Selected Issues of the Public Sector:
Special Interest Politics
Vincenzo Platino
Faculty of Economics and Administration
Masaryk University
—
Brno, April 11, 2016
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
What is a SIG?
In the idealized democracy, public policy is guided by the principle of “one
man, one vote”.
However, in all real polities, special interest groups participate actively in the
policy-making process.
What is a Special Interest Group (SIG)?
There is little consensus among social scientists, e.g.,
(i) Subset of voters with similar characteristics, beliefs, policy preferences.
(ii) Membership organizations that engage in politicy-related collective
activities.
We will restrict ourselves to broad definition:
Any organized group that can engage in a given policy-relevant activity.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Some researchers have estimate that in the 2000 there were circa 7000 Special
Interest Groups in the United States.
Most of the SIGs are organized around economic concerns, e.g.,
trade associations (e.g., the Semiconductor Industry Association,
American Iron and Steel Institute),
labor unions (e.g., the American Postal Workers Union, American
Federation of Teachers),
professional associations (e.g., American Medical Association, National
Bar Association).
There are SIGs who pursue noneconomic policy objectives, e.g.,
drug and alcohol policy (Mothers Against Drunk Driving, National
Organization for the Reform of Marijuana Laws),
capital punishment (Citizens for Law and Order, National Coalition to
Abolish the Death Penalty),
gun control (National Rifle Association, Handgun Controls, Inc.),
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Interest groups engage in a variety of activities and tactics to promote their
political objectives:
(i) SIGs as information providers, (i.e., lobbying activities)
Source of intelligence for lawmakers.
(ii) SIGs as education providers
Incentive to shape public opinion to their favour.
(iii) SIGs as resource providers
Campaign contributions.
Buying a chance to meet with a lawmaker.
Buying credibility.
Buying influence.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Questions addressed
One would like to provide answers to the following questions:
(i) What determines the extent to which SIGs are able to affect policy
outcomes?
(ii) Which channels (activities and tactics) do they exploit to bias policy
outcomes to their favour?
(iii) What happens when SIGs with different objectives compete for influence?
This lesson aims at providing a partial answer to the first point, considering
SIG as information provider.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Methodology and Assumptions
Which tools to analyze interactions policymakers and SIGs?
Actors are self-interested and behave rationally.
Game-theoretical approach: in search for an equilibrium.
- Actors operate in a strategic (intra-stage) environment.
- Forward-looking (inter-stage) behavior.
- Learning from observables in an imperfect information
world.
→ Coherent analytical framework(s) to study policy issue, both economic
and noneconomic.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
What’s lobbying?
Lobby: organized group serving SIG’s purposes.
Lobbying: (collection and) strategic dissemination of information.
More narrowly: lobbying as an attempt at persuading policymakers
about importance of lobby’s interests.
Need to take care of (i) informativeness of lobbying (ii) verifiability of
information.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Basic framework of analysis
.
(i) One lobby (L), one policymaker (P).
(iia) P’s welfare (objective) as a function of policy p (endogenous, controlled)
and state-of-the-world θ (exogenous, unknown).
(iib) Ranking of policy options depends on underlying state-of-the world. For
simplicity G(p; θ) = −(p − θ)2
.
Remarks:
1. policymaker’s preferences might represent for example, the welfare of the
members that are residents in the policymaker’s district, and they will
vote in the next election.
2. The pertinent facts are described by the variable θ. It might indicate the
tightness of the labor market, which would affect a policymaker’s
evaluation of alternative interest rates.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
(iiia) L’s welfare (objective) as a function of policy p (endogenous, not
controlled) and state-of-the-world θ (exogenous, known).
(iiib) Preferred policy outcomes depend on underlying state-of-the-world. For
simplicity U(p; θ) = −(p − θ − δ)2
, with δ > 0.
(iv) Asymmetric information: L knows θ, P has prior beliefs on its possible
values (i.e., probability distribution over r.v. θ)
(v) Preference structure (i.e., quadratic) and preference misalignment (i.e., δ)
are common knowledge.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Two states of the world
Suppose that there are only two states of the world: Θ = {θl, θh},
with θl, θh ∈ R and θl < θh.
θl θl + δ θh θh + δ
p
G
U
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
The policymaker initially regards the two values as equally likely, i.e.,
Prob(θ = θh) =
1
2
.
If L reveals the true state of the world, then a trusting policymaker solves
max
p∈R
−(p − θ)2
which implies p = θ.
If no information is disclosed, then the policymaker solves
max
p∈R
Prob(θ = θl)[−(p − θl)2
] + Prob(θ = θh)[−(p − θh)2
]
which implies p = Ep(θ) according to ex-ante beliefs, where
Eea(θ) = Prob(θ = θl)θl + Prob(θ = θh)θh.
Similarly, if P remains uncertain, then p = Eep(θ) according to ex-post
beliefs.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Two states of the world: Equilibria
Our task: we want to investigate the conditions under which lobbying can be
informative
Given θ ∈ {θl, θh}, L can report two possible messages m(θ) ∈ {θl, θh}.
We assume that the policymaker takes the lobbyist’s claims at face value.
• Case 1.: θh is the true state of the world.
(i) If L reports m(θh) = θh, then a trusting policymaker sets p = θh.
(ii) If L reports m(θh) = θl, then a trusting policymaker sets p = θl.
Remark: For simplicity of notation we will write p instead of p(m(θ)).
Notice that: U(θh; θh) ≥ U(θl; θh) if and only if θl ≤ θh. Prove it!
Thus, L has no incentive to misrepresent the fact when the state is θh.
• Case 2.: θl is the true state of the world.
(i) If L reports m(θl) = θh, then a trusting policymaker sets p = θh.
(ii) If L reports m(θl) = θl, then a trusting policymaker sets p = θl.
Notice that: U(θl; θl) ≥ U(θl; θl) iff δ ≤
1
2
(θh − θl). Prove it!
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Summarizing:
(i) If δ ≤
1
2
(θh − θl) there exists an equilibrium with informative lobbying.
The equilibrium is fully revealing.
(ii) If δ >
1
2
(θh − θl) the lobbyist’s report lacks of credibility:
The lobbyist has the incentive to announce the state θh no
matter what the true state is.
The policymaker do not care about the report of the lobbyist
and sets p = Eeaθ = 1
2 (θh + θl).
θl
θl + δ
θh p
U
Babbling Equilibra: If δ ≤
1
2
(θh − θl) there exist equilibria in which, the
policymaker do not care about the report of the lobbyist and sets p = Eeaθ.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Three states of the world
Assume that:
Three states of the world: Θ = {θl, θm, θh}, with θl < θm < θh.
The policymaker perceives the three states of the world as equally likely.
Given θ ∈ Θ, L can report three possible messages m(θ) ∈ {θl, θm, θh}.
When can the lobbyist advise the policymaker of the true state of the world?
• Case 1.: θl is the true state of the world.
(i) L prefers to report θl to θm if U(θl; θl) ≥ U(θm; θl). This leads to the
condition
δ ≤
1
2
(θm − θl) (1)
(ii) Notice that a false report of θh is unattractive to L anytime it is
unattractive to report state θm. Why?
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
• Case 2.: θm is the true state of the world.
(i) L has no incentive to report state θl because L prefers a policy larger that
θm, and a false report would induce the policymaker to choose a policy
that is smaller than p = θm.
(ii) L prefers to report θm to θh U(θm; θm) ≥ U(θh; θm). This leads to the
condition
δ ≤
1
2
(θh − θm) (2)
• Case 3.: θh is the true state of the world.
L has no incentive to report either the state θl or the state θm.
Each of these false reports would induce the policymaker to choose a
policy that is smaller than p = θh, and the SIG regards this policy to be
less than what it desires.
Therefore, conditions (1) and (2) are necessary and sufficient for an equilibrium
with truthful report.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Equilibrium with limited communication
What will be happen when one of the previous conditions is violated?
Suppose that δ ≤
1
2
(θm − θl) and δ >
1
2
(θh − θm).
In this situation, given θ ∈ Θ, L can report two possible messages
m(θ) ∈ {θl, ¬θl}, where ¬θl = (θm ∨ θh), i.e., θm or θh.
Then, a policymaker who takes the lobbyist’s claim at face value sets p = θl
when m(θ) = θl and p =
1
2
(θh + θm) when m(θ) = ¬θl. Why?
When does the lobbyist have incentive to report falsely in any of the three
states?
• Case 1.: θh is the true state of the world.
L has no incentive to report falsely. If L claims m(θh) = θl, he would
induce a policy of p = θl. This is outcome is less desiderable than
p =
1
2
(θh + θm) since
1
2
(θh + θm) > θl and the SIG regards the policy
p =
1
2
(θh + θm) to be less than what it desires.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
• Case 2.: θm is the true state of the world.
A truthful report report of m(θm) = ¬θl will result in a policy level of
p =
1
2
(θh + θm), while a false report of m(θm) = θl will result in a policy
level of p = θl.
Thus, L will not report falsely if and only if U(θh+θm
2
; θm) ≥ U(θl; θm),
which leads to the condition
δ ≥
θh − θm
4
−
θm − θl
2
(3)
Notice that this condition is satisfied when condition (2) is violated.
• Case 3.: θl is the true state of the world.
A truthful report report of m(θl) = θl will result in a policy level of
p = θl, while a false report of m(θl) = ¬θl will result in a policy level of
p =
1
2
(θh + θm)
Thus L will not report falsely if and only if U(θl; θl) ≥ U((θh+θm)
2
; θl),
which leads to the condition
δ ≤
θh − θm
4
+
θm − θl
2
(4)
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Therefore, when conditions (3) and (4) are satisfied, then there exists an
equilibrium with partial transmission of the SIG’s knowledge.
In the case in which conditions (1), (2), (3) and (4) are satisfied, we have three
different kind of equilibria.
1. an equilibrium with truthful report;
2. an equilibrium with partial transmission of the SIG’s knowledge;
3. Babbling equilibria.
Each of these reporting strategies may be justified by consistent beliefs of the
policymaker.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Costly Lobbying
So far, no costs for engaging in lobbying activities
Reasonable?
Hiring lawyers, policy experts, high-priced spokespersons.
Advertising, entertaining policymakers.
We study incentives for SIGs to invest in lobbying...
... and the effects of these costs on lobbying efficacy.
Three costs’ types:
(i) Exogenous: lobbying costs that are relatively fixed and outside the
control of the SIG. They do not vary with the content of the SIG’s
message;
(ii) Endogenous: SIG may choose to run a costly advertising campaign
rather than a more modest one. In this way, the SIG might be able
to indicate something about the nature of the policy environment;
(iii) Access: these costs are imposed by the policymaker, e.g., the
policymaker may insist that a SIG contribute to his campaign before
to schedule a meeting.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Exogenous lobbying cost: two states of the world
(i) One lobby (L), one policymaker (P).
(ii) P’s welfare (objective) as a function of policy p and state-of-the-world θ
G(p; θ) = −(p − θ)2
(iii) L’s welfare (objective) as a function of policy p, state-of-the-world θ and
fixed lobbying costs c
U(p, c; θ) = −(p − θ − δ)2
− c, δ > 0
where c = 0 if L chooses not to lobby, or c = cf > 0 otherwise.
(iv) Asymmetric information: L knows θ, P has prior beliefs on its possible
values (i.e., probability distribution over r.v. θ).
(v) Preference structure and preference misalignment (i.e., δ) are common
knowledge.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Two states of the world: Timing
Assume Θ = {θl, θh} and Prob(θ = θl) = 1/2. Game unfolds as follows:
L learns θ and chooses whether to bear the cost cf or not.
If so, P updates beliefs over θ conditional on L’s message m(θ).
If not, P may still update beliefs based on L’s choice not to lobby.
In any case, the game ends with P selecting p which maximizes his
expected welfare, and payoffs are realized.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Equilibrium
First, notice that incentives for truthful reporting are not altered by
cf > 0. Indeed,
(i) L will report truthfully when true state is θh (check it), yet might do so
even when true state is θl.
(ii) Potential gain from false reporting might not overcome the lobbying cost:
best not to lobby when true state is θl.
=⇒ SIG might do better when lobbying is costly!
P would learn from this that the true state is θl indeed
Information conveyed by L’s decision to show up, not by message’s
content.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Let’s construct an equilibrium according to which lobbying is observed iff θh is
the true state.
Failure to lobby makes P infer θl and viceversa.
Such beliefs system must be consistent in any PBE. Therefore,
(i) The SIG is willing to pay cf when the state is θh if and only if
U(θh, cf ; θh) ≥ U(θl, 0; θh), that is if and only if
cf ≤ (θh − θl)(2δ + θh − θl) =: k
(ii) The SIG prefers to refrain from lobbying when the state is θl if and
only if U(θl, 0; θl) ≥ U(θh, cf ; θl), that is if and only if
cf ≥ (θh − θl)(2δ − θh − θl) =: k
Remark: k > 0 if and only if δ > 1
2
(θh − θl). This makes sense, for it says
that a positive lobbying cost is required for credibility if and only if the
SIG has an incentive to report falsely when the state is θl.
Since k > k, there exists range of cf for which the previous inequalities
are both satisfied. =⇒ Such an equilibrium arises.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Question 1.
Consider Θ = {θl, θh}, and suppose that δ ≤ 1
2(θh − θl). If the
lobbyist sends the message m(θl) = θl to the policymaker, which
of the following beliefs are consistent with the fully revealing
equilibrium?
1. Prob(θl|m(θl) = θl) = Prob(θh|m(θl) = θl) = 1
2;
2. Prob(θl|m(θl) = θl) = 0 and Prob(θh|m(θl) = θl) = 1;
3. Prob(θl|m(θl) = θl) = 1 and Prob(θh|m(θl) = θl) = 0;
4. Prob(θl|m(θl) = θl) = 1
3 and Prob(θh|m(θl) = θl) = 2
3.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics
Question 2.
Which of the following costs are endogenous?
1. Salaries of the technical experts;
2. To run a costly advertising campaign rather than a modest
one;
3. A contribution to the policymaker’s campaign;
4. Salaries of the lawyers.
Vincenzo Platino Selected Issues of the Public Sector: Special Interest Politics