Lobbying and Corruption Dr James Tremewan - - PowerPoint PPT Presentation

Lobbying and Corruption

Dr James Tremewan (james.tremewan@univie.ac.at) Lobbying in Brussels / The Tullock Model

Lobbying in Brussels 2/27

Lobbying in Brussels

Lobbyists in Brussels: who are they?1

• Precise data is hard to come by, but estimated:
• 15,000-30,000 actively employed to influence EU institutions.
• Two-thirds work for business interests (the rest for civil society

and trade unions).

• Corporate lobbying well in excess of one billion euros.
• In-house lobbyists:
• 500 corporations have their own lobbying offices: BP, Rolls

Royce, E.ON...

• 1,500 industry lobby groups: International Association of

Amusement Parks and Attractions, European Chemical Industry Council (over 170 employees representing 29,000 companies)...

• NGOs: Amnesty International, The Smoke Free Partership,

Transgender Europe...

• For-hire firms: Think tanks; Public relations and law firms.

1http://corporateeurope.org/sites/default/files/publications/ceolobbylow.pdf

3/27

Lobbying in Brussels

Lobbyists in Brussels: what do they do?

• Work in advisory groups (Set up by EC in early stages of

legislation development and policy initiatives to provide technical advice).

• Draft amendments to legislation (sometimes used as is by

MEPs).

• Cultivate informal relations with MEPs: cocktail receptions,

expenses-paid trips, dinners.

4/27

Lobbying in Brussels

Lobbyists in Brussels: problems and safeguards

• Fine line between gifts and bribes, lobbying and corruption.
• Cash-for-influence scandal:
• Sting operation by the Sunday Times in March 2011.
• 60 MEPs invited to assist fake lobby firm for 100,000 per year.
• 14 agreed to meet, and three appeared to accept: Ernst

Strasser (Austria), Zoran Strasser (Slovenia) resigned; Adrian Severin (Romania) did not.

• Transparency of lobbyists:
• Register of lobbyists2 including who they work for and how

much they spend.

• Currently voluntary, but plans to make it compulsory by 2017.
• Transparency of officials:
• European Commissioners and MEPs must declare financial

information.

• But no evidence this is systematically verified3.

2ec.europa.eu/transparencyregister 3www.transparencyinternational.eu/focus areas/eu-integrity-study

5/27

Lobbying for rents

Lobbying: what is it all for?

• As we have seen, a huge quantity of resources is put into

lobbying EU institutions (the same goes for national institutions all around the world... Washington has an even higher density of lobbyists.)

• NGOs may highlight environmental problems, human rights

causes, etc.

• Much industrial lobbying is about obtaining or defending

economic rent:

• ”The big industrialists grumble that the politicians don’t give

them enough protective tariffs, and the politicians grumble about industry for not coming up with enough money for their election campaigns.” Count Leinsdorf in The Man Without Qualities by Robert Musil.

6/27

Lobbying for rents

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Price Quantity A B C MC MC+tariff QMC Qtariff

• International firms and domestic monopoly produce at cost MC.
• With no tariff: production at QMC, total surplus = A + B + C.
• With tariff: production at Qtariff , total surplus = A + B.
• B is just transfer from consumers to domestic monopoly, so

welfare loss of tariff to country = C (?)

7/27

Lobbying for rents

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Price Quantity A B C MC MC+tariff QMC Qtariff

• How much would monopoly spend on lobbying for tariff?
• If lobbying activities are assumed wasteful, welfare loss to

country could be up to B + C.

• The Tullock model considers firms/groups competing for

economic rent and asks what level of resources can we expect to be wasted on lobbying.

8/27

Lobbying for rents

The Tullock model

• Two firms are lobbying to receive a political prize worth Q (e.g. .
• Each firm chooses simultaneously how much to spend on

lobbying (x1, x2).

• The probability with which firm i wins the prize is

πi(xi, xj) =

• 1/2 if xi = xj = 0

xR

i

xR

i +xR j otherwise,

where R > 0 determines the expected returns from lobbying:

• R = 0 ⇒ lobbying pointless.
• R = ∞ ⇒ biggest spender wins for sure.

9/27

Lobbying for rents

The Tullock model

• For 0 < R ≤ 2, there is a unique symmetric pure strategy Nash

equilibrium.

• Each firm maximises

xR

i

xR

i + xR j

Q − xi

• First order conditions:

QRxR−1

i

xR

j

(xR

i + xR j )2 − 1 = 0

• Solve for xi, xj:

xi = xj = QR 4

10/27

Lobbying for rents

The Tullock model

• For 0 < R ≤ 2, the bigger R, the more resources are wasted in

lobbying:

• R = 1 ⇒ half the prize is wasted.
• R = 2 ⇒ the whole prize is wasted!
• For R > 2 there are only mixed-strategy equilibria.
• For R = ∞:
• Unique symmetric NE where both firms uniformly randomize
• ver [0, Q].
• Sometimes total spending is more than Q, sometimes less.
• Total level of spending is exactly Q on average.

11/27

Lobbying for rents

How do buyers respond?

• Firms lobby for a monopoly position. If successful:
• Price increases from competitive level (PC) to monopoly (PM).
• Quantity decreases from QC to QM.
• Winning firm gains B, consumers lose B+C.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Price Quantity A B C PC PM QC QM

• So what do consumers do?

12/27

Lobbying for rents

”Rent-defending Buyers”

• Not just firms but also consumer groups lobby for favourable

legislation.

• Examples:
• Consumers International

(http://www.consumersinternational.org/)

• Free Software Foundation (http://www.fsf.org/)
• European Beer Consumers Union (http://ebcu.org/)
• How to model this?
• Value of ”prize” is greater for consumers (B+C) than firm (B).
• Timing: Consumers may lobby
• at the same time as firms to prevent monopolization, or
• after monopolization to implement regulation.
• Key questions:
• Does the presence of rent-defending buyers increase or decrease

wasteful lobbying?

• Does the timing of consumer lobbying affect social efficiency?

13/27

Lobbying for rents

Davis and Reilly (1998)

• ”Do too many cooks always spoil the stew? An experimental

analysis of rent-seeking and the role of a strategic buyer”, Davis and Reilly, Public Choice, 1998.

• Environment with four sellers (value of prize=200) and one

buyer (value of prize=400).

• Theoretical predictions derived for six different ”treatments.”4
• Lottery (Tullock model with R = 1) or ”perfectly discriminating

auction” (R = ∞)

• Either no ”buyer” (Baseline), buyer lobbies with firms

(1-stage-with-buyer), or buyer lobbies after monopolist is determined (2-stage).

• Theoretical predictions tested in laboratory experiment.

4For more general results see Strategic Buyers and the Social Cost of

Monopoly, Tore Ellingsen, American Economic Review (1991).

14/27

Lobbying for rents

Baseline lottery

• Seller’s objective: πi = bi

s1200 − bi, where s1 = b1 + b2 + b3 + b4.

• FOC implies: bi = 200s1−s2

1

200

, i ∈ {1, 2, 3, 4}.

• Solution symmetric, so s1 = 4bi, and bi = 37.5.
• Total lobbying expediture = 150, DWL = 200, so total social

cost = 350.

15/27

Lobbying for rents

1-Stage-with-buyer lottery

• Seller’s objective: πi = bi

s2200 − bi, where

s2 = b1 + b2 + b3 + b4 + bB (bB is buyer’s bid).

• Buyer’s objective: πB = bB

s2 400 − bB.

• Optimal bids:
• bi = 200s1−s2

1

200

= 19.75, i ∈ {1, 2, 3, 4}.

• bB = 400s1−s2

1

400

= 98.77.

• Total lobbying expediture = 177.77, expected DWL = 88.89

(200 x probability a seller wins), so expected social cost = 266.67.

16/27

Lobbying for rents

2-Stage lottery

• Backward Induction: Suppose firm i becomes the monopolist.
• Objectives in second stage:
• πi =

bi bi+bB 200 − bi.

• πB =

bB bi+bB 400 − bB.

• Optimal bids and expected profits in second stage:
• bi = 44.44; πi = 22.22.
• bB = 88.89; πB = 177.78.
• Objectives for firms in first stage:

πj = bj

s122.22 − bj, j ∈ {1, 2, 3, 4}.

• Optimal bids in first stage: bj = 4.17, j ∈ {1, 2, 3, 4}.
• Total lobbying expediture = 150, expected DWL = 66.67, so

expected social cost = 216.67.

17/27

Lobbying for rents

Summary

Lotteries Baseline 1 Stage 2 Stage Seller bid 37.5 19.75 4.17/44.44 Buyer bid

• 98.77

88.89 Total lobbying expenditure 150 177.77 150 Expected DWL 200 88.89 66.67 Expected social cost 350 266.67 216.67 All Nash equilibria for perfectly discriminating auctions involve mixed

• strategies. Complicated, so forget about proofs in this course.

PDAs Baseline 1 Stage 2 Stage Average seller bid 50 14.5 0/50 Average buyer bid

• 83.6

100 Total lobbying expenditure 200 141.6 150 Expected DWL 200 58.4 50 Expected social cost 400 200 200

18/27

Lobbying for rents

Experiments: Design

• In each of 24 sessions, five participants play one auction 15

times, then another 15 times.

• Designed such that each auction is sometimes first, sometimes

second (to control for ”order effects”).

• In half the sessions the participants are ”experienced”, i.e. they

have already participated in a previous session.

19/27

Lobbying for rents

Experiments: Design

• Participants given $6 at beginning of each sequence of games

(plus $6 showup fee).

• At beginning of each auction, ”buyer” determined randomly.
• Participants visually isolated and bids submitted anonymously

(in baseline auctions the buyer does nothing).

• Bids collected by monitor and publically anounced.
• For lottery auctions, probabilities calculated, winning range from

0-99 announced for each bidder, and number drawn randomly to determine winner of prize ($2 for sellers or $4 for buyer).

20/27

Lobbying for rents

Experiments: Results (Social Cost)

21/27

Lobbying for rents

Experiments: Results (Social Cost)

• Across the six auction types, Nash equilibrium predictions

correctly rank-order mean observed social costs.

• Social costs are lower in the baseline lottery than in the baseline

perfectly discriminating auction.

• The addition of a strategic buyer reduces social costs.
• However, in all auction types, aggregate social costs

substantially exceed Nash equilibrium levels.

• The social cost savings generated by introducing a buyer tend

to be smaller than predicted.

• But experience brings things closer to predictions.

22/27

Lobbying for rents

Experiments: Results (Bids)

23/27

Lobbying for rents 24/27

Lobbying for rents 25/27

Lobbying for rents

Experiments: Results (Bids)

• Overbidding is pervasive in all auction types. In contrast to

social cost predictions, Nash predictions fail to rank-order bidding outcomes.

• Experience diminishes overbidding, and with experience, Nash

predictions for sellers perform well.

• Buyers account for much of the total aggregate overbidding

(probably because participants on average play only one out of five auctionss as buyer, so have less experience in that role).

26/27

Lobbying for rents

Conclusion

• In line with the theory, the experimental evidence finds that:
• The possibility of competing for rent is socially costly.
• A ”rent-defending” buyer reduces social costs.
• Social costs are higher than predicted because of overbidding,

but the discrepancy decreases with experience.

• Does it matter that the precise mechanism predicted by theory

isn’t accurate if the overall results are in line with the predicted effects?

• Consumer lobbies increase social welfare and benifit consumers.

Why are they so few compared to corporate lobbyists?

• A firm faces a large gain for itself from lobbying.
• Although consumers as a whole have more to lose, for each

individual this will be a small amount, and lobbying by one consumer has positive externalities for others: possibility of free riding.

27/27