Erik Moberg:

The Expanding Public Sector - a Threat to Democracy?

ã Erik Moberg and the City University of Stockholm



1 The expansion of the public sector
2 Political parties and party systems
3 The distribution of power
4 Delegation and instruction
5 General and specific instructions
6 The competition for a place in the executive
7 The competition for votes
8 An asymmetry favoring the public sector
9 The empirical evidence
10 Conclusions
11 Implications



The political competition for power is considered as a main force propelling the expansion of the public sector in democracies. The character of the competition, and thereby the expansive force, is thought to depend on the type of constitution in the country concerned. The most rapid expansion will probably occur in parliamentary countries with proportional elections. Anyway, expansion seems to be an almost unavoidable side-effect of democracy. The possibility that a democracy's life is limited to the time it takes for the public sector to grow from low to impossibly high levels is therefore finally mentioned and commented on.

1 The expansion of the public sector

It is well known that the public sector has increased continuously and considerably in a lot of countries, at least in modern times. In several Western European countries the public sector is now well above 50% of GDP. Now, even if there may be some occasional exceptions to this general pattern, even if the size of the public sector differs considerably between the countries, and even if the speed of the sector's expansion differs considerably between different periods of time, the general tendency towards expansion seems nevertheless almost unavoidable and in the nature of things. It is therefore hardly surprising that this tendency is widely and since long observed and acknowledged, and that many efforts to explain it have been done.

The most well-known of these explanations is perhaps the one suggested by Adolph Wagner, who argued that increasing per capita income, and increasing standard of living, in a society, gave rise to an increasing public sector (Wagner's law). Other scholars have explained the expansion, at least in democracies, as a side-effect of the political struggle for power. Still other types of explanations have been brought forward.1 In spite of the many explanations it is, however, widely agreed that public sector expansion is still a poorly understood phenomenon. A lot of empirical studies have made it clear that no proposed simple explanation is capable of clarifying the involved mechanisms single-handedly. Rather several explanations together carry the truth. It is, I submit, also likely that further elements, which have not so far been taken into account, can contribute to our understanding of the expansion of the public sector.

Political competition in democracies undoubtedly contributes importantly to expansion. At the same time I believe, however, that, the political competition takes different forms, and therefore also creates different effects in different democracies depending on their type of constitution. This is the theme that I am going to develop in this paper.

The theme, it should be noted, fits well into an intellectual tradition initiated by Joseph A Schumpeter. In addition to his other achievements he made, as we know, a pioneering study of democracy and thereby also inspired the later emergence of public choice theory.2 At the heart of Schumpeter's analysis lies a distinction between the motives of the political actors, and the societal effects of their activities, analogous to Adam Smith's distinction between the butcher's, baker's and brewer's profit motives, and the effects of their undertakings. In either case the effects depend on the structure of incentives facing the actors. In Smith's case the incentives are, of course, those of the market, whereas, in the political case, the incentives are those of the competition for power. In the former case the effects are, as we know, on the whole beneficial. In the political case there is no such generally accepted conclusion. On the whole the analysis still remains to be done. This paper is an effort to contribute to that task.3

2 Political parties and party systems

In order to study the impact of constitutions it is necessary to classify them in some relevant way. A reasonable suggestion is that the distribution of power, as determined by the constitution, should be of importance. Here, I hypothesize that the constitutional elements most likely to affect that distribution are those determining the properties of political parties and party systems. These elements, in turn, I submit, are the rules for appointing the executive, and the electoral laws.4 When discussing the influence of these rules I am mainly interested in two clusters of properties of parties and party systems, namely:

There are two principal methods for appointing the executive: (1) the one used in parliamentary systems, and (2) the one in presidential systems. According to the parliamentary method the people first elect the legislature, which then, in turn, appoints the executive. In a pure parliamentary system the executive, furthermore, can remain in office only as long as it enjoys the confidence, or support, of a majority in the legislature. This requirement is therefore often referred to as the parliamentary principle. The presidential method, on the contrary, means that separate elections are held for appointing a president and thereby also the rest of the executive. In a parliamentary country there are thus, at the national level, popular elections only for electing the legislature whereas, in a presidential country, there are two main types of elections, those for electing the legislature, and those for electing the executive.

A parliamentary system depends, for its functioning, on the existence of stable, centralized and cohesive political parties in a way that a presidential system does not. The reason is that the parliamentary support, in order to be reliable and lasting, cannot be anonymous. A support expressed by an ad-hoc, transient majority of individual members of the legislature cannot, it is easy to realize, have much value. The support has to be expressed by a few stable and identifiable actors, which, in effect, means political parties.

A parliamentary system is, however, not only dependent on stable, centralized and cohesive parties; conversely it also gives strong incentives for the formation of such parties, and sometimes also for forming big parties.5 The reason is that those properties enhance a negotiating party's credibility and reliability and thereby its chances to become a member of the executive, a membership which often is quite attractive, or even lucrative. This attractiveness, in its turn, is, at least to some extent, related to the fact that, in a democracy using the majority rule, a majority can exploit the outsider minority, for example by taxing it. In a parliamentary democracy this majority power is permanently anchored to the executive.6

Continuing with the electoral laws used for appointing the members of the legislature it is enough to consider two main types of systems. First there is the system with single-member constituencies in which, in each constituency, the candidate who gets a plurality of the votes, is elected. Then there is the system with multi-member constituencies in which the mandates are distributed to the parties in proportion to their votes.

The electoral system affects the parties in two ways. First, the plurality system has a strong tendency to reduce the number of parties, in the extreme to two parties, whereas there are no such reductive forces operating in the proportional system.7 Second, in contrast to the plurality system, the proportional system puts strong means for enhancing discipline, and thus for the creation of stable and cohesive parties, in the hands of the party leaderships. The main factor here is that the candidates for the legislature are largely dependent on the party leadership, both for nomination and for campaigning.

Now, by combining the methods for appointing the executive, and the electoral systems, we get four main types of democratic constitutions.8

1) Parliamentary constitutions with proportionalism. The main examples are in Western Europe. Usually there are some five to ten disciplined, stable and cohesive parties. Some of these parties may also be substantially bigger than required by a contingent threshold rule. Parliamentarism gives the incentives to discipline and, occasionally, to size, and proportionality the means. There are, however, no strong forces reducing the number of parties.

2) Parliamentary constitutions with plurality. This system is characteristic for the United Kingdom and some other countries in the Commonwealth. Due to the plurality system the number of parties are usually few. In spite of the incentives given by Parliamentarism the discipline is, however, lower than in constitutions with proportional elections, since the means are weaker.

3) Presidential constitutions with proportionalism. Several Latin American constitutions are of this type. Often there are many parties since the number-reducing forces operating in the elections for the legislature are weak. This is, however, often, to some extent, offset by a number-reducing effect of the presidential elections. The parliamentary incentives for discipline are absent, but there may be other incentives, and the means, given by proportionality, are there. Thus, in some countries the parties have a low coherence, while, in others they exhibit a considerable discipline.

4) Presidential constitutions with plurality. The main example here is the US. There we find two large parties with low internal discipline.

This fourfold classification is useful for my purpose, even if not exhaustive. In particular the constitutions which simultaneously have elements of presidentialism and parliamentarism, as for example the French constitution, are not represented - although, as we know, they are gaining popularity. Anyway, in the following, and due to lack of space, I will restrict the discussion to the first and fourth types of constitutions above. I will be comparatively detailed about the first type, and use the fourth type mainly as an illuminating contrast. Since these two types of constitutions can be looked upon as extremes in several respects this seems to be a defensible procedure.

3 The distribution of power

It is easy to see that the power distribution is very different in the two constitutional systems selected for discussion. In the parliamentary system with proportionalism the parties are so consolidated and disciplined that they can reasonably be considered as unitary actors. This does not mean, of course, that individuals are not important. It means, however, that the individuals almost exclusively play their roles within the parties. The individuals have a say in determining the party positions, and more so the higher they are in the party hierarchy. When it comes to dealings with actors outside the party, for example with other parties, or with the electorate in campaigns, or with lobbying organizations, it is usually the party as such, or the party leadership, which acts. It is in that sense that the party is a unitary actor.

In the parliamentary system with proportional elections the power is thus concentrated at the hierarchical tops of the political parties, albeit not evenly. In fact, almost all the power during the current election period is held by the constellation of parties belonging to the executive, or to the parliamentary majority supporting the executive. The power is thus very concentrated and the main actors, the important parties, are often fewer than five.

The US presidential system with plurality contrasts strikingly to this pattern. There, the party restrictions on the behavior of the president, and on the members of the Congress, are very weak indeed, and all these individual human beings can therefore be considered as fairly independent actors. In such a system there are thus hundreds of actors in the legislature and the executive. The power is diffused, not only between the president and the Congress, but also among all the members of the Congress.

These different patterns should reasonably be of great importance. It thus seems likely, to put it that way, that the transaction costs of political processes depend critically on the number of independent actors taking part.9 10 Considering the particular processes one may also say that the number of independent actors is likely to affect the possibilities to build decisive majorities or blocking minorities, the character of lobbying-processes, and the expediency of various strategies in the political competition.

As for lobbying the commonly held opinion, that it is more developed and more influential in the US than in other countries, is probably wrong. This is rather an impression created by the fact that the targets for lobbying are so many and so dispersed that the activities unavoidably become open and visible for everybody. Lobbying cannot, as in a parliamentary system, be hidden in a few closed rooms. For the same reason lobbying in the US is also less effective, and requires more resources, than lobbying in parliamentary countries.

4 Delegation and instruction

The character of the relation between the voters (the principals) and the political main actors (the agents), whether individuals or parties, are important and dependent on the constitution.11 For the discussion of these matters I will differentiate between two types of such relations, which I call delegation and instruction.

Delegation is the simpler of the two and most people have experience of it from everyday life. When people in typically voluntary associations like the local sports club or charity association elect presidents, secretaries, and so on, they usually do not require more than having confidence in the persons elected. They just want to be able to rely on them to act in a way that is in accordance with common sense and the purpose of the club. Feeling such confidence they delegate the decision-making to the people elected. Mostly such a system works well but if some functionary, for some reason, starts to act in ways of which the members disapprove, there are usually provisions in the club's charter for displacing the functionary. This rather simple kind of relation occurs not only in clubs, but also in politics.

Instruction, on the other hand, prevails when the voters do not limit themselves to a simple confidence in the ones elected but rather require that they execute a certain program, which may be worked out in a rather detailed way. Therefore, when people are elected, a program, or an instruction, for the elected to realize, is also adopted. The program may very well be, and often is, formulated by the people who want to get elected. Different candidates for political positions thus offer the voters to carry through different programs if they are elected. This, however, is fully consistent with the view that the program, once a candidate is elected, can be considered as an instruction from the voters to the elected.

It is easy to see that mixtures of delegation and instruction often appear in reality. Sometimes the element of delegation dominates, sometimes the element of instruction. One may therefore ask about the conditions favoring the one or the other type of relation. My hypothesis is that the parliamentary system with proportionalism has a tendency towards instruction, whereas the US presidential system has a tendency towards delegation.

The reason is simple. In the parliamentary system a campaigner, which in that case is a party, will be able to fulfill its promises if its electoral success is big enough. If, for example, a party alone gets more than fifty percent of the seats in the legislature, it can, by itself, form an executive and effectuate all its promises immediately. In fact, as we shall see later on, even a small party has a good chance of delivering on its promises if, after the election, it manages to join the executive.

A system of the US type is, in this respect, quite different. Imagine, for instance, a person running for the presidency, or for a seat in the Congress. In both cases everybody knows that the person, after the election, and however great the electoral success, will not, without further cumbersome and yet uncertain negotiations, be in a position to deliver on his or her campaign proposals. Exactly for that reason it would not be particularly clever, and perhaps even a bit ridiculous, to let detailed proposals dominate the campaign. It seems more expedient for the candidate to emphasize his or her own personal qualities, thereby indicating a capacity for prudent action in various future situations which, at the moment of the election, are impossible to foresee. That, on the whole, is also what candidates seem to do, and the resulting relation to the voters, hence, is primarily that of delegation.

5 General and specific instructions

Having thus argued that instructions are likely to play a relatively important role in a parliamentary, proportional setting I will now discuss the possible nature of those instructions. My main distinction is between general and specific instructions.12

General instructions may be based on ideological ideas about the ideal character or construction of society, or they may be derived from ideas about the common, or public, interest. The implementation of general instructions will thus usually affect the society at large.

Specific instructions, on the other hand, satisfy the interests of particular groups of people, that is special interests. Specific instructions are thus independent of notions about ideal societies, or about the common good - they exist just because people want more of the goods of this world, and can use majority politics for satisfying such wants. The implementation of specific instructions usually means that some people satisfy their interests at the expense of others.

I will now argue that, on the whole, it is easier for political parties to make deals (or logroll) about specific instructions than about general ones. If, for example, one party is committed to a particular specific instruction, and another party to another one, they can easily agree about supporting each other - if you support my instruction, I support yours. General instructions, on the contrary, are often in conflict with each other, and are therefore not easily reconciled - you cannot have socialism and capitalism at the same time. Sometimes, however, there may be deals in which a general instruction of one party is knit together with a specific instruction of another party. They may agree that the first party supports the second party's specific instruction, if the second party supports, or perhaps just tolerates, the first party's general instruction. It may also be argued that general instructions, which give a prominent and far-reaching role to the state, are easier to reconcile with specific instructions than those giving a limited role to the state. The reason, of course, is that specific instructions often are natural parts of state interventionism.

6 The competition for a place in the executive

We are now able to deal with the political competition in some detail. In a parliamentary, proportional setting there are two stages in that competition, first the competition for votes in a general election, and after that the positioning and bargaining in order get into the executive. The competitors are, as we have seen, the parties. Which strategies are most likely to lead to success in this context?

Focusing first on the second stage, the problem of getting into the executive, each of the parties knows that it is unlikely to get a majority of its own, and that it therefore must be able to make deals about a governmental program with other parties. This gives these hypotheses about the likely character of the party programs:

In addition to this we also get these hypotheses about the executive coalitions likely to emerge, and about their governmental porgrams:13  
7 The competition for votes

The competition for votes is more difficult to analyze than the competition for places in the executive, and we have to be satisfied with a few hypothetical preliminary steps. One reason for the difficulties is that the process, by which a voter decides which party to support, usually is complex. The issues are not presented one by one, as in a number of consecutive referendums, but rather as parts of complete party programs. In principle the voter thus has to weigh the pros and the cons, for each of the parties, in order to arrive at a final decision. Some voters may find that simple, but others feel uncertain almost all through.

In order to get a grasp of a mechanism, which may be important in parliamentary democracies with proportionalism, we can consider a party S of modest size which, in an election campaign, tries to attract a particular group of voters by offering them some advantages at the expense of other voters. This offer, or proposal, which we can call P, is thus an example of a specific instruction. Then some members of the target group, who appreciate the proposal, vote for S, which thereby becomes somewhat bigger than it would otherwise be. After the general election S will be considered a possible executive member and, in the negotiations preceding the formation of the executive, S promises to support important points in the other prospective member parties' programs in return for their support of P. The parties reach an agreement along these lines and form an executive. P thus becomes part of the executive's program and will therefore become implemented. Is this kind of scenario, we may ask, likely, or even typical, of a parliamentary democracy with proportionalism? Since the conditions therefore are fulfilled, my answer is in the affirmative.

The main condition is that the parties have the capacity to act in the way described. What is required, basically, is that the parties have reasonably clear voices when talking to the voters, that they can make firm deals with other parties, and finally that they can fulfill their own promises to the other parties by controlling their own people in the legislature. All of these requirements are better fulfilled in a parliamentary, proportional setting, with its unitary partisan actors, than in any other type of democratic system - or, in other words, the campaigners' capacity for credible commitment towards the voters is greater than in any other constitutional setting. The conclusion that the parties can act as described therefore seems reasonable.

This, however, does not settle the issue. It is obviously not sufficient for the parties to be able to act in the way described. They must also find it expedient to do so, it must pay in terms of votes. More exactly, the action must be expected to result in a net gain in votes - the number of voters attracted from other parties must be greater than the number of voters repelled.

In principle this is all practicable. The negative effects may for example be spread out so thinly, and over so many people, that those hit hardly notice. With some shrewd maneuvering it may even be possible to allocate the negative effects mainly on voters who would not have voted for the party anyway.14 Beside these problems about the management of the negative effects it is, however, also necessary to consider the reactions of those favored by the proposal. Are they really likely to feel attracted and thus to change their minds in favor of the proposing party? Some may perhaps do so immediately, but there may also be those in the target group who, although favored by the particular proposal, generally dislike the system of politically distributed goods and clientelism, and therefore want to change the system rather than to take part in it.15

Such voters, according to the terminology used here, favor some general instruction rather than the specific instruction at issue. But perhaps there is no party committed to the general instruction which these voters endorse, or if there is such a party its chances of becoming big enough for getting the instruction into a governmental program may be slim. Such dilemmas are, in fact, as we saw in the preceding section, quite likely. The voters may thus find it best to play safe and vote for the party offering the favors. Voting for the second best may, after all, seem more prudent since it may give a payoff even if the favored party, after the election, is still quite small.16

The conclusion thus is that strategies including specific instructions about favors to particular target groups may be quite profitable, and are more likely to be profitable in the parliamentary, proportional setting than in other constitutional contexts. Although other types of strategies will certainly also be used this is an important conclusion which, as we shall see in the next section, is relevant for the expansion of the public sector. Furthermore it paves the way for the following hypotheses about parliamentary, proportional systems:

In a presidential system with plurality, such as the US system, all of this is likely to be different. First, we remember, there is a tendency towards delegation rather than instruction. Candidates thus have an inclination to sell themselves, rather than programs. Furthermore, since strategies deliberately designed for attracting particular groups of voters do not work well the candidates must, rather, concentrate on not repelling voters.18 Since the electorate to a large extent is not committed from the beginning this means that they must be cautious not to repel anybody. There is, in fact, no opposition in the sense that there is in a parliamentary system. To this we can, however, also add that obligations of the state, once they exist, and whatever the reasons for their existence, are probably more difficult to remove in a presidential system of the US type than in a parliamentary system. The reason is the expediency, or even the necessity, of non-offending strategies in such a country.19

8 An asymmetry favoring the public sector

This paper basically argues that the democratic competition for power propels the expansion of the public sector. Campaigners, in order to win, promise their voters, if they win, a better life at the expense of the losers. One prerequisite for this, however, is that a majority rule is used. If the unanimity rule were used there would be no outside minority for the majority to exploit. But, as we know, there is no democracy in which the unanimity rule plays an important rule; all democracies rely mainly on various kinds of majority rules.

Another important prerequisite for the expanding public sector is that the exploitation, in fact, is done by taxation. This, however, is not necessarily the case. Those exploited must not necessarily pay by increased taxes. Another possibility is that they pay by having some already existing politically decided privileges reduced. Obviously we have to consider this second possibility since, in contrast to the first, it leaves the public sector unaffected.

There is however, I submit, an asymmetry which makes the second possibility much less likely than the first, namely that the task of taxing a large part of a population evenly is, administratively, much easier than the task of reducing marginally a lot of privileges of various groups, with the same final distributional effect.

It thus seems reasonable to expect expanding public sectors in all democracies. Still, the speed of expansion may differ depending on the type of democracy. A main reason for this, as I have argued, is that exploiting majorities are more easily formed in some democracies than in other others.

9 The empirical evidence

The general expansion of public sectors is a phenomenon which is well documented. Here, we are therefore only interested in data relevant for explaining the expansion mechanisms. The data of that kind that I know of is, however, utterly sketchy and inconclusive. To a large extent each reader has to use his or her own knowledge about things, and rely on the judgments that follow from that. There are, however, a few facts worth mentioning.

10 Conclusions

The discussion can now be concluded with the following two main hypotheses:

11 Implications

An almost unavoidable expansion of the public sector seems to imply that a democracy breaks down when the public sector gets impossibly large. If that is true, it is also true that the life of a democracy is limited to the time it takes for the public sector to grow from low to impossibly high levels. Are these conclusions really compelling, we may therefore ask, or are there ways to avoid the disastrous outcome? I will conclude by making three brief comments on this topic.

The first comment is related to the nature of the threat. Using Hirschman's (1970) well-known distinction between voice and exit democracy may be considered as rule by voice. Within that perspective the breakdown of a democracy is equivalent to an undermining of the voice option. Exit thus becomes the only remaining option for dissatisfied citizens, and a lot of different exit reactions are, I think, likely to occur. Same people may move to other countries, but many people may also exercise the exit option while staying where they are. They may, for example, turn from white to black markets, or from paying taxes to not doing so. If this happens the state, when it finally collapses, may be quite empty and void of authority, and it may have lost most of its tax-collecting power. Within this general framework many different specific outcomes seem possible. Some may be relatively fortunate, but it is also easy to imagine anarchical conditions with a lot of uncontrolled violence.

The second comment is related to the difficulties, in general, of undertaking preventive measures in controlled forms, in good time before the impending break-down. A main problem is that an expanding public sector automatically makes more and more people dependent on the sector, for example as receivers of transfers, or as public employees. Since these people are voters as well, and likely supporters of the public sector in that capacity, the problem of halting the growth, and contracting the sector, becomes more difficult the greater the sector is. These problems are dealt with in Eliasson (1986).

The third comment is related to the constitutional aspects of the challenge just mentioned. Above, in the section "the competition for votes," I made the point that although the public sector probably expands more slowly in normal times in a presidential democracy of the US type than in a parliamentary democracy with proportionalism, it may nevertheless be more difficult to reverse the expansion, that is to contract the public sector in an emergency situation, in the former constitutional setting. It may be easier to mobilize the power necessary for that purpose in the top-heavy parliamentary system. If this is correct it would really be a disaster to substitute a constitution of the US type for a parliamentary constitution with proportionalism in a country with a big and oppressive public sector. Rather, it would be necessary to reduce the public sector before the introduction of a new constitution. There are thus several different, difficult problems involved in the challenge, and the order in which their solutions is implemented is of critical importance. At last: Even if the expansion of the public sector is relatively slow in a presidential constitution with plurality, it is still important to design democratic constitutions which automatically contain public sector growth.


1Larkey, Stolp & Winer (1981) is a survey of theories of public sector growth.
2Schumpeter, J A (1976), Capitalism, socialism and democracy. New York: Harper & Row Publishers. In particular Downs (1957) was inspired by Schumpeter.
3To a large extent the paper is a condensed version of some main parts of a forthcoming book of mine provisionally titled Democracy: constitutions, politics and welfare effects.
4The importance of the electoral laws for the parties is widely discussed in political science but there is, to my knowledge, no corresponding interest in the effects of the rules for appointing the executive.
5One advantage for a big party, as we shall see later on, is that it may impose its ideology on the rest of the society. A big party often also has the advantage of being the component by which a coalition building-process starts. Even small parties may, however, have advantages by fitting well into minimum winning coalitions in Riker's (1962) sense. The incentives related to size are thus complicated.
6The idea that parliamentarism is dependent on stable, cohesive parties is generally accepted in political science. The opposite idea, that parliamentarism enhances stability and cohesiveness, is however, to my knowledge, not discussed in a systematic way at all, and when the topic occasionally arises for same reason, the idea is sometimes supported, sometimes discarded. An example of the latter is given by Sartori when he writes (1994, p 95) that "... party solidification and discipline (in parliamentary voting) has never been a feedback of parliamentary government."
7Maurice Duverger claimed (1964, p 217) that the tendency of a plurality system to enhance a two-party system came close to being "a true sociological law." This relationship, often referred to as "Duverger's law," is, however, not generally accepted in political science.
8This fourfold classification of constitutions is not totally absent in the political science literature. It is thus clearly indicated in for example Powell (1982) and Sartori (1994), and it is explicitly emphasized in Lijphart (1991). Neither of these authors do, however, stress the importance of the classification for parties and party systems.
9The transaction cost concept was, as we know, introduced by Ronald H Coase (1937) in economics, and by James M Buchanan and Gordon Tullock (1962) in constitutional analysis. In economics low transaction costs is generally considered as desirable, but in politics, where the majority rule usually reigns, it is not necessarily so. Low transaction costs may, for example, facilitate the formation of majorities exploiting the outsiders.
10When talking about transaction costs here I am thinking about the transactions between political actors, and thus about their outward activities in relation to each other. A disciplined political party, considered as a unitary actor, is however also characterized by a lot of inward activities. These activities are important since they, to a large extent, can thwart the intentions embedded in a constitution. A system of checks and balances can, for instance, be completely put out of function if the real decisions are taken within disciplined political parties which control the different branches of government. These matters have been analyzed by, among others, Donald Wittman (for example 1989, a and b).
11The nature of the relationship between voters and politicians has, it should be emphasized, been an important subject in the social sciences since long. Early and important contributors to the discussion were for example Edmund Burke and John Stuart Mill.
12This distinction is closely related to a distinction between generality and particularity made by Buchanan (1993a).
13The ideas expressed here are similar to, although not identical with, those in the so-called portfolio allocation approach. According to this approach, which has been elaborated by, among others, Laver & Shepsle (1994, 1996), the governmental program is determined by the allocation of ministries, or portfolios. Thus, if the head of the ministry of agriculture comes from party X, the agricultural policy of the government will, more or less, be the agricultural policy proposed in the party program of party X, and so on. In that way the total policy, or program, of the governmental executive will thus be composed of the relevant elements in the participating parties' programs
14The prerequisite that the campaigning politicians are able, to a sufficient extent, to recognize "their own people" is, as I will soon argue, likely to be valid. When that is so the campaigners can also afford, and perhaps even find it expedient, to offend other people. In Sweden there is even a particular term for politicians who choose to offend their opponents explicitly, namely politicians of confrontation. If that is taken to mean that not only opponent politicians, but also their presumptive voters, are railed at, then, I think, such politicians are hardly imaginable in a presidential setting with plurality.
15This requires, though, that the voters are endowed with a certain amount of idealism. If that is not so the target group, and in particular the individual voters in it, should rather, using Mancur Olson's concept, be considered as extremely narrow actors who do not care about the general effects of their behavior at all.
16An interesting implication of this is that the idea of rational ignorance has only limited application. The individual voter may be thought of as giving some additional weight to some instruction, which is already effectuated to some extent, and thereby, marginally, increase the advantages which are delivered according to that instruction. The returns on an individual vote are therefore greater than usually allowed for in the arguments supporting the idea of rational ignorance. This should lead to a higher turnout in general elections in parliamentary democracies with proportionalism than in other democratic systems, and seems to do so.
17This, it may be noted, is in agreement with Buchanan's contention that "Political players who might seek to further some conception of an all-encompassing general, or public, interest cannot survive" (1993a and b). Buchanan is, however, discussing democratic constitutions in general, whereas the focus here is on parliamentary constitutions with proportional elections, which, I submit, are extreme in the respect considered.
18Perhaps this explains the great importance of money and advertising in US election campaigns. There, policies do not speak for themselves to the extent that they do in systems encouraging political clientelism.
19In his book Demosclerosis (1994, p 124f) Jonathan Rauch, calling it an "asymmetry," says that "(t)o create a new subsidy or anticompetitive deal is hard, but to reduce a subsidy that already exists is much harder."
20In fact, this hypothesis indicates that the average expansion of the public sector per year of democratic rule may be an interesting concept. The Swedish public sector is, for instance, roughly twice as big as the one in the US, whereas democracy in the US is roughly twice as old as in Sweden. Therefore the average Swedish rate of expansion is roughly four times higher than the corresponding US rate.


Buchanan, J M & Tullock, G (1962), The calculus of consent - logical foundations of constitutional democracy. Ann Arbor: The University of Michigan Press.

Buchanan, J M (1993a),"How can constitutions be designed so that politicians who seek to serve 'public interest' can survive and prosper?" in Constitutional Political Economy, Vol 4 (1993), No 1.

Buchanan, J M (1993b), "Public choice after socialism," in Public Choice, 1993, pp 67-74.

Coase, R H (1937), "The nature of the firm," in Economica, Vol 4 (1937).

Downs, A (1957), An economic theory of democracy. New York: Harper Collins Publishers.

Duverger, M (1964), Political parties - their organisation and activity in the modern state. London: Methuen.

Eliasson, G (1986), "A note on the stability of economic organizational forms and the importance of human capital - a proposition about the endogenous, market-induced disintegration of the non-market sector," in Day, R H & Eliasson, G (eds), The dynamics of market economies. Amsterdam: North Holland, 1986.

Grilli, V, Masciandaro, D & Tabellini, G (1991), "Political and monetary institutions and public financial policies in the industrial countries," in Economic Policy, 1991 (October).

Hirschman, A O (1970), Exit, voice and loyalty -responses to decline in firms, organizations, and states. Cambridge, MA: Harvard University Press.

Larkey, P D, Stolp, C & Winer, M (1981), "Theorizing about the growth of government: a research assessment," in Journal of Public Policy, Vol 1 (1981).

Laver, M & Shepsle, K A (eds) (1994), Cabinet ministers and parliamentary government. Cambridge: Cambridge University Press.

Laver, M & Shepsle, K A (1996), Making and breaking governments - cabinets and legislatures in parliamentary democracies. Cambridge: Cambridge University Press.

Lijphart, A (1991), "Constitutional choices for new democracies," in Journal of Democracy, Vol 2 (1991), No 1.

Moberg, E (1992), Offentlig produktion av individuella och kollektiva nyttigheter. Stockholm: Näringslivets Ekonomifakta (mimeo).

Powell, G B (1982), Contemporary democracies - participation, stability and violence. Cambridge, MA: Harvard University Press.

Rauch, J (1994), Demosclerosis - the silent killer of American government. New York: Times Books.

Riker, W H (1962), The theory of political coalitions. New Haven: Yale University Press.

Sartori, G (1994), Comparative constitutional engineering - an inquiry into structures, incentives and outcomes. Basingstoke: Macmillan.

Schumpeter, J A (1976), Capitalism, socialism, and democracy. New York: Harper & Row Publishers.

Ståhl, I (1991), "Hur kan de svenska skattema sänkas? - svensk offentlig sektor i ett internationellt perspektiv," in Den nya svenska modellen. Stockholm: Svenska Arbetsgivareföreningen.

Wittman, D (1989a), "Why democracies produce efficient results," in Journal of Political Economy, Vol 97 (1989), pp 1395-1424.

Wittman, D (1989b), "The constitution as an optimal social contract - a transaction cost analysis of The Federalist Papers," in Grofman, B & Wittman, D (eds), The Federalist Papers and the new institutionalism. New York: Agathon Press.