Public Choice In a Representative Democracy

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

The findings of Figure 4.4 (upper part) very much are similar to the outcome under direct democracy (median).

However, the assumptions are unrealistic:

• one issue dimension (left or right),

• single peaked preferences,

• symmetric and unimodal preference distribution,

• all individuals vote,

• two candidate parties (the terms candidate and party are used synonymously).

Two reasons may be responsible for a relaxation of the assumption that all voters do vote:

• indifference and

• alienation.

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

© Freytag 2013

Public Choice

Then, still the median voter outcome holds, as long as the preference distribution is unimodal and symmetric. The median voter determines the outcome.

This may change if either preferences are asymmetrically distributed (mid part of Figure 4.4) or if the distribution is multimodal (lower pat of Figure 4.4).

In both cases, the degree of alienation is decisive; and in both parts of the figure, it is too small to change the median voter outcome.

Like in direct democracy, a problem of instability may arise if the world is multidimensional, i.e. if a program comprises of many issues.

Combined with multimodal preference distribution and alienation, multidimensionality might cause that extreme

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

Again, log-rolling may arise, causing cycles.

A candidate preferring all alternatives, maximises social welfare.

However, a candidate preferring issues 1 and 2 and preferring not issue 3 may win over the first one, as voters A and C can gain at the expense of voter B.

Cycling, as every platform can be defeated.

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

In a representative democracy, cycles should not be observed in single elections, as individual candidates cannot rotate within one campaign.

However, it should be observed in a permanent change of government (revolving doors).

Alternative hypotheses:

• random hypotheses and

• conspiracy hypotheses.

In the US (the appropriate example for two-party competition), the incumbent party left government only in a fourth of the cases.

Why so much stability?

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

b) Two-party competition in a constrained policy space

Candidates may restrict their platforms only to a subset of the policy space.

Imagine two parties and three voters (Figure 4.5) with equally long contract curves. No point in the (x1,x2)- diagram can defeat all others (see Figure 3.10).

It seems rational for candidates to restrict their platforms to points close to the voters’ ideal points.

The restricted set of platforms is called the uncovered set. These still can be defeated, but only in a cycle.

___

In figure 4.5, the uncovered set is the Pareto set ABC.

1.Introduction

Democracy

Figure 4.5: Cycling outcomes for a three-voter electorate

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

A

C

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

Relaxing further assumptions

1) If candidates have personal preferences,

• they may only reluctantly remove policy positions,

• the number of dimension may be reduced,

• ideology may become the single dimension.

2) If candidates are not exogenously given, but can enter and leave the contest, the result is similar. Under the median voter outcome, it does not make sense to enter, if one candidate has already declared her platform.

On the other hand, this relaxation shows that under twoparty elections both candidates will not have identical platforms.

Empirical tests don’t support to the median voter outcome.

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

c) Two-party competition (probabilistic voting)

Figure 4.6: Cycling possibilities

y

C UC

M

A B

UA UB

© Freytag 2013

x

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

So far, we have assumed that the voters are easily caught by and immediately react to the candidates’ moves (see Figure 4.6). A slight change to the left wins certain votes and loses others.

This “jerky” response by the voters is implausible:

• the voters are unlikely to be fully informed aboutt the candidates’ positions;

• the candidates may not know where the voters’ optimal points are;

• there may be random events having influence on the voters’ decisions.

If we assume probabilistic voting instead of deterministic, i.e. if voters behave continuously instead of discontinuously, the median voter outcome is confirmed.

Andreas Freytag

1.Introduction

2.Origins of the State

3.Public Choice in a Direct

Democracy

4.Public Choice in a Representative Democracy

5.Application of Political

Economy Models

6.Normative Public

Choice

Public Choice

The equilibrium is within the Pareto set, i.e. within the triangle ABC. Therefore, the normative properties of the equilibrium are desirable.

What happens if interest groups are introduced into the analysis?

• The existence of interest groups implies a bias inn the probability of a candidate of being voted for (see Figure 4.7);

• the candidates know the distribution of the bias term, but not the individual bias;

• still, the Pareto set can be reached;

• however, different interest groups receive different weights (welfare);

• “one man, one vote” does not hold any longer.