Summary of Key Findings

This thesis examines the effects of market frictions and distortions on allocative inefficiency and aggregate TFP in the Zimbabwean formal and informal manufacturing sectors. The thesis contributes to the growing literature on identifying and measuring misallocation and its implication for aggregate TFP in emerging economies.

The empirical analysis in this thesis makes use of a matched employer-employee dataset of Zimbabwean formal and informal manufacturing firms and workers that was collected between 2015 and 2018. The survey design and implementation of some aspects of the informal sector survey were conducted specifically as part of this thesis. The data contains detailed information for both firms and workers that allows a comprehensive analysis of the thesis objectives.

The thesis constitutes of three main empirical chapters. The first main chapter (Chapter 2) investigates the extent of allocative inefficiency within and between the formal and informal sector firms. The objective is to assess the importance of factor market distortions in contributing to misallocation and the associated consequences on aggregate TFP losses. Using the firm-level dataset for formal and informal sector firms we collected in 2015 and then applying the HK methodology, the study finds the following evidence: first, the results show firm productivity heterogeneity for both the formal and informal sectors. The TFPQ distribution for the small formal firms shows a large thick left tail, suggesting that low- productivity small formal firms do not exit production, thereby creating zombie firms. The existence of zombie firms is an indication of allocative inefficiency. The distribution of the aggregate measure of allocative efficiency, TFPR, show large dispersion, highlighting the existence of misallocation of resources within and between the formal and informal sector.

Second, the results show a positive correlation between the indicator of misallocation (TFPR) and productivity. This indicates that misallocation is considerably higher for higher- productivity firms in both the formal and informal sectors. To understand the sources and nature of distortions, the aggregate measure of misallocation (TFPR) is decomposed into capital and output distortions. The results show that there is a positive correlation between firm productivity and measures of capital and output distortions in the formal and informal sectors.

The results suggest that relatively more productive firms face high output and capital

131

distortions, which induce them to produce lower than optimal output. A comparison between the formal and informal sector firms indicate that informal sector firms suffer relatively more from both the capital and output distortions. While there is a slight difference in the extent of output distortions between these two sectors, capitals distortions are strikingly larger for the informal sector firms. These results stipulate that distortions act as a tax to comparatively more productive firms, causing more productive firms to become inefficiently small while less productive firms become inefficiently large, thereby suppressing aggregate TFP.

Third, the results show that by removing misallocation of resources, aggregate TFP gains of 153.6 percent can be realized. The results further indicate that the elimination of misallocation in the formal sector leads to 156.6 percent improvement in formal sector TFP, while in the informal sector the TFP gains increases by 151.2 percent. The results point out the importance of efficient allocation of resources to improve aggregate TFP.

The chapter concludes by presenting the robustness of the results through the use of WBES as an alternative dataset to our primary dataset. It also uses the OP covariance approach as an alternative to HK methods. The robustness check results corroborate our baseline findings indicating the existence of large misallocation in the Zimbabwean manufacturing sector.

Chapter 3 examines how financial access constraints contribute to misallocation and hinder productivity-enhanced firm performance for the informal manufacturing sector firms. The chapter draws from the panel dataset of informal sector firms that we collected between 2015 and 2018. The chapter first examines the association between financial access constraints and indicators of misallocation and determine if financial constraints amplify aggregate TFP losses via the misallocation channel. It then looks at how financial access constraints impede productivity-enhanced firm performance. Firm performance is measure by employment growth and change in the capital (investment), while measures of financial access constraints are constructed directly from the information on firm financing activities available on the questionnaire.

The results show heterogeneity in access to finance by informal sector firms. Evidence indicates that young and relatively small firms are more financially constrained. Further, results reveal that more productive firms are financially constrained relative to less productive firms.

Empirically, the chapter first shows evidence indicating a significant and positive association between firm TFP and the measures of misallocation, suggesting that aggregate TFP losses due

132

to misallocation are amplified (a result that is consistent with findings in Chapter 2). Second, we find a positive association between financial constraints and indictors of misallocation. This indicates that financial constraints impose an additional source of misallocation. Third, we find a positive association between indicators of misallocation and the interaction between financial constraints and firm TFP. This result signifies that financial constraints further compound aggregate TFP losses through misallocation. Fourth, we find that financial constraints magnify aggregate TFP losses through MRKP, capital distortions and output distortions. Thus, financial constraints act as a tax, leading to firms using less capital relative to labour than is optimal.

With regard to the link between financial access constraints and productivity-enhanced reallocation of employment growth and investment, the results reveal a negative and significant association between financial constraints and firm investment. Financial constraints thus reduce a firm’s investment capacity, leading to shrinking firm performance and, eventually, aggregate TFP. The results further indicate an insignificant relationship between firm productivity and investment. There is also evidence suggesting an insignificant relationship between firm investment and the interaction between financial constraints and productivity.

This indicates that productivity-enhanced capital reallocations are not realised in the informal manufacturing sector. Moreover, the results show an insignificant relationship between financial constraints and employment growth. Evidence also reveals that productivity- enhanced allocation of labour is not achieved in the informal sector. This implies that financial access constraints do not have a direct link to employment growth. Thus, the negative effects of financial access constraints on misallocation run through the investment channel and not squarely through employment growth.

Lastly, chapter 4 examines the efficiency of labour markets in allocating labour in the manufacturing sector. In particular, chapter 4 addresses the question of the extent and sources of labour market segmentation within and between the formal and informal sector employees.

Labour market segmentation theoretical models have traditionally been used to test the efficiency of labour markets by analysing the source of the wage differential between labour market sub-groups. The study is inspired by the rent-sharing models that theoretically explains the channels through which bargaining can lead to private but social sub-optimal allocations of labour. The empirical questions in this chapter are addressed using the matched employer- employee dataset that includes the panel dimension for employees and informal sector firms.

133

The chapter considers the labour market divide between the regulated formal and unregulated informal labour market segments, and the divide between permanent and contract (part-time) employees within the formal sector. We also consider the divide between the contract workers in the formal sector and workers in the informal sector. The comparison between formal and informal sector workers shows a significant raw wage gap of 52 percent and a conditional wage gap of 25 percent. These results indicate that the labour markets between the formal and informal sector firms are highly segmented. Likewise, a comparison between permanent and contract workers in the formal sector labour markets reveals a wage gap of between 16 and 28 percent, thus highlighting that within the formal sector, labour markets are also segmented. The RIF decomposition results show that the unexplained part accounts for relatively more of the wage, thus indicating evidence of labour market segmentation across the entire wage distribution. We also find evidence of segmentation between permanent and contract workers within firms in the formal sector (a raw wage gap of 28 percent and conditional wage gap of 15 percent). The RIF decomposition illustrates that the segmentation is higher at the top tail of the wage distribution.

Concerning rent-sharing as a source of wage differentials, the study finds a positive and significant association between firm profit-per-worker and employee wages in both the formal and informal sector. We also find evidence suggesting that more bargaining power is associated with higher wages. The results suggest evidence of rent-sharing as a source of segmentation within the formal sector, thus confirming that competitive labour models do not apply in the Zimbabwe labour markets. Thus, labour markets are inefficiently allocating labour resources.