Corruption and Levels of Economic Development: A Cross-Country Assessment with Special Reference to Africa

Cross-country studies of corruption have reported a negative relationship between corruption and economic development. Yet the causal direction is unclear. We posit that the relationship between corruption and economic development is innately interdependent or bidirectional because they coevolve within a country. Moreover, corruption and economic development interact with governance, culture, and social development, forming an interdependent system of quintuple helix. We examine these factors with a sample of 157 countries, as well as a subsample of 46 African countries. The results indicate that as countries move away from a factor-driven economy towards an innovation-driven economy, they observe lower levels of perceived corruption. However, the relationship between actual incidence of corruption and economic development is highly sensitive to covariates in governance, culture, and social development, consistent with interdependence among these factors. The factor that consistently explains perceived and actual corruption is a country’s governance systems, measured by institutions or press freedom. and GDP_PPP, and actual level of corruption across countries. Corruption is proxied with WB_bribery1. The results indicate that, in the presence of controls for governance (proxied by Institutions and Press_freedom), culture, and social development, Eco_stage is still significant in explaining the variation in the incidence of briberies, but GDP_PPP is not. When both Eco_stage and GDP_PPP are in the loading on is significantly positive, contrary to the conventional belief that economic development constrains to


Introduction
Whereas corruption practices exist all over the world, the degree thereof differs for different countries. Some countries show lower levels of corruption than others; none, however, shows absence. One inescapable conclusion is that corruption is related to a country's level of economic development. However, the mensional perspective.
There are different forms of corruption affecting different parts of the society, such as public sector, political sector, and private sector; they can be domestic or international [2]. The common feature is that corruption tends to impose and constitute a "tax" on the society, where gains are privatized by the local and global corruption networks and losses from corruption schemes are shared in different degrees by the rest of the society. Corruption deeply affects a country's quest to achieve an inclusive, equitable, and sustainable economy as well as political and social development [3] [4] [5] [6] [7].
Effective governance is an integral component of a country's quest to implement transparency at all levels in its social, economic, and political dimensions. A balanced and effective governance approach will eventually lead to and foster a country's structural transformation. On the other hand, poor levels of effective governance can lead to a number of economic, social, and political disruptions including corruption. Corruption is motivated by individual and institutional choices in a society. These choices, while shaping the expansion of corruption networks, are themselves shaped by the country's political, economic, social, and cultural environment. Thus, the corruption phenomenon needs to be understood from a broader perspective [8].
We posit that the relationship between corruption and economic development is innately interdependent or bidirectional because they coevolve within a country. Moreover, corruption and economic development interact with governance, culture, and social development, forming an interdependent system of quintuple helix. We examine these factors with a sample of 157 countries, as well as a subsample of 46 African countries. The results indicate that as countries move away from a factor-driven economy towards an innovation-driven economy, they show lower levels of perceived corruption. However, the relationship between actual incidence of corruption and economic development is highly sensitive to covariates in governance, culture, and social development, consistent with interdependence among these factors. This pattern is more pronounced among African countries. The factor that consistently explains perceived and actual corruption is a country's governance systems, measured by institutions or press freedom.
Our findings have policy implications for the worldwide battle against corruption and especially the anti-corruption efforts of African countries. Corruption is not a fixed feature of underdeveloped economies; but to combat poverty and corruption, a country must address its governance issues by establishing institutions of law and order and promoting transparency in executive, legislative, and The remainder of the paper is organized as follows. Section 2 discusses the special challenges faced by African countries in economic and social development. Section 3 develops the theoretical framework and describes the empirical methodology. Section 4 reports the empirical results. Section 5 concludes with discussion of limitations and policy implications.

Corruption: An African Perspective
Since the African colonies started gaining independence, Africa has attracted much attention for its endemic corruption. Despite its rich endowment in natural and human resources, Africa remains one of the poorest, undeve-  [12].
Corruption in Africa takes many forms, from "petty corruption", bureaucratic, economic, and "grand corruption." Many cases of "grand corruption" have ties to international networks of corruption. Some foreign companies engage in corruption schemes to protect their interests in African countries. In addition to unlawful financial transactions, trafficking of humans and animals, illegal drugs and arms are also sources of funds for corruption schemes in the region [  There is also collusion of African elites with foreign networks of corruption.
This collusion explains to some degree why a number of structural adjustment programs and policies funded by foreign government and foreign agencies have not resulted in the expected outcome [7] [9]. The aforementioned ills are exacerbated by the conflict of interests with the former colonial powers and devel-

Methodology and Research Design
Given the discussion in Sections 1 and 2, we posit that as a country moves along the path of economic development the causality question becomes more answerable in the sense of "chicken first or egg first". Our supposition is that corruption within a country coevolves with its economy and governance systems; their evolution, and hence the demand and supply of corruption, is subject to a country's social development and cultural tradition. We would like to refer to this framework as the quintuple helix. At the center of the system are three helices: governance, corruption, and economic development. Accountable governance systems is the pivotal force that curbs corruption and promotes economic development. In the absence of a strong governance system, corruption flou- . Corruption in a system of quintuple helix. This figure illustrates a quintuple helix system, in which corruption interacts and coevolves with governance, economy, social development, and culture. At the center of the system are three helices: governance, corruption, and economic development. Accountable governance systems is the pivotal force that curbs corruption and promotes economic development. Corruption, in turns, erodes the governance and checks economic development, while economic development helps alleviate corruption and leads to effective governance systems. Social development affects and responds to the interactions among governance, corruption, and economic development, while all the four strands coevolve with their cultural environment. Since SEM requires a large dataset, which is inherently difficult to obtain for a cross-country study with a focus on Africa, we resort to a parsimonious OLS model to test the quintuple helix framework: where Eco_stage and GDP_PPP are proxies for economic development, Institutions and Press_freedom for governance systems, ELF for cultural environment, and HDI for social development.
More specifically, Eco_stage is the stage of development based on the World Economic Forum's Global Competitiveness Report 2017-2018, coded 1, 3, and 5 for factor-driven, efficiency-driven, and innovation-driven economies, respectively; it is coded 2 for economies in transition from factor-driven to efficien- The country with the strongest institutions ranks first. Press_freedom is the press freedom index of 2015compiled and published by Reporters Without Borders. The country with the highest level of press freedom ranks first. ELF is the Ethno-linguistic fractionalization score between 0 (zero fractionalization) and 1 (highest fractionalization) based on the work of [33] [34] and [35]. HDI is the Human Development Index based on the United Nations Human Development Reports of 2016. It is a composite index measuring average achievement in three basic dimensions of human development-a long and healthy life, knowledge, and a decent standard of living. The country with the highest human development ranks first. We multiply the original numerical ranks of press freedom, institutions, and HDI by −1, so that a greater value denotes greater press freedom, stronger institutions, and higher level of human development. We argue that corruption within a country coevolves with its governance systems such that effective governance constrains corruption but rampant corruption erodes governance. Similarly, corruption slows down a country's social development, while a developed society stunts the demand and supply of corruption. Under these arguments, we expect negative correlations between corruption and the transformed variables of Institutions, Press_freedom, and HDI.
With regard to the dependent variable, we recognize that corruption has many dimensions and comes in different shapes and forms. Given corruption's shady and secret nature, measuring corruption is not an easy task. We employ three WB_bribery2, the percent of firms identifying corruption as a major constraint, also based on [36]. Most data are from 2015 or 2016. Not all measurements are undertaken every year. Thus, if a variable did not have the most current data, we used the last available data. In various specifications of the tests, when an observation (country) misses the data required to compute a variable in the model, the observation is dropped. See the Appendix for more details about the variables.

Results and Analysis
We mentioned earlier that various strands of research suggest that the stage of development and corruption have negative correlation, i.e., higher the development, lower is the corruption. Figure 3 and Figure 4 illustrate the negative relationship between CPI and different levels and stages of economic development. The CPI ranks are plotted for countries grouped by the stage of economic development: factor-driven economies (Stage 1), economies in transition from Stage 1 to Stage 2, efficiency-driven economies (Stage 2), economies in transition from Stage 2 to Stage 3, and innovation-driven economies (Stage 3). Figure 3 is constructed with all countries with available data, and Figure 4, with African countries only. In both cases, the bivariate plots show that as a country moves  away from resource-driven towards innovation-driven economy it becomes less plagued by corruption.   In Table 4, we report the multivariate tests of the relationship between economic development and actual level of corruption across countries. In Panel A,   Table 3, the r-squares also drop substantially, suggesting that actual level of corruption results from more complex factors than perceived corruption. Overall, these patterns indicate that the negative relationship observed between corruption and economic development may be driven largely by some common factors related to a country's political or governance systems. Table 5 and the two panels of Table 6 show the results from the subsample of African countries when corruption is proxied by TI_CPI, WB_bribery1 and WB_bribery2 alternatively. In Table 5, neither Eco_stagenor GDP_PPP is significant in explaining the variation in TI_CPI. The two governance measures (Institutions and Press_freedom) are still consistently significant and having the expected sign. However, Panel B of Table 6 reveals that neither Eco_stage nor GDP_PPP, by itself or together, is significant in explaining WB_bribery2, the percent of firms experiencing bribe payment request, or WB_bribery2, the percent of firms identifying corruption as a major constraint. The only variable statistically significant is Institutions. Taken together, these results indicate that the effect of economic development on corruption is minimal in Africa, where the most salient driving force of both corruption and economic growth is institutions that defines a country's political, regulatory and business environment.
Since the estimates drawn from the OLS regressions on cross-sectional data can suffer from statistical inefficiency because of nonspherical error terms, we undertake heteroscedasticity tests on three specifications of the model with both    [37]. The robust regression produces maximum likelihood estimates to mitigate the influence of outliers. Table 8 and Table 9 summarize the robustness tests on the full sample and the Africa subsample, respectively. Tests presented in Table   8 on the full sample confirm that GDP_PPP, either by itself or in combination with Eco_stage, is not significant in explaining the variability in perceived or actual experience of corruption in multiple regressions. Eco_stage is significant in explaining the variation in TI_CPI and WB_bribery1 but not WB_bribery2. Results from the Africa sample (Table 9) also confirm the OLS finding that Eco_stage is only significant in explaining the variation of WB_bribery1, but not TI_CPI or WB_bribery2. In contrast, governance, especially measured with Institutions, is consistently significant in explaining the variation in both perceived and actual corruption. In summary, the robustness tests confirm the results of OLS regressions.

Discussion and Conclusion
This paper assesses to what extent corruption is related to a country's stage of economic development, and how this relationship is sensitive to a country's governance systems, culture, and social development. We argue that corruption coevolves with a country's political processes (or effectiveness of governance), culture, social norms, and similar factors. Thus, it becomes imperative to approach corruption from a multidimensional perspective. We have taken the most parsimonious viewpoint in operationalizing these various concepts quantitatively. The conceptual difficulties are matched in severity by the difficulties of data and therefore the statistical methods employed. The concepts are nebulous, the indicators are many but imprecise and data are sparse. Therefore, we use regression analysis to extract as much of relationship between explanatory variables and the variable of corruption as possible.
Given the results of the full sample and the African subsample, we surmise that as a country moves along the path of economic development, corruption tends to recede. However, the causality question becomes more answerable in the sense of "chicken first or egg first". Governance related factors such as institutions and freedom of the press appear to be salient across samples and robust to different measures of corruption. As the appendix shows, the institution variable covers a country's political, regulatory, and business environment; freedom of the press represents the legal, political, and economic environments for the press.
Our findings have policy implications for the worldwide battle against corruption and especially the anti-corruption efforts of African nations. Effective governance thrives in developed economies, where the rule of law and transparency permeate the country's economy and political systems, resulting in a functioning state with commitment to lawful use of public goods and services. Corruption is not a fixed feature of underdeveloped economies; but to combat poverty and corruption, a nation must address its governance issues by establishing institutions of law and order and promoting transparency in executive, legislative, and judiciary decisions. Freedom of information and the press are important components for the creation of a transparent, equitable, and inclusive economic, political, and social environment. It is of paramount importance for African countries to invite the participation of their civil societies in their political, social, and economic transformations. The implementation of governance measures must be well aligned with the country's local realities in terms of social development and heterogeneous cultural traditions.