Web C Political Economyhomepage.ntu.edu.tw/~luohm/econ2015f/chapter_w03_text.pdf · Mobutu. The country, now called the Democratic Republic of the Congo, is the third largest country

W3-2

Political Economy

3Do governments and politicians follow their citizens’ and constituencies’ wishes?

CHAPTER OUTLINE

Conflict of Interest and Political Economy

W3.1

Does Democracy Work?

W3.2

Controlling Politicians

W3.3

Market Failure Versus Government Failure from a Political Economy Perspective

W3.4

Do governments and politicians follow their citizens’ and constituencies’ wishes?

EBE

Governments play a central role in modern economies. In the United States, for example, spending by the federal, state, and local governments makes up about 35 percent of GDP.1

But can we trust the government to enact policies that reflect the interests and the will of the people? Can we trust that the officials we elect to speak for us will actually do so? Should these officials consider their own points of view? What about their party’s view? Do they look after their own interests rather than those of the communities they are meant to serve? It is questions like these that form the basis for this chapter.

Web Chapter

© IT

AR-

TASS

Pho

to A

gen

cy/A

lam

y

KEY IDEAS

Because the economic preferences and views of citizens are often opposed, conflicts of interest over government policies are inevitable.

The Median Voter Theorem stresses the moderating influence of the median voter, who is the voter with the political bliss point (most preferred policy) in the middle of the distribution of political bliss points among voters.

In the presence of perfect political competition, which allows political candidates representing different options to freely enter and compete in elections, there will be a tendency for parties to converge toward the median voter’s political bliss point.

In practice, politicians often do not follow the wishes of the median voter or the electorate.

Some elected officials are politically corrupt in that they use the advantages of their office to engage in illegal and unethical actions that benefit or enrich themselves.

What determines whether, in practice, governments work for the public good or toward different objectives?

Stepping back a bit, let us ask the basic question, “What is the ‘public good’?” Imagine that the government intervenes in order to ensure a more equitable distribution of resources in society. The government can do so by taxing high incomes and redistributing to others in the society including, but not limited to, those who would otherwise be in poverty. But those whose income falls due to taxation are bound to see the intervention in a different light. Is such redistribution in the interest of “the public”? It is difficult to answer this ques-tion when some people are in favor of a policy and some are against it. In such situations, we say that there is a conflict of interest within the population. Some individuals and groups will benefit and favor one policy, while others will see their welfare decline as a result of this policy and prefer and support an alternative.

Although some policies create improvements whereby all parties benefit, in many situations such universal improvements (or what economists refer to as Pareto improvements) are not possible. Instead, there are winners and losers from each feasible policy, creating conflict of interest over policy choices. In practice, a major cause of conflict of interest is that the economic interests of dif-ferent individuals or groups are opposed. But different

economic interests are not the only source of conflict of interest in society. It might also arise because people have different views and oppose or favor different policy proposals for this reason.

Political economy is the study of the conflict of interest between individuals and groups, how this conflict is resolved, and how citizens influence government policy and control

Conflict of Interest and Political Economy

W3.1

Section W3.1 | Conflict of Interest and Political Economy W3-3

Conflict of interest refers to opposing preferences concerning policies between individuals or groups.

Political economy is the study of the conflict of interest between individuals and groups, how this conflict is resolved, and how citizens influence government policy and control politicians.

There are winners and losers from each feasible policy, creating conflict of interest over policy choices.

W3.2

W3.3

W3.4

W3-4 Web Chapter 3 | Political Economy

W3.1

politicians. Political economy clarifies how to think about the preferences of a group of in-dividuals or of the entire society. It also studies to what extent governments, politicians, and bureaucrats respond to these preferences and how they can be made more responsive.

To study political economy, we must first understand how society makes decisions or delegates decisions to politicians—processes that are intimately related to the social arrangements and political regimes under which the society functions. We also need to examine how the conflict of interest among individuals and groups exhibits itself in collec-tive choices under different social arrangements.

Monarchy, Dictatorship, and DemocracyMost of us have experienced life only under a single political regime, which may be called representative democracy. In representative democracy, citizens vote for different candi-dates who, if elected, hold public office for a period of time and then make policy choices. At the end of their term, or sometimes even before, they can be voted out of office—for ex-ample, if the citizens are not satisfied with their performance or a better candidate emerges.

The term democracy has its roots in ancient Greece, in particular, in the city state of Athens. Democracy, as the Athenians envisaged and practiced it, was very different from the representative democracy we know today. First, it was not representative. The (male) citi-zens of Athens would gather and vote over policies. The qualifier “representative” is key in understanding that our democracy is not direct like that of the Athenians. It involves voting only for who will represent us, with the major policy decisions delegated to these representa-tives until they are voted out of office. Because today almost all democracies are representa-tive, we will simply refer to them as “democracy” rather than “representative democracy.”

Democracy, though it seems like the only political game in town to many of us, has been rare throughout history. Democracy in Athens came to an end in the third century BC. A similar, though more limited, form of democracy in ancient Rome also came to an end when the relatively short-lived republic was replaced by the Roman Empire.

In fact, most societies throughout history have been ruled by various forms of mon-archies. In a monarchy, the monarch—a king, queen, emperor, or empress—has the ultimate ruling power. He or she chooses policies and decides whether to go to war and how much tax to impose on subjects, though in practice many decisions are delegated to administrators and relatives. Although the power of monarchs is not unlimited, as they can be deposed (for example, because of uprisings, revolutions, or coups), we normally think of their power as nearly absolute. Monarchies have been gradually replaced by more representative regimes throughout the nineteenth and twentieth cen-turies, though there are still many countries that have powerful monarchies, such as Saudi Arabia, Nepal, and Thailand, or largely symbolic monarchies, such as the United Kingdom, Sweden, Denmark, Norway, the Netherlands, and Spain.

Many countries are also ruled by dictators. Dictatorship is the general name given to non-democratic regimes ruled by a single person (a dictator) or a junta consisting of a small group of soldiers or rulers. Crucially, dictatorships do not hold elections to elect the key executives, and those in power are not replaced through elections. Policies in dictatorships are not responsive to the views or wishes of the citizens. Dictators often pursue policies that are in their own interests or sometimes those that they deem to be in the interest of the country as a whole, even though these conflict with the wishes of the citizens.

We should therefore not expect dictatorships to pursue policies in line with the interests and wishes of their citizens. The situation, at least in theory, is different in democracies. Because citizens have the right to vote politicians out of office, we might expect the politi-cal process to deliver policies that citizens, or at least a large fraction of them, would agree with. But to investigate whether this is the case or not, we first need to discuss how to think about the preferences of citizens when there is a potential conflict of interest among them.

How Well Does Voting Work? The Condorcet ParadoxWhen we think of an individual’s preferences, we implicitly assume that they satisfy some minimal amount of consistency. If Crystal prefers a Kindle Fire to a Samsung Galaxy and would rather have the new iPad than a Kindle Fire, then we would know that she surely prefers the new iPad to a Samsung Galaxy also. This is the situation depicted in panel

Representative democracy is the political system in which public officials are democratically elected and empowered to act on behalf of the citizens.

A dictatorship is a regime ruled by a single dictator or a junta.

© R

ob B

yron

/Shu

tter

stoc

k

W3.2

W3.3

W3.4


W3.1

(a) of Exhibit W3.1. In such situations, we would say that our preferences are transitive. Most economists take transitivity to be a minimal requirement for rational decision making, because without it an individual could end up in an infinite and frustrating cycle.

To see this, consider Melanie. Suppose that Melanie’s preferences are such that she would first choose to exchange her Samsung Galaxy for a Kindle Fire, which she prefers, then swap her new Kindle Fire for an iPad if given the chance because she prefers the iPad to the Kindle Fire. But if she were given an opportunity, she would also give up the iPad for the Samsung Galaxy. But she had happily given up the Samsung Galaxy for the Kindle Fire in the first instance! This is all because Melanie’s preferences are not transitive, as shown in panel (b) of Exhibit W3.1. The cycling arrows in this panel indicate that Melanie will keep on swapping her current choice for something else, never settling into an item that she prefers to the alter-natives. Basically, preferences are transitive if they can be put in a sensible order from best to worst, as we have done. All of the individual preferences we have seen so far in this book have been transitive. In fact, this is a sufficiently natural property that we do not dwell on it much.

Unfortunately, when a society makes decisions by voting, such transitivity is no longer guaranteed. This was first noted by the French mathematician and philosopher Marquis de Condorcet. For this reason, the failure of transitivity in collective preferences implied by voting behavior is referred to as the Condorcet Paradox. Consider a society consisting of three individuals, Adam, Bob, and Charlie, who seek to reach a decision by expressing their preferences through voting. Suppose that the item on the agenda is the tax rate and there are three possibilities: a low tax rate, an intermediate tax rate, and a high tax rate. The three

Preferences are transitive if the following is true: if an individual prefers choice (a) to choice (b) and choice (b) to choice (c), then she prefers choice (a) to choice (c).

The Condorcet Paradox refers to the failure of transitivity in collective preferences implied by voting behavior.

Few other cases illustrate how dictatorial regimes look af-ter the interests of the dictator and his cronies at the ex-pense of society better than the case of the Congo under Mobutu. The country, now called the Democratic Republic of the Congo, is the third largest country by area and one of the most natural resource-rich countries in Africa. Its people have suffered for hundreds of years, however. From the 15th century to the late 19th century, the area was governed by the absolutist and repressive Kingdom of Kongo. It was replaced by the African Congo Free State in 1877, which despite its name, was just the opposite of free— essentially a state that was the private property of King Leopold II of Belgium, who acquired the area for his personal enrich-ment and as a source of natural resources. The Belgian rule came to an end in 1960 but by 1965, the Congo was back under authoritarian rule, following a coup by Joseph-Desire Mobutu, who quickly set up a highly authoritarian and corrupt regime. Mobutu ruled the Congo (which he named Zaire, a name immediately dropped after he was deposed in 1997) as his personal fiefdom. He was continu-ously trying to enrich himself even as Zaire’s population was becoming poorer and poorer. We do not know exactly how poor the Congo became because under Mobutu, Zaire did not collect national statistics. The World Bank es-timates that between 1965 and 1997, income per capita (per year) in the Congo/Zaire fell from 320 PPP- adjusted dollars to 104 dollars (or by a stunning 67.5 percent). But Mobutu himself was likely one of the richest men in the world. Some reports put his wealth at $5 billion. His cronies, who the locals referred to as the “great vegetables,” also likely gathered vast fortunes from corruption and thievery.

We will never know, since all of Mobutu’s and much of his cronies’ wealth was never declared officially and was often stored in suitcases and secret Swiss bank accounts. What we do know is that Mobutu felt it was a bargain to buy a 16th-century castle in Spain!

CHOICE & CONSEQUENCE

The Congo under Mobutu2

© L

aufe

r/Fo

tolia

W3.2

W3.3

W3.4


individuals will vote for any two of the three options presented to them, and whichever of these two options receives the majority of the votes (two against one or three against zero) will be implemented. In some sense, you can think of this as a very simple form of direct democracy (though nothing here depends on this feature and exactly the same paradox would emerge if we considered votes over three different candidates who would get elected to public office).

Suppose that the preferences of Adam, Bob, and Charlie over the three options are as shown in Exhibit W3.2 (the exhibit presupposes that the preferences of each of the indi-viduals are transitive).

Let’s break down the preferences by examining them in batches of two. Suppose that Adam, Bob, and Charlie are choosing between low and intermediate taxes. Imagine that each of them votes in line with their preferences, as given in Exhibit W3.2. So Adam prefers intermediate taxes to low taxes and will thus cast his vote in favor of the intermedi-ate tax rate. Charlie will do the same, whereas Bob will cast his vote for low taxes. The tally is thus two votes in favor of intermediate taxes and one in favor of low taxes, so the collective choice of Adam, Bob, and Charlie between low and intermediate taxes would be intermediate taxes.

Next consider a vote between intermediate and high taxes. According to the preferences in Exhibit W3.2, both Adam and Bob prefer high taxes to intermediate taxes, so they will cast their votes in favor of high taxes, while Charlie will vote in favor of intermediate taxes. So in this case, high taxes will be selected over intermediate taxes, with two votes to one.

But next comes the strange part. Consider a vote between high and low taxes. Look-ing again at the preferences in Exhibit W3.2, we see that Adam prefers high taxes to low, but both Bob and Charlie would vote in favor of low taxes. This means two votes for low taxes and one for high, so low taxes will receive the majority of the votes. But this

iPad is preferred tothe Kindle Fire and

the Samsung Galaxy

Transitive Preferences

(a)

Kindle Fire ispreferred to the

Samsung Galaxy

iPad is preferred tothe Kindle Fire

Non-Transitive Preferences

(b)

Samsung Galaxy ispreferred to the iPad

Kindle Fire is preferredto the Samsung Galaxy

Exhibit W3.1 Transitive and Non-Transitive Preferences

In panel (a), Crystal prefers an iPad to both a Kindle Fire and a Samsung Galaxy and also prefers a Kindle Fire to a Samsung Galaxy. This implies that Crystal’s preferences are transitive. In panel (b), Melanie prefers a Kindle Fire to a Samsung Galaxy, and an iPad to a Kindle Fire, but then prefers to give up the iPad for a Samsung Galaxy, imply-ing that her preferences are not transitive.

W3.1

© A

gen

cja

Foto

gra

ficzn

a C

aro/

Ala

my;

© M

artin

Shi

eld

s/A

lam

y; ©

Od

neok

o/Sh

utte

rsto

ck

W3.2

W3.3

W3.4


amounts to a lack of transitivity in the collective preferences implied by voting: the society expressed, through its voting behavior, first a preference for intermediate taxes over low taxes, and then for high taxes over intermediate taxes. And yet, whereas transitivity would have implied preference for high taxes over low taxes, it also expressed a preference for low taxes over high taxes.

How Well Does Voting Work? The Arrow Impossibility TheoremHow typical is the situation we just went through? The answer depends on whether you look at the glass as half empty or half full. On the half-empty side, a celebrated result in political economy from Kenneth Arrow, sometimes referred to as the Arrow Impossibility Theorem, shows that if we do not put some more structure on individual preferences, the Condorcet Paradox is far from unusual.3

Here is a quick outline of the Arrow Impossibility Theorem. Arrow wanted to investi-gate not only the properties of simple elections, where the policy that obtains the majority of the votes is implemented, but of any possible voting scheme that would “aggregate” the preferences of all citizens. For example, a voting scheme might give greater weight to some citizens than others or might use different procedures for choosing between different outcomes. Arrow argued that any reasonable voting scheme should satisfy the following three conditions:

1. Unanimity: if there is a vote between two options, 1 and 2, and all voters prefer 1 to 2, then 1 should be selected.

2. Transitivity: which, as we just explained, requires some degree of consistency in choices (and we saw in the example of the Condorcet Paradox that this is not neces-sarily trivial to satisfy).

3. Independence of irrelevant alternatives: if there is a vote between 1 and 2, then the preferences of voters concerning the comparison of one of these against some third option, say option 3, should have no bearing on the outcome of the vote.

Arrow also noted that another desirable property is that a voting scheme should not be dictatorial, meaning that there should not exist a single individual whose preferences always prevail regardless of the preferences of everybody else.

The remarkable, and somewhat depressing, result of Arrow’s Impossibility Theorem is that the only way of designing a voting scheme that always satisfies unanimity, transitiv-

ity, and independence of irrelevant alternatives is to accept the existence of a dictator—an individual who “calls the shots.” Or, put differently, Arrow’s result says that we can-not find a voting scheme that satisfies unanimity, transi-tivity, and independence of irrelevant alternatives and the additional requirement that there should be no dictators. This is disconcerting in part because democracy, based on voting as a way of aggregating preferences and making social choices, is the most obvious and attractive alterna-tive to dictatorship. But according to Arrow’s Impossibility Theorem, it will not succeed in making consistent choices.

Exhibit W3.2 Policy Preferences of Adam, Bob, and Charlie

This exhibit shows the preferences of three indi-viduals, Adam, Bob and Charlie over three policy options, low taxes, inter-mediate taxes, and high taxes.

Most preferred In between Least preferredAdam High taxes Intermediate taxes Low taxesBob Low taxes High taxes Intermediate taxesCharlie Intermediate taxes Low taxes High taxes

Arrow’s Impossibility Theorem is not meant as a critique of democracy . . . it highlights the tensions of aggregating preferences in the presence of a conflict of interest among the citizens in a society.

W3.1

W3.2

W3.3

W3.4


But this pessimistic interpretation is incomplete. As we will see, even with its many flaws, democracy is a good model of government and Arrow’s Impossibility Theorem is not meant as a critique of democracy. Instead, it highlights the tensions of aggregating preferences in the presence of a conflict of interest among the citizens in a society. Conflict of interest is key to this result: if there had been no disagreement among the citizens, the issue of aggregating preferences would not have arisen in the first place. The “glass is half full” interpretation, which we turn to next, goes further and shows that in many plausible situations the Condorcet Paradox will not arise and Arrow’s Impossibility Theorem doesn’t apply.

How Well Does Voting Work? Single-Peaked PreferencesThere is something a little unusual in Bob’s preferences shown in Exhibit W3.2. At some level, Bob looks like a “small government conservative” voter, who prefers low taxes to high taxes: from Exhibit W3.2 we see that Bob’s political bliss point is low taxes. The political bliss point of an individual refers to his or her most preferred policy option among all available policies. Most people with a political bliss point of low taxes would also put high taxes at the end of their ranking. But Bob is not a typical small government conservative. He will not vote for a smaller government no matter what: in fact, he prefers high taxes to intermediate taxes. This might be because, he might reason, if the government is going to impose intermediate taxes, which will reduce the disposable incomes of citizens but may not be enough to provide high-quality public services, that’s perhaps worse than having higher taxes and decent public services.

Leaving the issue of whether we find Bob’s preferences reasonable, let us first un-derstand his preferences a little better by plotting them together with those of Adam and Charlie in Exhibit W3.3.

The leftmost part of Exhibit W3.3 shows Adam’s preferences. A higher bar means a more preferred policy. Adam is a good contrast to Bob, because his political bliss point is high taxes, so we can think of Adam as a typical “big government progressive” voter who prefers high taxes (because he supports high social spending, for example). Adam has the highest bar at the high tax rate, then the next highest at the intermediate rate, and the lowest at the low tax rate. Adam thus has the typical preferences of somebody who prefers high taxes and redistribution. We say that Adam’s preferences are single-peaked, meaning that there is a single peak in the patterns shown by the bars. Single-peaked preferences, in particular, imply that an individual prefers policies closer to his or her political bliss point to those that are farther away. Put dif-ferently, if we look at two policy choices both to the left or both to the right of an individual’s bliss point, then the individual always prefers the one closer to his or her bliss point. Clearly, Adam’s preferences satisfy this property. His political bliss point is high taxes. Intermediate taxes are closer to high taxes than low taxes are, and Adam indeed prefers intermediate taxes to low taxes. Similarly, Charlie’s preferences have a single peak at intermediate taxes.

The middle part of the exhibit shows Bob’s preferences. Bob has the highest bar at low taxes, but then the next highest is at high taxes and the lowest bar is at intermediate taxes. We can see from the exhibit that his preferences don’t look single-peaked. Rather, it looks

The political bliss point of an individual is his or her most preferred policy option.

Preferences

200 2,000 20,000Adam Bob Charlie

Low taxes

Intermediate taxes

High taxes

Exhibit W3.3 Policy Preferences of Adam, Bob, and Charlie

This exhibit shows diagram-matically the preferences of Adam, Bob, and Charlie from Exhibit W3.2. A higher bar means a more preferred policy. So for Adam, the highest bar is for high taxes, the next highest bar is for intermediate taxes, and the lowest bar is for low taxes, reflecting his preferences.

An individual’s policy preferences are single-peaked if he or she prefers policies closer to his or her political bliss point to those that are farther away.

W3.1

W3.2

W3.3

W3.4


like there are two peaks separated by a valley (contrast this to the preferences of Adam, shown on the left, and of Charlie, shown on the right, which both have a single peak). In fact, some might view Bob’s preferences as somewhat unusual: if he prefers low taxes to high taxes, shouldn’t he then prefer low taxes to intermediate taxes also?

It seems fairly natural that preferences should be single-peaked in this example. This is not always the case, because many political choices are multifaceted. Nevertheless, single-peaked preferences are a basic benchmark. And in fact, the Condorcet Paradox is inti-mately linked to the absence of single-peaked preferences, as we will see next. You might conclude that if we are willing to assume single-peaked preferences, we may be able to get out of the conundrums implied by the Condorcet Paradox and Arrow’s Impossibility Theo-rem. If you thought so, you would be right, and this is the essence of an important result in political economy named the Median Voter Theorem.

How Well Does Voting Work? The Median Voter TheoremThe Median Voter Theorem goes as follows. As a first step, take all of the voters in society and plot their political bliss points. We do this in Exhibit W3.4 for a hypothetical group of voters, and to make things more concrete, we have recorded their most preferred tax rate. The exhibit shows that there are a few people with extreme political bliss points (0 percent and 99 percent tax rates, for example). But most voters have more moderate preferences—their political bliss points are not as extreme and are concentrated between 20 percent and 50 percent.

We then look at which voter is in the middle of this distribution, meaning that he or she has equal numbers of voters to his left and right (to simplify, we have taken the number of voters to be odd). This person is called the median voter. In the exhibit, for example, the individual with a political bliss point of 32 percent in the middle of

the distribution is the median voter and is shown in green. There are 10 voters with lower political bliss points (0 percent, 10 percent, 12 percent, 13 percent, 20 percent, 25 percent, 26 percent, 27 percent, 29 percent, and 31 percent), and there are 10 voters with greater political bliss points (33 percent, 34 percent, 38 percent, 39 percent, 42 percent, 43 percent, 50 percent, 52 percent, 60 percent, and 99 percent). The Median Voter Theorem then says that, provided the preferences are single-peaked, in an election between the political bliss point of the median voter and any other policy option, the political bliss point of the median voter will win a majority. This finding, of course, implies that there cannot be a Condorcet Paradox (since the political bliss point of the median voter beats any other alternative).

With the preferences shown in Exhibit W3.2, the Median Voter Theorem did not apply, because the preferences of all voters were not single-peaked. In particular, Bob, as we just

200 2,000 20,000

100%

90

Median voter

80

70

60

50

40

30

20

10

0

Mostpreferred

tax rate

Individuals ranked bytheir political bliss points

Exhibit W3.4 Political Bliss Points of Voters and the Median Voter

The median voter is an indi-vidual whose political bliss point is in the middle of the distribution. In this example, the median voter is the indi-vidual, shown in green, with the political bliss point of a tax rate of 32 percent who has 10 voters to his left and 10 voters to his right (that is, 10 voters with higher bliss points and 10 voters with lower bliss points).

The median voter is an individual whose political bliss point (most preferred policy) is in the middle of the distribution of political bliss points among all voters.

The Median Voter Theorem states that, when preferences are single-peaked, the political bliss point of the median voter will receive a majority of votes against any other alternative policy.

W3.1

© M

aste

rzp

hoto

fo/F

otol

ia

W3.2

W3.3

W3.4


saw, did not have single-peaked preferences. To show that this absence of single-peaked preferences is responsible for the Condorcet Paradox, let us consider exactly the same preferences as in Exhibit W3.2 for Adam and Charlie, but let us suppose that Barbara has replaced Bob and she has single-peaked preferences (in particular, her preferences are the same as Bob’s except that the rankings of intermediate and high taxes are swapped). Exhibit W3.5 shows these preferences, and Exhibit W3.6 plots the political bliss points of these three voters.

From Exhibit W3.5, you can verify that all three voters have single-peaked preferences. Thus according to the Median Voter Theorem, the political bliss point of the median voter—the voter in the middle of the distribution—should prevail in an election. Exhibit W3.6 plots the preferences of Adam, Barbara, and Charlie, again with the highest bar corresponding to the most preferred option. We have them in the order of Barbara, Charlie, and Adam since Barbara’s political bliss point is low taxes, Charlie’s political bliss point is intermediate taxes, and Adam’s political bliss point is high taxes.

Exhibit W3.6 shows that the median voter is Charlie, whose political bliss point is the intermediate tax rate. Now let us compare the votes between intermediate taxes and the other two alternatives. If there is a vote between intermediate and low taxes, Charlie will vote in favor of intermediate taxes (that is, after all, his political bliss point) and Adam will as well, so intermediate taxes obtain a majority. If there is a vote between intermediate and high taxes, then Charlie and Barbara will vote in favor of intermediate taxes against high taxes, and intermediate taxes again obtain a majority. Therefore, consistent with the Median Voter Theorem, an intermediate tax rate beats any alternative, and there is no Condorcet Paradox.

Exhibit W3.5 Policy Preferences of Adam, Barbara, and Charlie

This exhibit shows the preferences of Adam, Barbara, and Charlie over high, intermediate, and low taxes. In contrast to Exhibit W3.2, all three vot-ers now have single-peaked preferences (in particular, Barbara, who has single-peaked preferences, replaces Bob from Exhibit W3.2, who did not have single-peaked preferences).

Most preferred In between Least preferredAdam High taxes Intermediate taxes Low taxesBarbara Low taxes Intermediate taxes High taxesCharlie Intermediate taxes Low taxes High taxes

W3.1

Preferences

200 2,000 20,000Barbara Charlie Adam

Low taxes

Intermediate taxes

High taxes

Exhibit W3.6 Political Bliss Points of Adam, Barbara, and Charlie

This exhibit shows diagram-matically the preferences of Adam, Barbara, and Charlie from Exhibit W3.5.

W3.2

W3.3

W3.4


When it applies, the Median Voter Theorem makes elections a much more predictable and orderly process than the Condorcet Paradox and Arrow’s Impossibility Theorem imply. We will next see that it also creates a powerful moderating influence on the platforms of parties and on politics in general.

Political Competition and Convergence to the MedianThe Median Voter Theorem doesn’t, by itself, tell us what policies will be chosen in a so-ciety. Given the theorem, we might expect that the political bliss point of the median voter should be implemented. But to reach this conclusion, we still need to look in a little more detail into the political institutions of the society. First of all, in the United States we live in a representative democracy, so we do not directly vote over policies such as the income tax rate that will apply to high earners. Instead, we simply elect politicians and they legislate laws and policies on our behalf. So we first need to translate the insights of the Median Voter Theorem to a representative democracy. Second, we also need to ensure that there is perfect political competition, meaning that different options or candidates representing different options can freely compete in elections.

To understand the importance of political competition, assume that society consists of just Adam, Barbara, and Charlie. Further assume that Barbara is the dictator, meaning that she can choose whatever policy pleases her. In this case, she will simply choose her political bliss point—low taxes. Of course, there is no reason to expect the Median Voter Theorem to say anything about what a dictatorship will do, because it is a result about policy preferences in democracies, where the wishes of all those who vote count. But the key thing to notice about the situation in which Barbara makes the choices without consult-ing others is the lack of political competition. Political competition, like competition in the marketplace, is supposed to ensure a level playing field.

One of the major results in political economy is that when the Median Voter Theorem applies and there is perfect political competition, then the political bliss point of the median voter will be the policy implemented. Moreover, in this case there will be policy convergence among parties (or political candidates) competing for office: the par-ties will all run for office on a platform offering to implement this policy. This policy convergence result was first noted by Anthony Downs and is sometimes referred to as Downsian policy convergence.4

To understand this result, imagine the following institutional structure. There are two par-ties, both wishing to come to power. Each party can choose any one of the feasible policies as its policy platform, which means that if it receives the majority of the votes and gets elected, it is committed to implementing this policy. Suppose, in addition, that the Median Voter Theo-rem holds. For example, for concreteness, you might want to envisage a society consisting of 21 voters with political bliss points given as in Exhibit W3.4. In this society, two parties, which we will call Red and Blue, will announce their policy platforms and then the voters will choose which of the two parties to vote for. The party that obtains 11 or more votes (a major-ity) will come to power and implement the policy that it promised in its platform.

Fortunately, in this case, an equilibrium is straightforward to characterize. Recall from Exhibit W3.4 that the political bliss point of the median voter is 32 percent. Imagine that the Red party runs with a platform of a 25 percent tax rate. What should the Blue party do? Exhibit W3.7 shows the political bliss points of the 21 voters from Exhibit W3.4, together with the policy platform of the Red party. It also shows how votes will be distributed if the Blue party chooses a policy platform of a 32 percent tax rate. In particular, the exhibit indi-cates that the voters shown in red will vote for the Red party in this case (they prefer lower taxes than the Red party offers, but much lower taxes than the Blue party offers, so they are closer to the Red party). The voters shown in blue will vote for the Blue party (they prefer higher taxes than the Blue party offers but they are closer to the Blue party than to the Red party, which is offering an even lower tax rate). We do not know which of the voters in-between, shown in gray, will vote for the Blue or the Red party. But this is immaterial, because there are already 11 blue votes, so the Blue party will win regardless of how these voters vote. Therefore, if the Red party offers 25 percent as its platform, the Blue party can always choose a platform that will win the election.

You can now go through the motions and repeat exactly the same reasoning for the case in which the Red party offers some tax rate other than 32 percent. Exhibit W3.8, for

Perfect political competition refers to competition in the political sphere and in particular to political institutions that allow different parties to enter elections and offer any of the feasible options or policy choices.

Policy convergence refers to a situation in which two parties competing for office both offer the same policy platform.

W3.1

W3.2

W3.3

W3.4


example, shows the situation in which the Red party offers a tax rate of 34 percent. Once again, the Blue party gets at least 11 votes.

So what can the Red party do? Clearly, offering anything other than 32 percent—the po-litical bliss point of the median voter—will lead to defeat at the polls. What if it also offered 32 percent? If it did so, both parties would have exactly the same platform and all voters would be indifferent between the two parties. Therefore, we would expect either party to come to power with a probability of 50 percent. Since coming to power 50 percent of the time is better than being defeated for sure, it is a best response for the Red party to also offer 32 percent. By the same reasoning, it is a best response for the Blue party to offer 32 percent rather than some other policy platform. Thus there exists a unique equilibrium, in which both parties choose as their platform the political bliss point of the median voter. This, as you might have guessed by now, is the Downsian policy convergence result. It implies that when there is perfect political competition and the Median Voter Theorem applies, the two parties will have strong incentives to converge to a moderate position, corresponding to the preferences of the median voter, and the outcome will necessarily be the political bliss point of the median voter.

Does this result make any empirical sense? In the United States, for example, Demo-crats and Republicans bicker endlessly and have very different platforms about such things as abortion and taxes. Most recently, the Tea Party wing of the Republican party has pushed for policies sharply different from what Democrats and many Republicans used to agree on and has refused to compromise with the president, even when this threatened a government shutdown. So does this mean that there is no policy convergence in the United States? The recent events notwithstanding, there have been many periods in which the two parties have run on similar platforms, and the landmark policies implemented over the last several de-cades carry the imprint of both parties. For example, Medicare started under a Democratic

Exhibit W3.7 The Blue Party Wins a Majority

The Red party is offering a policy platform that promises a tax rate of 25 percent, while the Blue party is offering a policy platform of a tax rate of 32 percent (which is the political bliss point of the median voter). The voters to the left of the Red party, who prefer very low taxes, are closer to the Red party than to the Blue party, and will vote for the Red party (and are thus shown in light red). The voters to the right of the Blue party have political bliss points closer to the Blue party and will vote for the Blue party rather than the Red party (and are thus shown in light blue). The vot-ers in-between may prefer either party depending on their exact preferences (and are thus shown in gray). But regardless of how the voters in between vote, the Blue party, which will get at least 11 votes, will win the election.

200 2,000 20,000

100%

90

80

70

60

50

40

30

20

10

0

Mostpreferred

tax rate

Red party’splatform

Vote forRed party

Vote forBlue party

Blue party’splatform

Individuals ranked bytheir political bliss point

W3.1

W3.2

W3.3

W3.4


President Lyndon Johnson but was then expanded by Republican Richard Nixon and then again by Republican George W. Bush. Deregulation of various industries, including fi-nance, and strong anti-inflation policies started with Paul Volcker as the Chairman of the Federal Reserve under Republican Ronald Reagan, but continued under Democratic and Republican presidents that followed him. The aggressive fiscal stimulus plans against the recession of 2007–2009 started under Republican George W. Bush and continued largely unchanged under Democratic President Barack Obama. But we will also see below that the U.S. political scene is not one of full policy convergence, especially lately.

Self-Interested and Ideological Politicians: Implications of Limited CompetitionPerfect political competition is a useful benchmark but does not describe many real-world situations. In most democracies, political competition takes an imperfect form for at least three reasons:

1. Perfect competition in politics needs free entry—that is, candidates must be able to en-ter the political arena without any special legal or technical barriers keeping them from doing so. But free entry is difficult to achieve in practice because only well-organized and established parties can compete in elections, and this fact of political life has only become greater as the importance of money in politics has increased. For this reason, the United States has been a two-party system for over a century now, and third-party candidates have had very little success in national or even local elections.

2. Politicians and parties have ideological biases. So, for example, a party with a left-wing bias would not campaign with a promise of low taxes even if low taxes may be what the median voter wants.

3. As we will see in greater detail later, politicians are often highly self-interested and sometimes even untrustworthy. Such traits carry several implications. To start with, their promises might not be very reliable. George H. W. Bush campaigned in the 1988 presidential elections on a platform of no tax increases, famously saying: “Read my lips!”

200 2,000 20,000

100%

90

80

70

60

50

40

30

20

10

0

Mostpreferred

tax rate

Red party’splatform

Blue party’splatform

Vote forBlue party

Vote forRed party

Individuals ranked bytheir political bliss point

In most democracies, political competition takes an imperfect form.

Exhibit W3.8 The Blue Party Again Wins a Majority

The Red party is now offering a policy platform that prom-ises a tax rate of 34 percent while the Blue party is still offering a policy platform of a tax rate of 32 percent (the political bliss point of the median voter). This time, the voters to the right of the Red party, who prefer higher taxes, are closer to the Red party than to the Blue party and will vote for the Red party (and are thus shown in light red). The voters to the left of the Blue party have political bliss points closer to the Blue party and will vote for the Blue party rather than the Red party (and are thus shown in light blue). The Blue party again receives at least 11 votes and wins the election.

W3.1

W3.2

W3.3

W3.4


Yet two years after coming to power, he agreed to a budget that increased gasoline taxes; imposed new excise taxes on alcohol, tobacco, and other commodities; and raised the top marginal tax rate. Aside from failed campaign promises, politicians will often look after their own interests at the expense of the public, meaning that they will try to further their own careers, choose policies and public programs in line with their own personal and ideological biases, and at worst, (as we will see next), engage in corrupt and illegal activities to build their own personal wealth or to further their own interests.

These real-world situations restrict the applicability of the Median Voter Theorem and Downsian policy convergence. Suppose, for example, that preferences over taxes are given as in Exhibit W3.4 but there are only two parties competing. One of the parties favors an extremist left-wing position and would like to impose a tax rate of 70 percent on income, while the other one represents an extremist right-wing position and would like to abolish the income tax altogether. Moreover, suppose that, just as with the case of George H. W. Bush, the promises that politicians will make during the campaign don’t have any binding effect on the policies that the politicians will actually implement once in power. In this situation, the Downsian policy convergence result does not apply. Voters will have to elect one of the two parties and will have to put up either with a very high or a very low tax rate. Of course, if we were in a situation of perfect political competition, a third party would spring up and might offer what the median voter wanted. But in many situations, as in the United States and the United Kingdom, it is difficult for third parties to emerge and to successfully contest elections. Thus, in many relevant situations we might not get policies in line with the wishes of the median voter—even if the policy outcomes might not be as extreme as in the example we have just given. Whether politicians pursue policies in line with the interests of the public is a largely empirical question, which we turn to next.

Evidence-Based Economics

Recall that according to Downs’s model of political convergence, we would expect politicians from both parties to adopt policies in line with the preferences of the median voter in their constituencies. When this is the case, we would think that poli-

ticians are indeed responding to the wishes of the voters. In contrast, we have seen how ideologically motivated politicians might choose policies different from what the median voter or the majority of their constituencies prefer.

One way of investigating whether politicians choose the policies that their voters prefer is to focus on elections for the U.S. House of Representatives that either Republican or Democratic candidates won narrowly and then compare the actual policies they supported following the election. The idea can be understood by considering a specific example. Ignoring independent candidates, who are typically not important in federal elections in the United States, let’s consider two elections for congressman in the House of Representatives in two different districts. Suppose that in a district in New Jersey, the Republican candidate wins 50.1 percent of valid ballots and is elected, while in Pennsylvania the Democratic can-didate wins 50.1 percent and is also elected. In both cases, these are very narrow victories. The 0.2 percent difference between the New Jersey and Pennsylvania races could easily be due to small variations in exactly who turned out to vote, and one would not have great confidence that the same candidate would win if the vote were repeated.

On the basis of their voting patterns, we would therefore conclude that the political pref-erences of the citizens of New Jersey and Pennsylvania, at least insofar as their preferences between Republican and Democratic candidates are concerned, are very similar. However, almost by chance, New Jersey is sending a Republican to the House, while Pennsylvania

Q: Do governments and politicians follow their citizens’ and constituencies’ wishes?

W3.1

© ITAR-TASS Photo Agency/Alamy

W3.2

W3.3

W3.4


is sending a Democrat. If elected representatives do indeed represent the wishes of their constituencies, we would expect the congressmen from the two states to have a very similar voting record. For example, they should both oppose extreme right-wing or left-wing mea-sures brought for a vote because they are representing states that appear to have median vot-ers who are very much at the center. If, on the other hand, we find that the voting records in Congress of narrowly elected Democrat and Republican politicians are very different, then we could conclude that elected representatives are voting on the basis of their ideological biases, or party affiliations, rather than the wishes of their constituencies.

Economists David Lee, Enrico Moretti, and Matthew Butler investigated this implication by using data from the elections for the U.S. House of Representatives between 1946 and 1995.5 They focused on narrow elections and looked at information on how left-leaning or right-leaning a politician was and his or her voting record in the House (for example, whether they voted the same way as the Democratic or Republican leadership in the House). If these narrowly elected representatives from the two different parties simply follow the wishes of their constituencies, they should vote similarly on bills, since their constituencies have essen-tially the same composition (about 50 percent support for Democrats and 50 percent support for Republicans). But the researchers found a large difference between the voting records of narrowly elected Republicans and Democrats. In fact, their results suggested that a narrowly elected Republican votes in exactly the same way as an average Republican congressman (with a right bias), and the same applies to a narrowly elected Democrat (who votes with a left bias). These results therefore provide evidence against the view that elected representatives simply respond to the wishes and policy preferences of their constituencies. Instead, it appears as if different candidates have their own ideological or other policy agendas or follow those of their party, and if elected, they work toward implementing those agendas rather than pursuing policies that their constituencies or the median voter who elected them would have wished.

One shortcoming of the study by Lee, Moretti, and Butler is that we only observe the votes that the politicians cast in the House. We might worry that for many votes, the politi-cians we are focusing on are not pivotal, and thus their votes do not really matter for the overall policy decisions that will be implemented, and thus the fact that they are following the party line might not be so surprising when their votes do not matter.

A related study by economist Per Pettersson-Lidbom overcame this difficulty by applying a similar methodology to elections for Swedish municipalities.6 In such municipalities, politi-cians elected to local office decide local tax rates, influence unemployment rates, and have a say about the size of local government employment. Pettersson-Lidbom similarly found that the policy choices of narrowly elected right-leaning and left-leaning politicians were very dif-ferent. For example, narrowly elected politicians from the left chose on average 2–3 percent higher tax rates and employed 4 percent more workers in local governments than narrowly elected politicians from the right. Pettersson-Lidbom’s study provides even stronger evidence against the view that politicians simply implement the wishes of their constituencies.

Question Answer Data Caveat

No. The evidence suggests that politicians of different parties representing more

or less the same preferences support and choose very

different policies.

Vote shares in House of Repre-sentatives elections in the United

States and votes of elected representatives in the House.

Also vote shares in Swedish local elections and policies chosen

by local governments.

The extent to which politi-cians from different parties vote and choose different polices varies across coun-

tries, issues, and time as our next Letting the Data Speak

box illustrates.

Do governments and politi-cians follow their citizens’ and

constituencies’ wishes?

W3.1

W3.1

W3.3

W3.4


W3.2

LETTING THE DATA SPEAK

Political Polarization and Policy Convergence

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

1879 1889 1899 1909 1919 1929 1939 1949 1959 1969 1979 1989 1999 2009Year

House

Senate

Distancebetween

the parties

Exhibit W3.9 Polarization in the House and Senate, 1879–2009

Source: Updated data from Nolan McCarthy, Keith T. Poole, and Howard Rosenthal, Polarized America: The Dance of Ideology and Unequal Riches, Volume 5 (MIT Press, 2006); available at http://voteview.com/ polarized_america.htm.

Under perfect political competition, theory suggests that we should expect policy convergence. But as we have seen, the reality is quite different. Does the extent of divergence (or “polarization”) between politicians from different parties vary over time? To answer this question for the United States Congress, political scientists Nolan McCarthy, Keith Poole, and Howard Rosenthal devel-oped a statistical procedure to estimate the ideological positions of the members of Congress from their votes.7 Using these ideological positions, they then computed the average distance between Democratic and Repub-lican congressmen and senators. Their findings are

summarized in Exhibit W3.9, which shows a striking pat-tern both for the House and the Senate: from 1879 to the 1920s, there was a high degree of polarization between Democrats and Republicans. Then the U.S. political scene went through a rapid period of convergence between the two parties. This trend then got reversed in the 1970s, and by now, the Senate is as polarized as it was in 1879, and the House has become even more polarized than it was then. So the Tea Party wing of the Republican Party, which we discussed briefly earlier in the chapter, may not be an exception, but the continuation of an ongoing wid-ening gap between the two major parties.

Does Democracy Work?The evidence we just presented suggests that, in many instances, democracy doesn’t work exactly as we would’ve liked it to. Politicians do not just represent the wishes of the vot-ers (or the median voter). In the next section, we will also see that political corruption is widespread and can become a major problem for societies. Faced with these facts, some individuals despair and question whether democracy works at all, and whether some other form of government might be preferable. For example, the remarkable growth performance of China since the 1980s under a highly authoritarian regime has raised questions as to whether some forms of dictatorship might be preferable to democracy.

Probably the best answer to this question was given by the former British Prime Minis-ter Winston Churchill, who quipped: “Democracy is the worst form of government, except for all those other forms that have been tried from time to time.” The truth in this statement is that all of the other forms of governments suffer from the same problems as democracy,

W3.2

W3.1

W3.3

W3.4

Section W3.2 | Does Democracy Work? W3-17

W3.2

but only more acutely. If democracy fails at being representative, monarchy and dictator-ship are certainly far less representative. And even though democracies sometimes fail to take decisive action in response to crises, there is no evidence that other forms of govern-ments systematically do better.

In fact, there are several reasons why democratic forms of governments have desirable features. The first, as we have seen, is that they represent the wishes of the citizens, at least under perfect political competition. Even when they deviate from this ideal, as the evidence from the U.S. House of Representatives suggests, the deviation is rather limited. Republican congressmen elected from moderate districts still act as typical Republican congressmen rather than exactly representing the wishes of their median voter, but there is a limit as to how far the Republican or Democratic Party can deviate from the electorate and still win votes.

Another benefit of democracy was first noted by the same Marquis de Condorcet whom we encountered in the context of the Condorcet Paradox. The idea was further elaborated on by the famous nineteenth-century British scientist Francis Galton. It is related to the fact that a group of individuals might often make better decisions than any single indi-vidual might. To see this supposition in action, consider the TV show Who Wants To Be a Millionaire?, where a highly technical question that contestants are unable to answer often gets the correct answer from the audience. A contender is allowed to ask the audience for guidance only once and, after asking a question, is shown a summary of how many people voted for each of the four possible answers. The answer that receives the most votes is often the correct one.

This idea, in fact, goes back to Galton. Galton thought that a group of relatively uninformed individuals would collectively have much more knowledge than any single one of them and make highly accurate predictions. He also saw how relevant this obser-vation might be in an assessment of the success of the democratic form of government. To investigate his ideas, Galton went to an agricultural fair in Plymouth in 1906, where participants were asked to guess the weight of an ox (after the animal was slaughtered and dressed, meaning the head and other parts were removed). Eight hundred people ap-parently took part in this contest. Galton was well aware that these were no experts and in fact, in the Nature article that he wrote in the following year, he stated: “[. . . most of the contestants were . . .] as well fitted for making a just estimate of the dressed weight of an ox, as an average voter is judging the merits of most political issues on which he votes.”9

Perhaps the best way of deciding whether democracy works or not is to imagine that you are choosing which so-ciety to live in. You can choose to live in a non- democratic country like Kuwait, the United Arab Emirates, or China, where (1) individual freedoms are curtailed ( particularly so if you’re a woman in Kuwait or the United Arab Emirates), (2) you do not have the right to elect your leaders or express your views about their actions (unless you happen to be a sycophant of the regime), and (3) free access to information is all but absent. Or you can live in a democ-racy such as Portugal or Spain. The average citizen of Portugal and Spain is poorer than the average citizen of Kuwait or the United Arab Emirates and has never expe-rienced growth rates similar to those in China. But he or she can take part in a vibrant democracy at both local and national levels. Even if politicians are sometimes (even

often) corrupt, citizens feel that they have the right to vote them out of office, and certainly they have relatively free access to information about the politicians. With all of this, would you prefer to live in a non-democratic coun-try like Kuwait, the United Arab Emirates, or China, or a democracy like Portugal or Spain?

Some have argued that these benefits of democracy are counterbalanced by the greater efficiency of non- democracies in achieving economic reform and an effi-cient allocation of resources (without the bickering that democratic legislatures generate). Recent research by Daron Acemoglu, Suresh Naidu, Pascual Restrepo, and James Robinson, however, shows that the evidence points to the opposite conclusion: democracies are more likely to adopt economic reforms and they generate more economic growth relative to non-democracies.8


The Benefits of Democracy

W3.1

W3.2

W3.4


W3.3

The remarkable thing was that, as Galton discovered and reported in his article, when he looked at the 787 valid entries, he found that the middle estimate—the equivalent of the median voter in our model of elections above—was extremely close to the actual weight of the ox. This estimate was 1,197 pounds, while the actual weight was 1,198. Galton con-cluded: “The result seems more creditable to the trustworthiness of a democratic judgment than might have been expected.” To the extent that making judgments on political issues is similar to guessing the weight of an ox (admittedly a leap in comparison), democratic decision making might have additional benefits that most non-democratic forms of deci-sion making lack.

Political corruption refers to illegal and illegitimate actions, such as bribery and misuse of public funds, for private benefit by elected public officials.

Galton’s ideas have recently gained popularity both in academic research and in practice. Nowadays they go under the name of prediction markets. Exactly as in the agricultural fair that Galton visited, in prediction markets people, often non-experts, make predictions or bets over things ranging from outcomes of sports contests to elections. Consistent with Galton’s findings, these collec-tive predictions are often quite accurate. Today, several companies use prediction market software to aggregate information from their employees and customers, fore-cast sales, or manage manufacturing capacity. A manager can simply pose a question online—for example, whether a given task will be completed on time—and employees can bet on what they believe the answer will be. Those with accurate guesses are rewarded either with money or prizes. General Electric even used this approach to gen-erate and collect new business ideas.10

Prediction markets play an important role in politics, too. In the United States, an online betting site called

Intrade allowed users to bet not only on sports events and the Oscars but also on election outcomes. Some political scientists viewed the odds that Intrade offered to be a more accurate prediction about election outcomes than the results of pre-election polling surveys. For in-stance, several months prior to the 2004 presidential election, Intrade’s odds slightly favored George W. Bush, the eventual winner, while most public opinion polls still favored John Kerry (though in other instances Intrade’s odds have been volatile and arguably not very infor-mative). On November 26, 2012, the U.S. Commodity Futures Trading Commission sued Intrade, which is based in Ireland, for running an illegal online betting site, prompting the company to stop taking bets from U.S. users. Small-scale bets on political outcomes are only allowed on the Iowa Electronic Markets, a nonprofit Web site operated by faculty at the University of Iowa for research purposes.


Prediction Markets

Controlling PoliticiansIn a representative democracy, we elect politicians to public office to act as our agents. They derive their authority from the fact that they won the right to act on our behalf in an election. There is no presumption that they should always do exactly what we want. But there is a strong presumption that they are to act in our interest or in the interest of the “public good”—even if this concept is not easy to define, as we saw in the discussion of the Condorcet Paradox.

Politicians, Corruption, and PoliciesWe’ve learned that politicians often do not just follow the wishes of the median voter or the electorate. Congressmen elected in narrow elections, discussed above, are one example. Par-tisan politics is potentially costly to society and makes democracy less representative. But it is likely less damaging as a whole than another failure of representation: political corruption.

Political corruption refers to illegal and illegitimate actions, such as bribery and mis-use of public funds, for private benefit by elected public officials. The common forms of political corruption include the appropriation of public funds for personal use, bribes taken from companies in return for special treatment (including the awarding of public contracts), and support for policies and legislation against the interest of the citizens in

W3.3

W3.1

W3.2

W3.4

Section W3.3 | Controlling Politicians W3-19

W3.3

return for bribes and campaign contributions. Political corruption is related to but different from bureaucratic corruption, which entails similar illegal activities for private benefit by bureaucrats, judges, or law enforcement officers.

Corruption is costly to society and creates inefficiencies for several reasons:

1. It leads to wasteful use of tax dollars. Taxes that are raised to be invested in public goods or redistribution go into the pockets of corrupt politicians or their cronies.

2. It induces a misallocation of resources. This will be the case, for example, when a contract for a new government building goes not to the lowest bidder, which typi-cally comes from the most efficient firm, but to a potentially less efficient firm will-ing to pay a bribe to a powerful politician.

3. It potentially discourages investment: companies, anticipating that they will have to pay bribes to politicians in order to be successful, may be discouraged from enter-ing or investing.

4. It may have highly distortionary indirect effects. For example, to be able to obtain bribes, politicians might erect entry barriers, which could bar new firms from enter-ing the playing field and enable existing firms to make high profits. These profits might in turn translate into beefy bribes for the politicians. If instead there were per-fect competition and free entry, profits would be spread out among firms and conse-quently there would only be limited bribes that politicians could obtain or extract.

5. A political system that allows for bribes can also lead to the disadvantageous selec-tion of politicians. Instead of public-spirited politicians, public office might attract those who are unscrupulous enough to seek public office for personal enrichment.

Corruption is endemic throughout the world. Exhibit W3.10 shows the distribution of political corruption around the world according to the think tank Transparency Inter-national. The exhibit shows high levels of corruption in much of the world, including

Bureaucratic corruption refers to illegal actions for private benefits by bureaucrats, judges, and law enforcement officers.

9.0 – 10 8.0 – 8.97.0 – 7.96.0 – 6.95.0 – 5.94.0 – 4.93.0 – 3.92.0 – 2.91.0 – 1.9

0 – 0.9

Exhibit W3.10 Corruption Map of the World

The countries shown in dark blue have higher levels of corruption as measured by the international think tank Transparency International. (High values of the index of Transparency International correspond to lower corruption.)

W3.1

W3.2

W3.4



Value of Political ConnectionsCorruption at the highest level of office is likely to be particularly costly. The regime of the former Indonesian dictator and president Suharto is generally believed to have been extremely corrupt, with Suharto himself, his family, and many of his key political appointments re-ceiving large bribes in return for political and economic favors. Anecdotal and case study evidence also suggests that firms closely aligned with Suharto received favor-able treatment and thrived during his reign. But docu-menting the presence or the extent of such corruption is not straightforward, since even when it is generally sus-pected, corruption is not publicly observed.

A creative paper by Raymond Fisman overcame this difficulty by investigating how the stock market value of firms that were politically connected to Suharto changed in response to news regarding his health.11 Fisman used data from a consulting group in Indonesia on political connections between different industrial groups and Suharto’s regime. Because of Suharto’s domination over Indonesian politics, it was generally believed that the advantage enjoyed by firms connected to him would disappear after he fell from power—for example, if he died. Between 1995 and 1997, there were episodic ru-mors about a deterioration in Suharto’s health. Fisman showed that following such events, firms connected to Suharto experienced a significant decline in their stock market value. In fact, the magnitude of the connection effect appeared to be very large. For example, the results suggested that the most highly connected firms would lose about 23 percent of their stock market value as a result of Suharto’s death.

Of course, political corruption is not confined to Indonesia or the other developing countries (though Exhibit W3.10 suggests that corruption tends to be more of a problem in poorer countries). As we saw in the case

of Illinois politics, similar problems afflict the United States, and European countries are not free of corruption, either. In the United States, for example, several congressmen, senators, and governors were convicted of corruption charges or were forced out of office because of alleged corruption in recent years. So perhaps you should expect to see in the United States exactly the same things as in Indonesia.

A follow-up paper by Fisman and colleagues applied the same methodology to the connections of Vice Presi-dent Richard (Dick) Cheney.12 Famously, not only was Cheney one of the most influential vice presidents but he was also closely tied to several companies, including the giant oilfield services company Halliburton. There were rumors that connections to Cheney mattered, for example, in the awarding of contracts in occupied Iraq. Fisman and colleagues looked at changes in the stock market values of companies connected to Cheney through his past cor-porate board linkages following several important events. These included news concerning Cheney’s heart attacks, the likelihood of a Bush victory in 2000, and the likelihood of war in Iraq, as measured by prediction markets. In con-trast to the Indonesia study, however, this research did not find any evidence that connections to Cheney increased stock market value of the connected companies. One in-terpretation for this finding is that political institutions in the United States are stronger than those in Indonesia and thus ensure that political corruption at the top is more limited. But recent research by Daron Acemoglu, Simon Johnson, Amir Kermani, James Kwak, and Todd Mitton shows that during the recent financial crises, connections may have played a more important role, so even under stronger political institutions, connections are valuable and need to be watched out for by the media and the electorate.13

W3.3

almost all of South America, Africa, and Asia. Notably, the United States has higher levels of corruption than Canada and several European countries.

Although political corruption is a major concern for political economy, the social costs of bureaucratic corruption may be just as severe. Although in most societies only

a few politicians are in a position to extract large bribes and cre-ate major inefficiencies, smaller pockets of corruption can exist at lower levels of society. Corruption at these levels can be just as paralyzing for economic activity. If you have to pay a bribe every time you ship some goods, pass through checkpoints, need to get a license or a government document, or encounter a police officer, the implications can be quite severe both for the quality of life and for economic activity. And much case study evidence suggests that life in many developing countries involves exactly this type of en-demic bureaucratic corruption.

Some social scientists, such as Nathaniel Leff and Hernando de Soto, have suggested that bureaucratic corruption may help grease the wheels of the markets in developing economies, where regulations can ©

Mar

yna

Ples

hkun

/Shu

tter

stoc

k

W3.1

W3.2

W3.4

Section W3.3 | Controlling Politicians W3-21

The level of corruption in the United States in Exhibit W3.10 is higher than that in Canada or Scandinavian countries. This is not just a figment of the imagination of Transparency In-ternational. Local politics has been notoriously corrupt in the United States throughout much of the twentieth century. Chicago and Boston were among the most corrupt cities, and Chicago and Illinois politics are still viewed as highly corrupt. Some of this was revealed painfully to the public

in 2008 when Illinois Governor Rod Blagojevich massively misused his authority to appoint a replacement for for-mer senator Barack Obama, who was elected president in November. U.S. Attorney Patrick Fitzgerald presented evidence showing that Blagojevich attempted to auction off the newly vacated senate seat to the highest bidder. In December 2011, Blagojevich was convicted of eighteen cor-ruption charges and sentenced to fourteen years in prison.


Corruption in Chicago

W3.3

In addition to political corruption exhibiting itself as the value of political connections, corruption in the ranks of bureaucrats and law enforcement officers in Indo-nesia also appears to be endemic. A creative paper by economists Benjamin Olken and Patrick Barron over-came the difficulty of observing the extent of corruption by designing a field experiment to collect data on brib-ery in the Aceh province of Indonesia.15 They arranged enumerators to accompany truck drivers on 304 trips to and from Aceh. The enumerators’ only job was to pas-sively observe and record bribe transactions. This design enabled Olken and Barron to document the extent of bribery and also to test several simple ideas about the economics of corruption.

They found widespread corruption mostly at check-points, where truck drivers were allowed to pass through only by making illegal payments to police and military personnel. The enumerators documented more than 6,000 illegal payments, with an average value per trip equal to 13 percent of the total cost of the trip. They also

showed that police and soldiers at checkpoints near the end of the trip demanded higher bribes than those at the beginning. This is in line with the implications of a simple theory of corruption, where the bargaining power of bu-reaucrats or law enforcement officers demanding bribes determines the extent of the bribe. Preventing passage of a truck near the destination of its trip, after the driver had already paid and cleared all previous checkpoints, was more costly to the driver and thus increased the bargaining power of the police and military officers. The evidence suggests that this translates into higher bribes.

Finally, Olken and Barron also exploited the fact that during this period Indonesian soldiers were withdraw-ing from Aceh, so the number of checkpoints that truck drivers had to go through declined. Olken and Barron showed that when there were fewer checkpoints, the average bribe per checkpoint increased, which is also consistent with the idea that when the police and military officers have greater bargaining power, they expect and are able to demand greater bribes.


Bureaucratic Corruption in Indonesia

be stifling and bribes are just one way of overcoming them.14 Yet most economists believe and most evidence suggests that bureaucratic corruption lowers economic efficiency. What makes bureaucratic corruption perhaps even a more challenging problem than political cor-ruption is the difficulty of eliminating it. Keeping tabs on politicians is not easy, but it is doable. As we see next, the media plays an important role as the watchdog of politicians. Keeping tabs on and incentivizing a large bureaucracy is much harder, and once corruption comes close to becoming a social norm, eliminating it becomes very difficult.

Electoral Controls on PoliticiansThere are two ways to think about all of this evidence of political corruption. One is to wonder why politicians are often corrupt and unscrupulous. The other is to wonder why politicians are not even more corrupt. Many politicians in the United States and Europe control hundreds of millions of dollars and have the ability to misuse their power for per-sonal or ideological gains. However, although corruption is not a rare occurrence as we have indicated, it is still the exception rather than the rule, at least for national and federal

W3.1

W3.2

W3.4


Are voters more likely to vote against corrupt and incom-petent politicians? The view that elections are effective controls over politicians presumes that the answer to this question is yes, though direct evidence that supports this presumption is not easy to come by. A creative study by economists Claudio Ferraz and Frederico Finan used a large audit study of the Brazilian federal government con-cerning corruption at the local level.16 The federal anti-corruption program in Brazil, initiated in 2003, randomly selected municipal governments and audited their use of federal funds. The outcomes of these audits were made public. For example, the audits reported the number of violations connected to corruption for mayors. Ferraz and Finan showed that mayors who were found out to be cor-rupt in the audit report were significantly less likely to be reelected—a finding that supports the premise that vot-ers can use elections to expel corrupt politicians.

The Brazilian audit and election data also enabled an investigation of whether these mayors would have been voted out of office even without the audit study, for ex-ample, because voters already knew about their corrup-tion. To investigate this issue, Ferraz and Finan compared mayors audited right before and right after the 2004 municipal elections and compared mayors with the same level of reported corruption. Consider, for example, two mayors both with the same number of corruption-related violations in two neighboring municipalities. Suppose, however, that there is one difference between this pair of

mayors: one of them is audited just before the 2004 elec-tions and the other one just after. So they have exactly the same level of corruption, are running for office in similar areas, and have otherwise similar track records. The only difference is the information that the voters have about them in 2004. For the first mayor, they know that he has two corruption-related violations. For the second mayor, they suspect that he is corrupt, but they don’t have it doc-umented by the federal audit program.

Ferraz and Finan found that even in this comparison, those whose corruption was publicly disclosed by the au-dit study were significantly more likely to be voted out of office. In particular, a mayor with two corruption-related violations revealed before the 2004 municipal elections appeared to be 7 percentage points less likely to be re-elected relative to a mayor who also had two violations but whose corruption was revealed only after the elec-tion. This suggests that voters were not fully informed about the extent of the corruption of their politicians and that when they learned about it, they became more likely to vote the politicians out of office.

This study provides yet further evidence that corruption is endemic in many political and bureaucratic situations, but it also shows that providing more information to vot-ers about the extent of corruption might help, as it will enable them to use elections more effectively to control and unseat corrupt politicians and presumably reduce the scope for unscrupulous behavior by politicians.


Voting Corrupt Politicians out of Power

W3.3Voters can reward good behavior with reelection and punish bad behavior by switching their votes.

politicians. So why is it that selfish and ambitious politicians do not misuse their office much more than they currently do? While social norms of honesty might go some way toward explaining the relatively limited amounts of political corruption, a large part of the answer lies in the structure of democratic political institutions—the fact that elections are a good way of disciplining politicians.

In fact, when politicians are self-interested and fail to act as perfect agents of the voters, elections are a way of incentivizing politicians by offering them rewards and punishments—reelection versus being voted out of office. We can think of the voters as having delegated cer-tain tasks to the politician, and they need to incentivize the politician so that he or she performs these tasks and does not attempt to look after his or her own interests at

the expense of the voters. In an electoral system, voters can attempt to control politicians by rewarding those who have pursued successful policies (assuming the availability of un-distorted information about the impact of enacted policies) and who have not deviated from their promises. Voters can reward good behavior with reelection and punish bad behavior by switching their votes. But how effective are such controls? The next Letting the Data box sheds some light on this question.

Media and PoliticsExercising control over politicians relies on information; it is impossible to control and regulate any kind of behavior unless we are informed about it. In day-to-day politics, it is almost impossible for most voters to know the exact actions that politicians take, and it is often difficult for them to know whether a politician is acting honestly or is knee-deep

W3.1

W3.2

W3.4

Section W3.4 | Market Failure Versus Government Failure from a Political Economy Perspective W3-23

in corruption. One of the many roles of free media is to act as a conduit for this type of information. Independent journalists report on what politicians do and sometimes find evi-dence of corrupt behavior. Such behavior then acts as the basis of the voting behavior of the citizens in future elections.

This view suggests that information provided by the media is critical for the proper functioning of democracies. If true, we should indeed observe news affecting voting be-havior, and we should even observe in some cases corrupt politicians attempting to control news and media outlets. The evidence is largely consistent with these expectations. We just saw how Brazilian voters use information on corruption in their voting behavior. Evidence from another Latin American country, Peru, presented in the next Letting the Data Speak box, provides a glimpse into politicians’ (particularly corrupt politicians’) efforts to control the media.

In fact, media can affect political views and voting behavior not only by uncovering cor-ruption but also by informing voters with news and commentary. Still, there is no guarantee that media outlets themselves will act as unbiased arbiters. Newspapers, magazines, and TV outlets often have their own ideologies, and might spin news in a particular manner either unwittingly or purposefully as a way of influencing voters. Existing evidence shows that media outlets do so, and they are often successful in swaying public opinion, at least to some degree.

Can we do better by paying politicians for their perfor-mance rather than simply voting them out of office when we are dissatisfied with them? Pay for performance might give politicians greater incentives to please the voters. So if it were your choice, would you be in favor of paying big bonuses to politicians?

There are both pros and cons to such a proposal. On the pro side, the large bonus for successful and hon-est behavior might discourage corruption and lead to better policies. On the con side, however, there might be more weighty matters. First, pay-for-performance or other types of “high-powered” incentives often backfire inducing economic agents to take actions in order to create the appearance of good performance

even when they are performing poorly. For example, politicians might choose costly actions to hide the true state of public finances if their pay depends on the bud-get deficit (though the example of Greece, where the government hid the true level of indebtedness of the country from voters and from the rest of the European Union, suggests that politicians might do that even when there is no pay-for-performance). Second, pay-for-performance for politicians might create the wrong type of selection. Instead of public-spirited politicians, it might be particularly attractive to those interested in making a quick buck. Put differently, the ethos of public service might not sit well with large performance-based bonuses.


Should We Pay Politicians for Performance?

W3.3

W3.4 Market Failure Versus Government Failure from a Political Economy PerspectiveAs we have seen in this chapter, even if politicians just implement the wishes of voters, the conflict of interest between different individuals makes the choice of policies a nontrivial matter. Part of the subdiscipline of political economy studies how these decisions are made under different sets of political institutions. And we have also seen that there is no reason to presume that politicians will not be self-interested economic agents. If so, they will pursue their own agendas and might even act in corrupt ways to enrich themselves at the expense of society, unless they are controlled effectively by elections, the media, or other institutions.

The distortions caused by political economy conflicts and politicians and bureaucrats acting in their own interests rather than as representatives of the electorate and society are referred to as government failure. Government failures imply that every time we increase

Government failure refers to potential social costs of government actions due to political economy conflicts and political and bureaucratic corruption.

W3.1

W3.2

W3.3


W3.4

The evidence from Peru suggests that politicians are very keen on controlling the media. But does the media indeed affect how people vote? Economists Stefano DellaVigna and Ethan Kaplan provide an answer to this question by studying the expansion of the Fox News cable channel across U.S. towns.18 It is generally accepted that the cov-erage at Fox News is more right-wing and pro-Republican than at other TV stations and most other media sources. One might then conjecture that exposure to Fox News might increase the support for the Republican Party, be-cause undecided voters who received their information from Fox News might be more likely to decide to vote Republican, or because Fox News might mobilize more right-leaning voters to turn out. On the other hand, it is also possible that voters recognize that some of the news they receive is biased and may filter it out.

DellaVigna and Kaplan exploited the fact that the Fox News channel was introduced in October 1996 and then

spread across different towns slowly, depending on its agreements and the decisions of local cable companies. It had reached only about 20% of towns by 2000. DellaVi-gna and Kaplan then compared the change in the vote share of the Republican candidates between the 1996 and the 2000 elections. The 1996 elections were held be-fore the introduction of Fox News, so differential change between towns with and without access to the channel could be interpreted as the impact of Fox News. They found that the Republican candidate gained between 0.4 to 0.7 percentage points in towns where the population had access to the Fox News cable channel compared to similar towns without such access. This evidence is con-sistent with the hypothesis that information provided by media sources, whether biased or unbiased, can have a significant effect on voting patterns.


The Fox News Effect

the set of activities that are delegated to the government, there will be more conflict of interest among voters about exactly what action should be taken, and there will be more areas in which politicians can pursue their own agendas or will have more money to use for their own benefit.

On the other hand, if left alone, markets themselves can fail to arrive at socially efficient outcomes—for example, in the provision of public goods or protection of common pool resources. Market failures motivate society to delegate certain actions to governments—for example, providing and enforcing policies to regulate pollution.

Dictators are often aware of the importance of free media and do their best to fight it. Their actions sometimes reveal the costs of free media to dictators and its value to society. A colorful illustration of this situation comes from Alberto Fujimori’s rule in Peru. Though originally democratically elected, Fujimori soon set up a dictatorial regime in Peru, mounting a coup while still in office in 1992. Thereafter, though elections continued, Fujimori built a corrupt re-gime and ruled through repression and bribery. To do so, he relied heavily on his right-hand man, Montesinos, who was heading the powerful national intelligence service of Peru. Montesinos was an organized man, so he kept good records of how much he paid different individuals in order to buy their loyalty, even videotaping many actual bribery incidences. There was a logic to this. Beyond just record-keeping, this evidence made sure that the accomplices were now on record and would be as guilty as Fujimori and Montesinos if they ever blew the whistle.

After the fall of the regime, these records fell into the hands of journalists and authorities and were studied by economists John McMillan and Pablo Zoido.17 The sums paid for bribery were very revealing about the value of the media to a dictatorship. A Supreme Court judge was worth $5,000–$10,000 a month, and politicians were also paid similar amounts. But when it came to newspapers and TV stations, the sums were in the millions. Fujimori and Montesinos paid $9 million on one occasion and over $10 million on another to control TV stations. They paid over $1 million to a mainstream newspaper, and they paid amounts to other newspapers ranging between $3,000 and $8,000 per headline. Fujimori and Montesi-nos thought that controlling the media was much more important than controlling politicians and judges. One of Montesinos’s henchmen, General Bello, summed up this sentiment in one of the videos by stating: “If we do not control the television, we do not do anything.”


The Value of the Media in Peru

Therefore, the costs of market failures often have to be stacked up against the costs of government failure. This perspective certainly does not imply or condone an extreme liber-tarian position where we should tolerate market failures because governments are not infal-lible. But it highlights that we should try to quantify both the costs of market failures—for example, what will society lose if a factory continues to pollute the air with no enforced abatement policies?—and the costs of government failures—for example, what will society lose if elected officials pursue mainly their own agendas? This type of cost-benefit analysis might then give us a better handle on what types of activities should be within the sphere of the government and which ones should be left to markets, even when markets are imperfect.

Taking into account the potential costs for government failure, should we have governments also intervene in order to prevent individuals from making mistakes? A paternalistic perspective suggests that such interven-tions might be justifiable. There are indeed some areas where paternalism might be both justified and its political economy costs might be limited. For example, perhaps mandating car seats for infants is justified since parents might underestimate the probability and the potentially severe consequences of accidents, which could be quite harmful to infants and small children. The room for cor-ruption and political economy conflicts in this case is probably limited.

But in many other cases, including regulation of alcohol, tobacco, and drugs, the room for political and bureaucratic

corruption might be great, even if there is some argu-ment for paternalistic actions. An interesting way of think-ing about this problem is to ask whether dictatorships or democracies intervene more in supposedly paternalistic ways. Although a quantitative analysis of this question seems difficult, evidence is quite clear that dictatorships often use paternalistic arguments to limit the access of their citizens to information. Dictatorships also seem to regulate such matters as religious observance and dress code much more than democracies. These observations certainly do not suggest that paternalism makes democ-racy less likely, but raises the possibility that paternalistic actions might sometimes be excuses for governments to intervene for their own agendas or increase their control over society or some subset of it.


Reevaluating Paternalism

Summary

Political economy is the study of the conflict of interest between individuals and groups over policy choices, how that conflict is resolved, and how citizens influence government policy and control politicians. Political economic analysis enables us to better evaluate the costs and benefits of government intervention in the economy when this intervention is carried out by politicians who are only imperfectly responsive to the wishes of the citizens.

We have seen that there is much more to governments’ and politicians’ choices of policies than first meets the eye. Like other economic agents, rulers, presidents, representatives, and other politicians—whether elected or self-appointed—are not selfless. When they choose policies, they consider the relevant trade-offs. This includes balancing the potentially conflicting preferences of different voters. But they also take into account the benefits that they themselves will derive from different policies and actions.

In monarchies and dictatorships, there are few checks on rulers’ behavior and not much reason to expect that they will behave in a way that is in the interests of the society at large. In democracy, politicians have to compete for office. Even if democracy is in practice far from perfect, it does empower voters to replace politicians who are corrupt, incompetent, or unresponsive to the wishes of their constituents.

Summary W3-25


Collective preferences, as expressed by voting behavior, do not always exhibit transitivity. We say that preferences are transitive if the following is true: if Choice A is preferred to Choice B and Choice B is preferred to Choice C, then Choice A should be preferred to Choice C. The Condorcet Paradox highlights how there can be a failure of transitivity in voting behavior. The Arrow Impossibility Theorem shows more generally that a clear order of preferences of three or more alternatives cannot be converted into a community-wide ranking in the voting process, thus implying that it is impossible to have an ideal voting system.

A voter’s most preferred policy is his or her political bliss point. Such preferences are said to be single-peaked if the voter prefers policies closer to his or her political bliss point than to those that are far away. A voter in the middle of a distribution of political bliss points is referred to as the median voter; the Median Voter Theorem states that when preferences are single-peaked, the political bliss point of the median voter will win a majority of votes over any other policy option.

Politicians do not always follow the wishes of the median voter or the electorate. Worse, they sometimes engage in corrupt and highly dishonest acts. Bribery and the misuse of public funds for private benefit are two common forms of political corruption. Illegal activities like these also exist on lower levels in the form of bureaucratic corruption. Whether and how much politicians will choose policies that the citizens prefer and refrain from corruption depends on the political institutions in place (for example, whether the country is a democracy) and on how much information is available about their actions and choices.

When politicians are not scrutinized by voters, the media, or the courts, they will be relatively free to pursue their own interests and neglect their constituents.

Ultimately, do politicians follow the voters’ wishes? Not always. But in a democracy, they at least have to listen. They have to pay attention to the views of the voters and their economic well-being. They have to show they’re worth a vote. Democracy is as imperfect as the representatives who rise to power. They might not be fully responsive to the wishes of the majority, but they can lose their job any day if they ignore too many people too many times. Democracy’s strength comes not just from the freedom to elect leaders but also from the capacity to expel them.

Key Termsconflict of interest p. W3-3political economy p. W3-3representative democracy p. W3-4dictatorship p. W3-4transitivity p. W3-5

Condorcet Paradox p. W3-5political bliss point p. W3-8single-peaked p. W3-8median voter p. W3-9Median Voter Theorem p. W3-9

perfect political competition p. W3-11policy convergence p. W3-11political corruption p. W3-18bureaucratic corruption p. W3-19government failure p. W3-24

Questions

1. Differentiate among the following three political regimes—monarchy, dictatorship, and representative democracy.

2. What does it mean to say that an individual’s preferences are transitive?

3. Explain the Condorcet Paradox.

4. According to Kenneth J. Arrow, what are the three condi-tions that any reasonable voting scheme should satisfy? What does Arrow’s Impossibility Theorem state?

5. When is an individual said to have single-peaked policy preferences?

6. Explain the Median Voter Theorem.

7. When is policy convergence, also known as Downsian policy convergence, likely to occur among political parties?

8. Are you likely to find perfect political competition in the real world? Explain.

9. What did the former British Prime Minister Winston Churchill mean when he said, “Democracy is the worst form of government, except for all those other forms that have been tried from time to time”?

10. How does political corruption differ from bureaucratic corruption?

11. How does political corruption affect society?

12. Explain how elections in countries with democratic po-litical institutions can place controls on corruption.

13. What are the likely consequences of providing monetary incentives to politicians for good performance?

14. Does government failure necessarily condone an extreme libertarian point of view where the government does not intervene in markets at all?

All questions are available in MyEconLab for practice and instructor assignment.

Problems

1. In his book Power and Prosperity, the economist Man-cur Olson theorized that a country is better off under a dictator who maintains power for a long period, such as Mobutu, than under a short-term, unstable dictatorship. Olson maintained that, while both are undesirable forms of government, the former is less disruptive to the econ-omy than the latter. Discuss why this assertion may or may not be true.

2. The following table shows the preferences of three groups of voters. The decision on immigration in this country is taken based on majority vote. There are three possible policy options—ban immigration, allow a limited number of migrants, or allow all migrants.

What would be the collective choice between the follow-ing three scenarios?

a. Allowing a limited number of migrants versus allow-ing all migrants?

b. Allowing all migrants versus banning immigration?

c. Banning immigration versus limiting the number of migrants?

d. Given that the individual preferences of each group are transitive, examine whether the collective prefer-ences of the three groups are transitive.

Group 1 Group 2 Group 3

Most preferred

Ban immigration

Allow all migrants

Allow a lim-ited number of migrants

In between Allow a lim-ited number of migrants

Ban immigration

Allow all migrants

Least preferred

Allow all migrants

Allow a lim-ited number of migrants

Ban immigration

3. Consider the collective preferences of Adam, Bob, and Charlie regarding taxes that are given in Exhibit W3.1. One of the more unexpected and intriguing aspects of preferences such as these is that the outcome can be pre-determined given the order in which pairs of the alterna-tives are voted on. This is called “pairwise voting.” So, for example, given alternatives A, B, and C, first voters choose between A and B, and then the winner of that con-test is voted on versus C.

a. Show the order of voting between pairs that will result in low taxes being adopted.

b. Show the order of voting between pairs that will result in intermediate taxes being adopted.

c. Explain what feature of voters’ preferences makes the order of voting matter.

All problems are available in MyEconLab for practice and instructor assignment.

Problems W3-27


4. The following table shows the popularity of three may-oral candidates among three groups of voters in a country.


Candidate A 50% 30% 20%

Candidate B 20% 50% 30%

Candidate C 30% 20% 50%

a. Plot the preferences of the three groups on a graph. Identify the political bliss point of each group of vot-ers. Would the Condorcet Paradox apply to these three groups of voters?

b. Now look at the following table. Plot the preferences of the three voter groups on a graph and identify the political bliss points of the three groups of voters. Do the three groups of voters exhibit single-peaked pref-erences? What does this imply for the outcome of a majority vote?


Candidate A 20% 30% 50%

Candidate B 30% 50% 30%

Candidate C 50% 20% 20%

5. There is an alternative social decision method called a Borda Count. In this technique, alternatives are assigned a number based on the ranking of the alternatives by each agent. For example, suppose a society is trying to decide among four alternatives. If a voter likes alternative A bet-ter than any of the other three alternatives, A would re-ceive a 4. If she likes C better than any of the remaining two alternatives, C would receive a 3. And so on.

Suppose Adam, Bob, and Charlie are now voting on the is-sue of gun control. They are considering four alternatives:

Ban private ownership of guns entirely (B)

Severely restrict ownership of guns (S)

Moderately restrict ownership of guns (M)

Lightly restrict ownership of guns (L)

Here are the preferences of the three, from most preferred to least preferred:

Adam Bob Charlie

B B L

M M B

L L S

S S M

a. Construct a Borda Count for the alternatives. What is the social ranking of the four based on the Borda Count?

b. Now assume the Supreme Court has ruled that bans and severe restrictions on guns are unconstitutional. So, neither B nor S is an option.

Construct a new Borda Count on the remaining alternatives—light restriction and moderate restriction. (Use 2 to designate the more preferred alternative and 1 to designate the less preferred.) How does the rank-ing of these two alternatives compare with the ranking between the two you found in part (a)?

c. Which condition of the Arrow Impossibility Theorem is violated in light of your answer to part (b)? Explain.

6. India and China have very different political institutions. Some commentators feel that China’s autocratic politi-cal regime allows decisions to be made faster and more efficiently. Implementing economic policies is also easier in China. India, on the other hand, is the world’s largest democracy. Decision making in India is sometimes lim-ited by electoral politics or judicial constraints. As a result, decision making is slow and implementation is some-times ineffective. This has led some people to believe that democracy is not always conducive to economic growth; autocratic countries can make and implement decisions faster. Would you agree with this analysis? Explain your answer.

7. You are closely following the elections in your country this year. The two main candidates who are running for president belong to the two biggest parties that are on opposing ends of the political spectrum—the Conser-vative Right and the Left Alliance. There are also a few independent candidates. You have decided that you will cast your vote depending on each candidate’s stance on various issues like foreign policy, gun rights, taxes, gay marriage, and healthcare. As the campaigns heat up, you notice that the viewpoints expressed by both parties seem to converge to the center rather than being extreme left or right opinions. Given that the candidate with the highest number of votes wins the elections, what do you think could explain this convergence?

8. The chapter discusses Transparency International’s rank-ing of countries by levels of corruption. It is interesting to note some of the least corrupt countries on the list and look at the correlation of their ranking with the perfor-mance of their economies. For example, the least corrupt countries are, in order, Denmark, New Zealand, Finland, Sweden, and Norway. Each is a wealthy country with a very high standard of living, low unemployment, and low inflation.

This is in contrast to countries rated most corrupt by Transparency International. Among those at the bottom of the list (that is, those with the most corrupt govern-ments) are Haiti, Syria, Sudan, North Korea, and Soma-lia. These are uniformly poor nations with a low standard of living and dismal economic growth performance.

Data from http://www.transparency.org/cpi2013/results

Speculate on the reasons why a more corrupt country would be more likely to generate slow or nonexistent economic growth than one with a more honest and re-sponsible government.

9. In many developing countries, starting a business in-volves navigating a labyrinth of rules and regulations and red tape. Paying a bribe to expedite paperwork is a com-mon practice. As Samuel Huntington puts it, “the only thing worse than a society with a rigid, overcentralized, dishonest bureaucracy is one with a rigid, overcentral-ized, honest bureaucracy.” The argument is that bribes increase economic activity because, without bribes to grease bureaucrats’ palms, starting a business would be an extremely long, drawn-out process and would prob-ably drive investors away. Is this always the case, or does bureaucratic corruption involve other costs? Explain your answer.

10. It is interesting to note that several countries around the world actually make voting compulsory. According to the CIA’s World Factbook, as of August 2013 voting is man-datory in 22 countries. Although this law is not enforced in many of these countries, in several nations citizens who do not comply are subject to fines.

Give arguments for and against making voting compulsory.

11. A study by the economist Rudiger Ahrend found that the presence of corruption in a country does not affect the freedom of the press but having a free press greatly lim-its corruption. The study suggests that strengthening the freedom of the press is extremely important in fighting

corruption. Why do you think this is the case? Explain your answer.

12. Much of the work in the subdiscipline of political econ-omy, which is the main focus of this chapter, was pio-neered by Nobel laureate James Buchanan. He believed that the basic framework of economic analysis was useful in studying political and governmental activity. Addition-ally, he advocated that economists pay more attention to the political realm within which economic activity oc-curs. Here is a quote from the lecture he gave at the Nobel awards ceremony in 1986:

The political economist who seeks to offer normative advice must, of necessity, concentrate on the process or structure within which political decisions are observed to be made. Existing constitutions, or structures or rules, are the subject of critical scrutiny.

Read Buchanan’s Nobel lecture in its entirety at: http://www.nobelprize.org/nobel_prizes/economic-sciences/laureates/1986/buchanan-lecture.html

Based on your reading, write a brief essay about what you think is the gist of his ideas concerning the relationship between economics and politics.

(Note: This is not an easy read, but you will still be able to glean much of the essence of Buchanan’s argument if you read it carefully.)

Endnotes 1 National Income and Product Accounts, Tables 1.1.5 and 3.1.

2 Michaela Wrong, In the Footsteps of Mr. Kurtz: Living on the Brink of Disaster in Mobutu’s Congo (New York: Harper Perennial, 2002).

3 Kenneth J. Arrow, Social Choice and Individual Values (New Haven: Yale University Press, 1951, 2nd ed. 1963).

4 Anthony Downs, An Economic Theory of Democracy (New York: Harper, 1957).

5 David S. Lee, Enrico Moretti, and Matthew J. Butler, “Do Voters Affect or Elect Policies? Evidence from the U.S. House,” Quarterly Journal of Economics 119, no. 3 (2004): 807–859.

6 Per Pettersson-Lidbom, “Do Parties Matter for Economic Outcomes? A Regression-Discontinuity Approach,” Journal of the European Economic Association 6, no. 5 (2008): 1037–1056.

7 Nolan McCarthy, Keith T. Poole, and Howard Rosenthal, Polarized America: The Dance of Ideology and Unequal Riches, Vol. 5 (Cambridge, MA: MIT Press, 2006).

8 Daron Acemoglu, Suresh Naidu, Pascual Restrepo, and James A. Robinson, “Democracy Does Cause Growth,” NBER Working Paper 20004 (2014).

9 Francis Galton, “Vox Populi,” Nature 75, no. 1949 (1907): 450–541.

10 Michael Totty, “Business Solutions: How to Decide? Create a Market,” Wall Street Journal, June 19, 2006.

11 Raymond Fisman, “Estimating the Value of Political Connections,” American Economic Review 91, no. 4 (2001): 1095–1102.

12 David Fisman, Raymond J. Fisman, Julia Galef, Rakesh Khurana, and Yongxiang Wang, “Estimating the Value of Connections to Vice-President Cheney,” The B.E. Journal of Economic Analysis & Policy 12, no. 3 (2012).

13 Daron Acemoglu, Simon Johnson, Amir Kermani, James Kwak, and Todd Mitton, “The Value of Connections in Turbulent Times: Evidence from the United States,” NBER Working Paper 19701 (2013).

14 Nathaniel H. Leff, “Economic Development Through Bureaucratic Corruption,” American Behavioral Scientist 8, no. 3 (1964): 8–14; and Hernando de Soto, The Other Path: The Invisible Revolution in the Third World (New York: Harper, 1989).

15 Benjamin A. Olken and Patrick Barron, “The Simple Economics of Extortion: Evidence from Trucking in Aceh,” Journal of Political Economy 117, no. 3 (2009): 417–452.

16 Claudio Ferraz and Frederico Finan, “Exposing Corrupt Politicians: The Effect of Brazil’s Publicly Released Audits on Electoral Outcomes,” Quarterly Journal of Economics 123, no. 2 (2008): 703–745.

17 John McMillan and Pablo Zoido, “How To Subvert Democracy: Montesinos in Peru,” Journal of Economic Perspectives 18, no. 4 (2004): 69–92.

18 Stefano DellaVigna and Ethan Kaplan, “The Fox News Effect: Media Bias and Voting,” Quarterly Journal of Economics 122, no. 3 (2007): 1187–1234.

Endnotes W3-29

Documents

Web C Political Economyhomepage.ntu.edu.tw/~luohm/econ2015f/chapter_w03_text.pdf · Mobutu. The country, now called the Democratic Republic of the Congo, is the third largest country