Rent-seeking by politicians and firms may distort the allocation of important public resources such as state-owned banks’ finance (Khwaja and Mian 2005), government bailouts (Faccio, McConnell and Masulis 2006) procurement contracts (Goldman, Rocholl and So 2007) and natural resources (Aaronson 2008). This is perceived to be a serious problem in many countries, especially during periods of economic crisis, as resources are scarcer and competition for them becomes stiffer (Johnson and Mitton 2003). However, evaluating the actual extent of the misallocation implied by rent-seeking activities is extremely difficult because such practices are often illegal and therefore remain hidden from official statistics.
At the same time, connections between private firms and the public administration are in many instances tolerated by the law. For example, in several countries (including the US) private companies contribute massively to electoral campaigns. Restricting the attention to those cases in which the relationship between the politician and the firm is known, we can identify rent-seeking and quantify the social cost it imposes on the rest of the economy.1
One such case occurs when politicians are directly involved in the economic activity of private firms. In recent work (Cingano and Pinotti, 2009), we merged administrative registries of Italian local politicians and national social security archives to identify this type of connections in a panel of about 1200 Italian manufacturing firms during the period 1985-1997.
The system of local governments in Italy comprises about 8000 municipalities, 100 provinces, and 20 regions. During our sample period, about 300,000 individuals were appointed in local governments, controlling directly about one-third of the total public budget (and retaining much discretionary power over the allocation of the remaining part). Unlike members of the national parliament, most local administrators maintain a stable occupation alongside their political career, so we could track 11,000 of them in our sample of firms (out of 1.5 million employees). In particular, about half of the firms experienced a change in connection status during our sample period, allowing us to exploit longitudinal variation to identify the impact on firm performance.
Figure 1 shows that profits experience a significant increase (about 5% on average) right after a firm becomes connected and an analogous decrease when the connection expires. This result is in line with the positive premium attached by financial markets to the stock price of connected firms (e.g. Faccio, 2006).
Figure 1. Firm profitability and political connection
Note: These figures plot the difference between the average profitability of connected and non-connected firms in the years around changes in connection status. Profitability is measured by the residuals of a regression of Earnings Before Taxes (at constant 1991 Italian Liras, converted into thousands of euros at official exchange rates) on firm, sector-year, and province-year fixed effects.
Our analysis suggests that the increase in profits is exclusively due to connections with politicians appointed with the majority coalition (connections with parties that lost the elections have no value). It reflects an analogous change in firm revenues, which in turn depends on favourable demand conditions (as opposed to productivity increases). Finally, such increases occur only in the domestic market, while exports are unaffected, and the gains are larger for firms operating in industrial sectors and geographic regions characterised by a higher incidence of public expenditure over total demand.2 Table 1 reports our estimates of the average effect of political connections on firm revenues, distinguishing between sectors and regions characterised by a different relative importance of public demand.
Table 1. Firm revenues and political connections
| |
|
SECTORS, by public expenditure |
| |
|
all |
high |
low |
| |
all |
5.2% |
6.2% |
0.8% |
| REGIONS,by public expenditure |
high |
16.4% |
24.3% |
2.8% |
| |
low |
3.0% |
4.3% |
-0.3% |
Notes: This table reports the estimated effect of political connections on firm revenues as an average in the whole sample and across firms operating in sectors and/or regions characterised by high vs. low public expenditure over total value added in manufacturing (above and below median values, respectively). All estimates control for firm, sector-year and province-year fixed effects. Bold characters denote estimates that are statistically significant at the 95% confidence level.
These findings confirm that the private returns to political connections stem from a distortion of public demand in favour of connected firms. The implied welfare loss, in terms of misallocation of public expenditure, varies (negatively) with the elasticity of substitution between the products of different firms; it ranges from 10% in markets characterised by low incidence of public demand to 20% in more heavily dependent markets.
References
Aaronson, Susan Ariel, “Oil and the public interest”, VoxEU.org, 12 July 2008.
Cingano, Federico and Paolo Pinotti, “Politicians at Work. The Private Returns and Social Costs of Political Connections”, Bank of Italy Working Papers 709 (2009).
Faccio, Mara, “Politically connected firms”, American Economic Review 96 (2006), 369-386.
Faccio, Mara, John J. McConnell and Ronald W. Masulis “Political connections and corporate bailouts”, Journal of Finance (2006), vol. 61(6), 2597-2635.
Goldman, Eitan, Jörg Rocholl and Jongil So, “Political Connections and the Allocation of Procurement Contracts”, EFA Ljubljana Meeting (2007).
Johnson, Simon and Todd Mitton, “Cronyism and capital controls: Evidence from Malaysia”, Journal of Financial Economics 67 (2003), 351-382.
Ölcer, Dilan and Helmut Reisen, “Extracting more from EITI”, VoxEU.org, 17 February 2009.
Ritva Reinikka and Jakob Svensson, “Publicity can be a powerful deterrent to corruption”, VoxEU.org, 2 July 2007.
1 Note that, in principle, the estimates obtained in this way are very conservative. Rent-seeking should in fact be more pervasive (and the associated welfare loss higher) when the connection is not manifest. On this issue, see the Vox column by Reinikka and Svensson (2007) on publicity as a deterrent to corruption in the allocation of public funds; Ölcer and Reisen (2009), on the other hand, find only mixed evidence in favour of the Extractive Industries Transparency Initiative, a process proposed by the British government in 2003 to enable more transparent and accountable governance in resource-rich countries.
2 At the sectoral level, the relative weight of public demand was computed according to the input-output matrices for the Italian economy. It is highest for pharmaceutical products (almost 40%), while almost negligible for electronic components. At the regional level, it is computed using national accounts and industry statistics, and it turns out to be higher in southern Italy.
