Doing Business while Holding Public Office: Evidence from Mozambique

• In Event: Bureaucracies

Abstract

We link the universe of owners of businesses formally registered in Mozambique since Independence to a new database of politically exposed persons (PEPs). This sheds light on channels through which serving politicians can reap private benefits in an institutional setting where legislation and capacity to regulate conflicts of interest are limited.

Our focus on the extent to which politicians’ personal business interests expand on account of holding office represents a particular form of cronyism, whereby active politicians straddle the public-private sector divide. This is conceptually distinct from crude forms of neo-patrimonialism or state capture, which respectively tend to caricature either politicians or existing private firms as the prime agents of rent-seeking behavior. In contrast, we hypothesize that the expansion of politicians’ own business network is one feasible outcome of a time-consistent and (semi-)legal bargain, whereby entrepreneurs agree to share the benefits (profits) of their business activities with firms owned by politicians in exchange for ongoing protection and/or other favors (e.g., permits, licenses etc.).

To shed light on how politicians’ personal business networks evolve we use public registry data from Mozambique to construct the complete network of companies established since 1975, including their registered owners. This registry information allows us to construct individual-level time-series metrics of business performance, including the number of companies owned, as well as indicators of their structural power within the network of firm-owners, such as indicated by their ‘godfather centrality’ (as per Jackson, 2019).

Linking this to a new hand-collected database of PEPs, we use both individual fixed-effects and lagged outcome models to bracket estimates of the contribution of holding political office to own business outcomes over the period 1985-2019. In comparison to individuals without so-called political exposure, we find that the average PEP achieved significantly faster growth in their personal business network along multiple dimensions. According to our core results, obtained from a sub-sample of current PEPs matched to other PEPs who only held office later in the observation period, becoming a PEP is associated with a: 7 percentage point (pp) faster growth (over five years) in the number of companies in which they have an ownership interest; 10 pp greater increase in their godfather centrality position; and 20 pp greater increase in both their degree and decay centrality positions. In other words, even when we compare current PEPs against future PEPs, current PEPs become owners in more companies and accumulate a stronger structural position in the business network (e.g., as a broker between other businesses). Our basic results hold across a range of alternative models used to address 2 potential confounding from unobserved selection into holding political office, including group-wise trends and controls for prior business performance, and are also confirmed from event study estimates.

Digging deeper, we find that while both party political and executive offices are associated with broadly similar outcomes over the period(s) individuals hold office, these benefits only persist for holders of executive positions once they leave office. Importantly, we also show that PEPs systematically tend to accumulate ownership in joint-stock companies, as opposed to sole-owned companies, and they predominantly accumulate interests in companies with activities in commerce (trade) and investment/finance. These findings suggest that PEPs act as gatekeepers to state contracts and genuine (foreign direct) investment.

In relation to previous studies our contribution is threefold. First, we provide one of the first quantitative (microeconomic) inquiries into the dynamics of politicians’ personal business networks, especially in a very low-income country where the (small) formal private sector is known to be heavily influenced by political elites. Second, we consider a distinctive and novel set of outcomes over an extended period of time, namely the size and importance of the private business networks associated with individual politicians, including measures of network centrality that capture their influence within the network of business-owners. In so doing, we provide support to existing work, largely of a qualitative nature in Mozambique and beyond, which contends politicians frequently act as unproductive rentier-brokers, captured by the metaphor of the ‘big man’ straddling private and public interests (e.g., Szeftel, 2000). Third, we demonstrate how publicly-available information from (digital) business registries, which have the advantage of near-universal coverage of the formal sector, can be deployed to investigate substantive political economy questions.

Authors

• Felix Sören Schilling, Copenhagen Business School

• Sam Jones, University of Copenhagen

• Finn Tarp, University of Copenhagen