Parties or Portfolio? The Economic Consequences of Africa’s Big Cabinets

Abstract

Does cabinet size have an impact on economic policy in Africa? The average number of ministers has increased steadily for four decades, yet we know little about the economic effects of new portfolios, despite popular complaints about costly cabinets. Comparative studies generate conflicting expectations, either blaming coalition governments for patronage or crediting them with economic restraint. Using data on 45 Sub-Saharan African countries between 1971 and 2006, our empirical analysis links parties and portfolios to budgetary policy performance. We show that cabinets with more ministries are associated with budget surpluses, but they are also slightly more likely to engage in patronage spending. Next, we find that cabinets governing through multiparty coalitions have no consistent impact on budget surpluses. However, they are strongly associated with less extractive government and lower rates of patronage spending compared with single-party cabinets. These results hold after controlling for the type of colonial legacy, economic conditions, population size, constraints on executives, level of democracy, oil income, type of party system and ethnic and religious fractionalization. We conclude that parties and portfolios are both important but they have different effects: adding portfolios to the cabinet may improve economic outcomes by enhancing specialization, but governance through multiparty cabinets generates incentives to both limit extraction and restrain patronage spending.

Government Type and Public Spending in Africa

“…Indeed, analyses that include both variables come to varied conclusions about the relative importance of the number of ministers and measures of the number of parties for spending levels (Perotti and Kontopoulos 2002; Volkerink and de Haan 2001). The only existing work on government composition and its effects on government spending in Africa looks at both the government composition and the number of ministers (LeVan and Assenov 2015), but the theoretical mechanism focuses on patronage pressures, rather than spending, and includes both democracies and autocracies. To facilitate the comparison of democratic governments in Africa to existing analyses, I propose a second hypothesis, which examines support for the argument that the CPR problem exists at the ministerial, rather than party, level:…”

“…Although scholars have increasingly looked to other regions to test arguments about levels of government spending (Elgie and McMenamin 2008; Hallerberg and Marier 2004; LeVan and Assenov 2015; Wehner 2010; Woo 2003), they frequently include both democracies and autocracies. This is also true of existing work on African cabinets, which has tended to include both democracies and autocracies (Arriola 2009; Arriola and Johnson 2014; Francois, Rainer, and Trebbi 2015; LeVan and Assenov 2015; Yabré, Semedo, and Ouédraogo 2019), or exclusively autocracies (Kroeger, 2020). Because electoral accountability is an important part of the CPR logic that underpins my theoretical mechanism, I focus on democracies, where we might reasonably expect that electoral accountability is more applicable.…”

“…Third, this article presents an empirical analysis that can more easily be compared to existing scholarship from other regions. I use a new dataset designed to capture specific information about government composition in African democracies, allowing scholars to move beyond the use of less specific index measures of government composition and stability (LeVan and Assenov 2015; Yabré, Semedo, and Ouédraogo 2019). I also focus on the empirical strategies discussed at length by Bawn and Rosenbluth (2006), facilitating a discussion about how the effects of CPR spending arguments compare across more or less institutionalized contexts (Elgie and McMenamin 2008).…”

“…This encyclopedic reference allows me to count ministerial portfolios according to the titles officially on record. There is a general sense that the number of ministers in African cabinets is increasing over time (Arriola 2009; LeVan and Assenov 2015), which is reinforced by a contested narrative regarding the perception that Africa’s bureaucracies are overgrown, or “bloated” (Goldsmith 1999; Mkandawire 2015). However, focusing on those democracies for which spending data is available, there is relative stability in cabinet size, as illustrated by Figure 1.…”

“…Many scholars have argued that existing theories are relevant to a broader sample of countries (Elgie and McMenamin 2008; Hallerberg and Marier 2004; Wehner 2010; Woo 2003) but most do not include many African countries in their sample 1 . Others include both democracies and autocracies (LeVan and Assenov 2015; Yabré, Semedo, and Ouédraogo 2019), which means that they treat spending dynamics as comparable across these settings, rather than a product of different underlying institutional mechanisms. Still other existing work focuses on budget balances (Elgie and McMenamin 2008; Woo 2003) whereas the primary outcome of the common‐pool‐resource theory is on overall levels of spending (Bawn and Rosenbluth 2006; Martin and Vanberg 2013).…”