This Is AuburnElectronic Theses and Dissertations

Three Inter-Related Essays in Applied Economics: A Quasi-Experimental Analysis of Ponzi-Financial Institutions




Musah, John Mustapha

Type of Degree

PhD Dissertation


Agricultural Economics and Rural Sociology

Restriction Status


Restriction Type


Date Available



This dissertation consists of three inter-related essays in applied economics, which delves deeper into different research questions with varying econometric models on the recent financial sector crisis in Ghana, sub-Saharan Africa. The emergence of financial institutions in recent times, coupled with lending models in the field of microfinance is celebrated as significant improvements, which further enhances the perceptive miracle of enabling previously un-bankable households by creating “safe nets” to replace the missing links that excluded them from access to more traditional forms of financial services like savings, credit, and investment. In recent times however, financial opportunists took advantage of the vulnerable households by creating unsustainable investments within Microfinance Institutions, Savings and Loans Companies, commercial banking institutions, and other financial houses that functioned as de facto ponzi schemes. Their ultimate goal was to exploit vulnerable households for selfish monetary advantage. The influx of these ponzi-financial institutions (PFI’s) has rendered Ghana’s financial sector unstable, leading to loss of confidence in the banking sub-sector. The government of Ghana and supervisory agency have done very little to safeguard depositor’s funds. Chapter 1 determines the impact of PFI’s on household savings, using a three-wave pooled cross-sectional dataset from World Bank’s Living Standards Measurement survey to analyse this impact by employing a difference-in-differences estimation strategy with regional and time fixed effects. The findings suggest that the activities of these PFI’s induced all households to increase their average savings by 502.00 Ghana cedis ($125). Further evidence shows that the non-poor income group also increased their average savings by 414.50 Ghana Cedis ($103.63). It is interesting to note that urban households increased their annual savings by an average of 914.30 Ghana Cedis ($228.57), relative to rural dwellers. The results in this chapter were further strengthened by a number of robustness checks. Elasticities of savings and investment in the post-closure of PFI’s would be negatively affected. In chapter 2, we employ a two-wave pooled cross-sectional dataset to investigate the post-closure of these PFI’s on household savings. We follow the approach adopted by Ravallion et al (2005) to execute a difference-in-difference-in-differences estimation strategy with geographical and time fixed effects. The findings adduced from this chapter suggests that post-closure of PFI’s led to a significant reduction in household savings among all households and the non-poor, consistent with previous findings. On the average, all households and non-poor households reduced their average savings by 788.70 Ghana cedis ($157.00) and 888.70 Ghana cedis ($222.18) respectively. Contrary to earlier narrative, previous period savings shows a negative relationship with household savings behavior. The results of this chapter were further strengthened by a number of robustness checks and falsification tests. And finally, chapter 3 examines the impact of PFI’s on household credit acquisition, by employing a Propensity Score Matching technique to test for and measure the treatment effect on the observations of interest. This is done by using the greedy matching approach to run through the list of treated units and select the closest eligible control unit to be paired with each treated unit. The results obtained from the analyses indicate that the presence of PFI’s (2006-2013) reduced credit acquisition on the average by a 3-percentage point among households, while the post-closure effect (2013-2017) increased credit acquisition on the average by a 2-percentage point. The combine effect shows that households reduced their credit acquisition on the average by a 2-percentage point between the period 2006-2017, which thus indicate that activities of PFI’s negatively impacted credit-acquisition.