A plethora of empirical studies attempted investigating the causal relationship between corruption and poverty. But the outcome of those studies produced mixed results in a regression model that captured only two variables. In this investigated bivariate model, economic growth as an important variable was not included. In fact, omitting such an important variable could seriously affect the research output. Therefore, the objective of this study is to fill the research gap by investigating the trivariate causality for Nigerian economy by incorporating economic growth as a third variable in the existing corruption and poverty bivariate model. The results of this study indicated that there is an existence of co-integration among the variables under investigation characterizing a long run causality relationship among corruption, economic growth and poverty in Nigeria. The results of the dynamic economic growth model indicate the degree of growth influence linkage on corruption.
The findings are suggesting that policies that enhance economic growth are expected to be precise in reducing corruption and poverty in Nigeria. Therefore, the government needs to give importance on the development of critical sectors that are growth enhancing. K EY WORD: Economic growth, poverty, corruption, co-integration, causality. 1.0 INTRODUCTION The Empirical studies that examined the causal relationship between corruption and poverty using bivariate granger causality frame works (such as Gupta et al., 2002; Negin, Abd Rashid, & Nikopour, 2010; Gyimah-Brempong, 2006; You and Khagram, 2004) have produced mixed results. The study of Negin, et al. (2010) suggested that there is a significant relationship between corruption and poverty, without including economic growth as the channel of transmission between corruption and poverty. Meanwhile, Gupta et al. (2002) found significant relationship between corruption and poverty using sample of African countries.
Arora (2009) provided evidence for reverse causality in their finding where they reached a conclusion that the imbalance between the poor and the rich make the poor weaker and incapable to monitor the rich. Aliyu and Elijah (2008) asserted that corruption and economic growth are cointegrated in Nigeria. But, their study failed to integrate the poverty variable into their regression analysis in order to determine the simultaneity of influence. Apart from methodological problems, these approaches have suffered from a limited number of observations and excluded relevant variables which may have a significant relationship with the two variables being considered which is the economic growth variable.
The study of Demery and squire (1996) is one of the early studies that used trivariate model but his study focused on inequality economic growth and poverty, and the authors found that inequality is inversely related to poverty implying that inequality affect poverty through mean income. This highlights the needs of including economic growth as a third variable in the relationship between corruption and poverty. Another approach that tried to establish the importance of economic growth variable even though his study uses bivariate model was the work Lambsdorff (2007). This study integrated the ratios of GDP to capital stock and proxy this ratio of GDP to capital stock as macroeconomic measure of the average capital productivity. The result indicated that there was an existence of significant negative impact of corruption towards this ratio. In fact, it is evident that the role of economic growth in corruption and poverty relationship has been neglected in corruption poverty literatures (e.g. Bardhan, 2006; Mauro, 1995; Knack and Keefer, 1997; Tanzi and Davoodi, 1997).
Moreover, majority of these studies are based on cross section data which cannot capture country specific issues. This paper differs significantly from existing body of literatures by introducing economic growth variable in the corruption-poverty relationship for Nigeria. We also address the general methodological problems of simultaneity by taking all variables in equal footing, without any prior distinction between endogenous variables. Here, we treat all variables as endogenous. We employ co-integration and vector error correction techniques based on the argument that if variables in the model are cointegrated and each is individually 1(1) that is integrated of order 1 (i.e. each is individually no stationary), then either one of the variable Granger- causes the remaining variables or the influence of the remaining variables Granger causes the latter variable. However, in order to address the issue of exogeneity identification in causality, we followed the argument in Gujarati (2004) which said that causality may not be necessary or sufficient estimator of establishing weak exogeneity; on that basis we refuse to develop a theoretical model that has to do with assumption of exogeneity amongst economic variables.
We, therefore, follow the approach of Negin et al. (2010) to directly test the causal relationship among corruption, poverty and economic growth. With this introduction the remaining sections of the paper cover why economic growth is important in the corruption poverty model in section two of the study. While section three dwells on Methodology. Then, in section four, we provide analysis on our empirical results and their major policy implication. Finally, we end the paper in section five with summary and conclusion. 2.0 CORRUPTION, POVERTY AND ECONOMIC GROWTH RELATIONSHIP The channels of causation between corruption and poverty have been intensively discussed in the literature. The transmission mechanism is that corruption affects poverty by first impacting on economic growth variables, while economic growth variables in turn generate impact on poverty.
This implies indirect linkage between corruption and poverty. The issue of controversy in both theoretical as well as empirical literature is on the direction of influence. The liberalists argued on the one hand that the direction runs from corruption to economic growth and to poverty (e.g. see Kaufmann et al., 2005, 2006 and 2009; Mauro, 1995). On the contrary, the heterodox, and statists argued that the direction of influence is from growth to corruption and finally to poverty. This means that improvement in economic growth is what is required for poor countries to achieve, before corruption and poverty can be reduced (e.g. see Khan 2006, 2007, 2009, and 2010). The consensus argument is that corruption constraints economic growth by hindering both internal and external productive investment, through tax and discouraging entrepreneur manpower development which will, in turn, reduce economic growth and decline in economic growth lead more poverty.
In another way, corruption reduces the quality of social infrastructures such as roads, electricity, housing, and water supply. Corruption also diverts marginal talent into rent seeking, which discourages the composition of public expenditure. Corruption also reduces tax revenue where entrepreneurs are diverted into informal arrangement of excessive rent taking which reduces taxes in exchange due to excessive rent taking by the officials. Another linkage between corruptions and economic growth is that corruption may lead lower output due to low level of investment and low level of output. The empirical evidences discussed, so far, were emphasizing on the need to take country specific study in order to investigate in more detail the causal linkage among corruption, poverty and economic growth simultaneously to help in drawing policy implication to Nigeria. The civilian government that came into power in 1999 has lunched Poverty Alleviation Programme (PAP) (World Bank, 2001).
The PAP precisely seeks create employment to 200,000 unemployed youth, provide access to credit to peasant farmers, provide access to education to 51percent to 70 percent adult by the year 2003, improve health care delivery system from 40 percent to 70 percent by the year 2003, improve Medicare delivery in the area of children immunization from 40 percent to 100 percent improve supply of rural water supply and provide rural electrification from 30 percent to 60 percent, provide technical training to bunches of unemployed tertiary institutions graduates and establish avenue of small scale development. Due to poor governance and institution, PAP did not achieve its desired objectives, thus, government phased it out and introduced new program called (NAPEP) in 2001 (Aliyu, 2001).
The new program came up with four broad of structures strategies which included Youth empowerment schemes (YES) aimed to provide youth employment opportunities and to provide technical training. This particular program was sub- divided into Capacity Acquisition Scheme, Mandatory Attachment Scheme. While the remaining three other program of NAPEP included Provision of Credit Facilities Program, Sustainable Development, and Harnessing of Solid Mineral Resources. The main target of poverty reduction strategies is to eradicate poverty completely from Nigeria by the year 2010. In mid-2004, government launched another reform policy called National Economic Empowerment Development strategies anchored on institutional and governance reforms with tripod approach wealth generation and general employment generation and addressing corruption and efficient property right, democratic accountability and rule of law.
The ideas here is that anti corruption will lead to market efficiency and market will provide faster development and faster poverty reduction. With this in mind, this study intends to examine whether institutional reforms through anti corruption governance agenda can ensure efficient market in the context specific to Nigeria. Is market enough efficient to ensure economic growth and poverty reduction at an early stage of economic development? 3.0 Econometric methodology and characteristic of the data In order to address the problems of measurement error in the previous studies, we employed the technique of co-integration in order to determine the long run relationship among variables.
However, having long run causality does not mean any short run causality. As such, we employed the error correction mechanism in order to adjust short run equilibrium into long run dynamic equilibrium. This is because establishing cointegration by using VAR approach at a first differencing may sometime provide wrong results. Therefore, we introduced VECM as an alternative way of establishing causality. This will enable us to understand the short run dynamic equilibrium relationship among variables.