Topic 6: Politics of Public Goods Provision

Public choice theory

Government may not act to maximise citizens' well-being, because politicians and bureaucrats follow their own interests.

Government failure

The inability or unwillingness of government to act mainly in citizens' interest.

Bureaucracies (public choice view)

Civil service organisations that provide government services but follow their own self-interest.

Leviathan theory

Voters cannot trust government to spend efficiently and must design rules (tax and spending limits, electoral competition) to restrain government greed.

Public vs private provision - evidence

Private provision is usually more efficient for most goods (steel, telecoms, banking). For goods people understand poorly (pensions, health insurance, education), private competition can mean wasteful advertising and higher costs.

Natural monopoly

Uniformly falling marginal (and average) cost means one firm can supply the whole market most cheaply (economies of scale).

Private natural monopoly - problem

Maximises profit by producing too little at too high a price. Government running or regulating it can restore the efficient quantity.

Contracting out

Government keeps responsibility for a service but hires private firms to provide it. Raises the risk of corruption.

Corruption

Abuse of power by officials to increase their own or their associates' wealth.

Corruption - what makes it worse

Weak electoral accountability, more red tape, and underpaid officials. Bribes act as informal taxes to access services.

Acemoglu and Robinson (2012) - Why Nations Fail

Inclusive institutions (broad political and property rights) grow faster than extractive ones (power held by a small elite). North vs South Korea: similar in 1948, now about 10 to 1 in income per head.

Mauro (1995)

Countries with lower-quality government (red tape, corruption) grow more slowly. Problem: these countries differ in other ways, so the estimate may be biased.

Buchanan vs Key

Buchanan: improve politics by reforming the rules, not by hoping for better people. Key: government ultimately depends on who we elect.

Clientelism - definition

A political system where private goods and services are exchanged for political support such as votes.

Clientelism - five elements

1) Dyadic relationships (patron and client, possibly via brokers). 2) Contingency (benefits only for supporters). 3) Hierarchy. 4) Iteration (repeated interaction). 5) Volition (in principle free to exit).

Clientelism - hierarchy

Turns accountability upside down: politicians hold voters accountable for their votes, instead of voters holding politicians accountable. The patron has higher status than the client.

Clientelism - iteration

Repeated exchanges build reciprocity norms and let each side punish the other for defecting. Unlike bribery, which can be one-off.

Clientelism - limits to volition

Threats of violence, no real chance to exit or voice, social ostracism, withheld benefits, hard collective action, and few credible challengers.

Patronage vs clientelism

Patronage uses public office resources (e.g. public jobs). Clientelism can use private or party resources too.

Vote buying vs clientelism

One-off vote buying with no obligation is not clientelism. It counts only if it has contingency and iteration.

Clientelism - consequences

Weaker state capacity, resources allocated by loyalty not need or merit, accountability to clients not the public, and more corruption.

Intergovernmental grants - normative role

Fund services with spillovers, reduce reliance on inefficient local taxes, and guarantee similar access to essential services nationwide.

Swing voter vs core supporter targeting

Swing voter view: target grants at swing areas to win votes (Lindbeck and Weibull; Dixit and Londregan). Core supporter view: reward areas that strongly backed the incumbent (Cox and McCubbins).

Why favour aligned governments?

Grants to opposition-run local governments help the opposition claim credit, so grantors get more electoral return from giving to aligned (same-party) governments.

Sole-Olle and Sorribas-Navarro (2008) - design

900 Spanish municipalities, 1993-2003, grants from central, regional and upper-local governments. Changes in alignment over 3 elections allow difference-in-differences, and grants from aligned vs unaligned grantors allow triple differences.

Sole-Olle and Sorribas-Navarro (2008) - findings

Partisan alignment substantially raises grants received, most of all when both layers are single-party governments. Coalition leaders also gain. Mere municipal coalition partners do not.

Banerjee and Somanathan (2007)

Rural India, about 500 constituencies, 1971-1991. Social group identity strongly predicts access to public goods: Brahman areas did better on some goods, Scheduled Caste and Tribe areas worse on almost all. Larger villages better served.

Asher and Novosad (2017)

Close-election regression discontinuity in India: constituencies represented by a ruling-party politician get more private employment, higher firm share prices and more output (night lights). Political favouritism.