** (7.036)	0.291*** (13.285)	0.152*** (11.21)
Net effect	0.086	−0.061	0.133	−0.067
Adjusted R^2	0.295	0.275	0.321	0.365
Observations	291	291	137	137
Number of Countries	16	16	7	7
First Diff.	Yes	Yes	Yes	Yes
Aid lagged	Yes	Yes	Yes	Yes

Notes. Table 3 presents the results of the robustness and sensitivity analysis. Years for HDI and GCI starts from 2004–2017 due to the availability of data. Aid is the net official development assistance and aid. Political orientation is a dummy variable of 1 if the government in power is leftist or 0 otherwise. Aid*Pol.Orientation is an interaction term for Aid and Political orientation. Political stability is a measure of the absence of violence and political instability. Corruption is the measure of perceived corruption from Transparency International. Trade Openness is measure as a sum of import and export. Gross capital is gross capital formation as a percentage of GDP. Inflation is the annualized consumer price index. Foreign investment is foreign direct investment inflow to the country. t-statistics in parenthesis. *, ** and *** represent 10%, 5% and 1% significant level respectively.

Although the coefficient of column 1 and 2 of Table 3 are smaller than column 3 and 4 of Table 2, the direction of the relationship between a variable of interest is similar to original estimations which confirm the robustness of model and proxies.

4.3. Sensitivity analyses

The impact of the difference in political orientation on the use of foreign aid could be affected by the political system in which the recipient government operated. Knowing very well the dominance of a single party and the lengthy stay of presidents in Africa, we use a sub-sample of seven multiparty democratic countries⁵ which have a good track record of a successful change of government over a long time. The results of this additional analysis are presented in column 3 and 4 of Table 3. The coefficient of both political, ideological proxies is much higher and significant than the original estimation in Table 1. Arguably, the difference in political-ideological orientation is best seen in a multiparty democratic system. It also implies that; multiparty democratic countries are better managers of aid in Africa, which is consistent with Boone’s (1996) findings that democratic countries perform better than the least democratic regimes.

5. Conclusion

The use of foreign aid has been extensively and controversially discussed, yet there is no compelling empirical evidence of an aid-growth pattern (Quibria 2014). This is partially due to the neglect of the political orientation of the recipient government. In this study, we have examined the impact of recipient governments’ political orientation on the use of foreign aid. We used data from 16 African countries over 21 years with alternative measures of variables.

Our analyses reveal the following findings. First foreign aid has a significant impact on economic growth regardless of the government being leftist or rightist because GDP is an overall measure of economic growth affected by negative and positive of different segments. However, foreign aid is used differently under each political orientation. Second, each political party uses foreign aid on a specific area of the county’s development in accordance with its political orientation. Leftists political parties are more likely to use aid on pro-poor growth projects such as poverty reduction, basic education, and health services. On the other hand, rightist political parties, are likely to use aid effectively in promoting their free-market and private sector growth agenda. Precisely, rightists are more likely to use aid in promoting private sector growth and investing in long-term infrastructure projects.

Our results imply that the political orientation of the recipient country matters in the debate of aid-usage. Hence, the expected use of aid in a country depends on the political orientation of the government in power. Consequently, our findings provide some convergence in the existing diverse literature on the topic. As the results show, foreign aid effectiveness yields mixed results because existing literature is limited to economic growth as the only measure of development. We argue that economic growth is a holistic measure which does not truly capture the particular intent of leftist and rightist classification. Our results which capture these particular measures bring some convergence in the debate making an incremental contribution to the existing literature. Equally, our results are robust to alternative measurements of development indicators and subsampling.

Although our findings provide a fair understanding of the use of foreign aid in Africa and are consistent with standard assumptions, they must be interpreted with care owing to two limitations. The crudeness of coding political ideologies as well as limited data on single party domination in several African countries.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Political ideology and political orientation are used interchangeably.

2 We use IPS because of its accuracy on short period data.

3 The net effect should be interpreted cautiously given the binary measure of political orientation.

4 Note that the results are interpreted in the inverse form because political orientation is binary where 1 is for leftist and 0 for rightist.

5 Cape Verde, Cote d’Voire, Ghana, Kenya, Malawi Senegal and Zambia.