Lobbying and Corruption Dr James Tremewan (james.tremewan@univie.ac.at) Transparency and Lobbying
Introduction http://www.oecd.org/gov/ethics/Lobbying-Brochure.pdf 2/21
Introduction http://www.oecd.org/gov/ethics/Lobbying-Brochure.pdf 3/21
Introduction Transparency • It is a common view (even amongst politicians and lobbyists!) that increasing the degree of transparency of lobbying is desirable. • Necessary to identify undue influence of vested interests in policy decisions. • Reveals policymakers’ conflicts of interest. • Increases public confidence in democracy/policymaking. • Policymakers may want to know who is funding lobbyists, e.g. to identify whether pressure comes from domestic public opinion or foreign interests. • Yet only one third of OECD countries have government regulation of lobbying (as of 2013). Regulations vary widely across countries, e.g. mandatory (Canada and US) or voluntary (France). 4/21
Introduction Arguments Against Transparency • Compulsory disclosure of conflicts of interest may reduce pressure for prohibition of such conflicts: • US Senator Philip Hart claimed disclosure works by “revealing the possibility of . . . conflict, leaving it to the voter to decide whether the conflict has influenced the official acts of the congressman or senator.” • Voters may be unable to infer how lobbying has affected policymaker/regulator: • Disclosure of lobbying could increase (unwarranted) trust/mistrust. • Lobbyists/politicians who suspect that people who know of financial arrangements discount their information may be encouraged to exaggerate even more. • ”Moral licensing”: people feel worse about taking advantage of others who are unaware. Transparency reduces moral costs. • May discourage submission of useful but commercially sensitive information. 5/21
Introduction Examples • Example 1 (Interest Group, Politician, Voter): • Politician must decide whether to support or oppose gun control and knows which is best for society. • NRA will fund Politician’s electoral campaign if they oppose it. • Voter’s must decide whether to believe Politician’s statements about impact of gun control, and if to vote for him/her. • Voters may or may not know about donations from NRA, depending on campaign-financing laws. • Example 2 (Interest Group, Research Institute, Regulator): • In 2013, the US Occupational Safety and Health Administration began public consultation on setting limits on working with silica dust (a major hazard to construction workers). • For the first time they requested that those submitting evidence should declare their funding sources. • A group of senators protested, claiming revealing funding sources would bias the OSHA’s evaluations of submissions. 6/21
Introduction Model: Timeline • A lobbyist wants a particular policy implemented ( B ) and offers a politician a payment of 4 if they recommend this policy. • Lobbyist’s interests may diverge or coincide with voters’ (politician’s private information). • Timeline: • ”Nature” chooses � L , w.p. 2 / 5 (interests diverge) s = R , otherwise (interests coincide) • The politician observes the true state of the world ( L or R ) and chooses whether to accept the payment and recomend B ( full commitment assumed ), or refuse payment and recommend either A or B . • The voter observes only the action of the policymaker (possibly including whether or not they accept the payment) and chooses A or B . 7/21
Introduction Model: Payoffs • Politician receives: • 6 if the policy they recommend is chosen (gets elected). • 4 if they accept the payoff. • 1 if the policy they recommend matches the state of the world (e.g. intrinsic motivation, or good for reputation if true state becomes common knowledge with some probability). • Voter gets 10 if the policy that they choose matches the state of the world and 5 otherwise. • No Transparency (NT), voter observes only which policy is recommended. • Full Transparency (FT), voter observes which policy is recommended AND whether or not the payment is accepted. 8/21
Introduction Theoretical predictions • Similar definition of equilibria similar to previous set of slides. • No Transparency - only corrupt equilibrium (CE): • Politician always takes payment and recommends B . • Voter always follows recommendation. • EU V = 8; EU P = 10 . 6; correct choice 60% of the time. • Full Transparency - CE, and honest equilibrium (HE): • Politician never takes the bribe, and recommends A in L and B in R . • Voter always follow the advice and believes that if the payment is accepted that they are more likely to be in A . • EU V = 10; EU P = 7; correct choice all the time. • Why no equilibrium where politician recommends A in L , and takes the bribe and recommends B in R ? • Voter will always follow recommendation, so politician always has incentive to take payment in L as well. 9/21
Introduction Research questions • Is the best outcome ( A or B ) chosen more often with transparency? • Are voters more likely to mistrust information when the payment has been take - are they more likely to follow a recommendation of B when the bribe has not been taken? • Do politicians anticipate this, and decline the payment in R . 10/21
Introduction Experimental Design • 10 rounds of stranger-matching in same role. Repeat in other role (surprise): to ensure Receivers fully understood Senders incentives. • Two randomly chosen rounds paid from each role (2 pts = e 1). • Strategy method for Receiver: • Easier to see if equilibrium strategies are really being played. • Make sure we get enough observations at every information set. • Strategy method for Senders too complicated to explain. • Six sessions (142 subjects): • NT: Three sessions of 24 subjects. • FT: Two sessions of 24 subjects, one of 22. • Two matching-groups in each session: 12 independent obs. • All statistical tests are Mann-Whitney rank-sum tests based on matching-group averages. 11/21
Results Means: NT=0.77; FT=0.71 (difference not significant) 12/21
Results Means: NT=0.49; FT=0.41 (p-value=0.0547) 13/21
Results Means: NT=0.01; FT=0.03 (p-value=0.0505) 14/21
Results Receiver Strategies (NT) 15/21
Results Receiver Strategies (NT) 16/21
Results Receiver Strategies (FT) 17/21
Results Receiver Strategies (FT) 18/21
Results Means: NT=0.78; FT=0.64 (p-value=0.025) 19/21
Results Preliminary Results: Summary • Transparency does not improve information transmission: • In FT, correct choice made less than predicted. • In NT, correct choice made more than predicted. • A substantial number of Senders reject the payment to tell the truth in L in both treatments (declining over time). There is more truth-telling in the NT treatment, consistent with “moral licensing”. • Only a small number of Senders reject the payment and tell the truth in R . This number declines over time, but increases again when the roles are switched. More common in the FT treatment. • In FT, following all recommendations is modal strategy (CE). • Very few Receivers follow when the payment is not taken, and don’t follow when the payment is taken (HE). • The second most common strategy is to follow a recommendation of B when the payment is taken, but not when the payment is not taken! (honesty is suspicious?) 20/21
Results Preliminary Conclusions • Transparency fails to induce Senders to tell the truth (in fact they lie a little bit more). • Possible explanations: • Moral license (see earlier slide). • It seems that some Senders initially believe that they need to decline the payment to be believed in R , as hypothesised. However, as they encounter the (strange?) strategy of Receivers only believing them if they take the payment, they start accepting the payment more often. • Lying aversion means a large number of senders decline the bribe in L to tell the truth, but take the payment in R as they can still tell the truth. Receivers recognise this so are willing to follow recommendations of B even when the payment is taken (and some Senders take advantage of this and take the payment all the time).... honesty destroys the honest equilibrium! 21/21
Recommend
More recommend
