This being so, one of my ambitions in the following is to show that we are, in fact, dealing with an attack. But I will also, of course, discuss the merits of the attack as such, as well as the merits of Lewin’s other arguments and conclusions.
The General Structure of the Book
Lewin’s book has the following five chapters:
1. Self-Interest and Public Interest
2. Does the Voter Follow His Pocket-Book?
3. Are Politicians Vote-Maximizers?
4. Are Bureaucrats Budget-Maximizers?
5. Individual and Collective Rationality
In the first chapter Lewin delimitates his subject-matter. He is, he says, interested in the roles of self-interest and public interest in Western politics. Self-interest, according to Lewin, is “to think only of the consequences for oneself”, whereas public interest is “to think also of others” (p 24). In order to make the study of the roles of self-interest and public interest manageable Lewin puts three questions (p 2) “structured”, as he says, “to three different levels of the political system”. The questions are:
1. Does the voter mainly follow the dictates of his pocket-book or some conception of the public interest?
2. Do politicians mainly strive to maximize their votes or to realize their electoral platform?
3. Do bureaucrats primarily try to maximize their budgets or to implement as loyal servants the political decisions of their masters?
Commenting on these questions it may first be noted that the actors in each of the three questions, that is the voters, the politicians and the bureaucrats respectively, together cover almost all important actors in democratic politics. Still, the interest organizations, and the people acting within them, are missing. This is remarkable in a book with the declared main ambition to assess the importance of self-interest and public interest in politics.
We also see that each of the three questions presents two possibilities, or alternatives, which at least seemingly exclude each other. Furthermore, and again in each question, the first possibility has a tendency towards some sort of selfish behavior, whereas the second alternative, rather, tends towards some kind of public or common good. The two alternatives thus correspond to the alternatives in the very title of the book.
In addition to this it is also obvious that the three questions correspond to the chapters 2, 3 and 4 as indicated above. This underlines the importance of these chapters. They contain, so to speak, the real substance, or the real investigation, in the book. The first chapter essentially paves the way for these chapters, and the fifth chapter comments, in a somewhat philosophical manner, on the results from them.
In the first chapter Lewin also tells us about the method he is going to use for dealing with the three questions. The important thing here is that he is not going to make any main investigations himself, starting from scratch as it were. Rather he will present the ways in which three different scholarly schools, or traditions, have dealt with the questions. These schools, using Lewin’s labels, are the rationalists, which in turn are divided into public-choice theorists and social-choice theorists, the philosophers, and the electoral researchers. They have all, according to Lewin, been interested in the questions about self-interest and public interest. However, and “since each field has adopted its own terminology and deals with these questions at different levels of abstraction, it cannot be said that there exists much dialogue between them” (p 1). This, of course, is to be deplored, and something giving Lewin a task. “To establish cross-fertilization is the aim of this book” he says on the very first page of his opus.
Later on in the first chapter Lewin presents the three schools in a more detailed way. Here, because of the attack on public choice, I will concentrate on Lewin’s presentation of that particular school or tradition. Among other things he mentions some works and authors, which, in particular, he associates with public choice (pp 4-6). These are An Economic Theory of Democracy and Inside Bureaucracy by Anthony Downs; The Calculus of Consent by James Buchanan and Gordon Tullock; The Politics of Bureaucracy by Gordon Tullock alone; and Bureaucracy and Representative Government by William Niskanen. (Complete references are given in the list of literature at the end of this comment.)
In addition to this Lewin also characterizes public choice. Most important, he emphasizes categorically, is that the public-choice theorists, without exceptions, assume that human beings are selfish. He thus writes (p 6) that, in the studies mentioned above, “Downs, Buchanan, Tullock, and Niskanen all held consistently to the fundamental premiss that the motives of the individual are selfish”. He also quotes (p 3) Dennis Mueller who, in his textbook on public choice, writes that “the basic behavioral postulate of public choice … is that man is an egoistic, rational, utility maximizer”. Then, immediately after this quotation, Lewin himself declares that “(w)henever the hypothesis of self-interest is made the object of scrutiny in this book, we thus take the public-choice school as our point of departure.”
Lewin thus explicitly associates the idea of self-interest, and thereby selfish voters, selfish politicians and selfish bureaucrats, with public choice. This means that the first alternative in each of Lewin’s three questions, the selfish one, becomes associated with public choice, and this, in turn, highlights the prominence of public choice in his undertaking. Following Lewin’s further treatment of the three questions and public choice we shall now continue to the middle chapters of the book – those devoted to the voters, the politicians and the bureaucrats respectively.
The Chapter on the Voters
The headline of this chapter is the question “Does the Voter Follow his Pocket-Book?”, which is a short version of the basic first question: “Does the voter mainly follow the dictates of his pocket-book or some conception of the public interest?”
For the understanding of this question we have to know the meaning of pocket-book voting. This is a kind of economic voting, but not the only kind. Basically Lewin makes a distinction between two kinds of economic voting. The first kind consists in voting against the incumbents when the voter, for one reason or another, has problems with his or her own economic situation, and supporting the incumbents when the voter’s own economic situation, on the contrary, is good and promising. The second kind of economic voting consists in voting against the incumbents when the economy of the country as a whole, for one reason or another, is not developing as it should, and supporting the incumbents when the country, on the contrary, is thriving and prospering economically. It is the first kind of voting which Lewin calls pocket-book voting. He also says that that kind of voting is motivated by self-interest, whereas the second kind of voting is motivated by a concern for the public good. I am not going to argue about any of these two judgements.
Now, it may be held that the voters mainly exhibit the first kind of economic voting described, or that they mainly exhibit the second kind. Lewin sometimes calls the first idea the self-interest hypothesis, sometimes the pocket-book hypothesis. The second idea he sometimes calls the public-interest hypothesis, and sometimes, using professional jargon, the sociotropic hypothesis. Thus we have:
- The self-interest hypothesis: The voter votes for or against the incumbents according to the development of his or her own economic situation.
- The public interest hypothesis: The voter votes for or against the incumbents according to the development of the country’s economy.
The two hypotheses obviously correspond quite well to the two alternatives in Lewin’s basic question about the voters. The problem is to decide between them, and that is not altogether easy. A difficulty is that data from aggregate analysis are likely to be insufficient. If, for instance, the incumbents are defeated in an election in economically bad times, it is not possible to tell if the reason is that particularly the suffering people vote according to their pocket-books, or that people in general are concerned about the nation’s economy. At an aggregate level the result could be the same. Disentangling these matters requires data from survey investigations about individuals and their motives. Or, in Lewin’s words (p 35): “What does an observed correlation between a decline in income per capita and a set-back for the incumbent party actually tell us? Is it those who are afflicted who react? Or is the party also abandoned by those who, without having suffered any personal loss, reject the government’s economic policy out of consideration for what is best for the society as a whole? That is the vital point in this debate. / If it is the afflicted, we should be entitled to regard the vote as an expression of self-interest. If it is not, the vote should be seen as a sign of public interest.”
In order to find out about these matters Lewin quotes and comments on a number of empirical investigations, mostly of the survey type, about voters’ behavior, in the US as well as in Western Europe. This abundant evidence, according to Lewin, shows that the pocket-book hypothesis is false. This position may be reasonable, but only if restricted exactly to the kind of pocket-book voting defined above. People may however, perhaps, follow the dictates of their pocket-books in other ways than the one described by Lewin. If so the information assembled by Lewin is without relevance. Still he concludes:
- That people, generally speaking, do not vote according to their pocket-books.
From this, in turn, he derives two further conclusions, namely:
- That the public interest hypothesis is proved correct, and
- That the public-choice school, insofar as its ideas about voters are concerned, is proved wrong.
I think that all of these conclusions are wrong.
The basic problem here is that the distinction between two kinds of voting, as represented by the two hypotheses above, is far from exhaustive. For seeing this it is enough to realize first, that self-interest may take other forms than pocket-book voting in the sense defined, and second, that interest may concern other things than oneself or the society at large, for instance the class, or the professional group, to which the voter feels attached.
For the discussion of these possibilities it should at first be noted that both pocket-book voting and sociotropic voting, as defined above, always involve change. Voting, according to the definitions, will change as a result of changed private finances in the one case, and changed public finances in the other case. This is also underlined by Lewin, for instance when he writes (p 40, my italics) about “… the usual questions about changes in the private economic situation and party preferences of the persons interviewed, …”. Now, let us consider, as a contrast to this, a voter who is a worker, and constantly, independently of his own current economic situation, and also independently of who currently governs, votes for a labor party. What kind of voting is this?
In a few places Lewin himself actually addresses this question, although quite superficially. In a context where he is discussing a methodological issue he writes, for instance (p 21, my italics), that “(a)ccording to class voting workers vote for a socialist party and the middle class for a non-socialist party. The blue-collar and white-collar worker may each have good reasons for their different choices, reasons that have nothing to do with group interests. It may be the considered opinion of the blue-collar worker that the welfare policies of the Social Democrats are best for the country as a whole, regardless of the fact that he himself will enjoy some of the benefits; the white-collar worker may similarly consider that the public interest is best served by deregulation and more elbow-room for market forces even if he does not personally stand to gain by the change.”
Yes, it may be like that, and that is also the important and perfectly legitimate point for Lewin’s in his methodological discussion. If, however, we confine ourselves to just considering the factual aspects of the example, it is equally obvious that it may be the other way round. It may be the considered opinion of the blue-collar worker that he should vote for the Social Democrats since that is dictated by his proper self-interest and by the interests of his class. Correspondingly the white-collar worker’s vote may be dictated by self-interest and class-interest. To this it may just be added that Lewin himself, in another place (p 53), describes “the voters’ occupational group” as “a strong determinant of party preference”.
We thus see that voters to a considerable extent vote according to their occupational and class membership, and that this may be interpreted as self-interested and/or class-interested voting. Things being like this it is obviously impossible to conclude, as Lewin does, that voters do not follow their pocket-books, just because they do not do it in the very particular sense defined by him.
As for the second conclusion, that the voters, in fact, are public-minded, it should be clear by now that it is not valid. The reason, of course, is that the voters, as we have just seen, may be self-interested in other ways than the one excluded by Lewin. The conclusion is however weak in another sense as well. The point is that the conclusion is just a conclusion, not a result established by direct empirical evidence. The thrust of the empirical data quoted by Lewin is against the pocket-book hypothesis (in a particular sense as we have seen), not for the public interest hypothesis. This, in fact, is in accordance with Lewin’s own recommendation in the quotation above: If it is not the afflicted who abandon the incumbent party, he says, then (my italics) “the vote should be seen as a sign of public interest”. This is hardly convincing.
Lewin’s third conclusion, the one about the shortcomings of the public-choice school, requires for being valid that the public-choice school is convincingly attached not only to self-interest in general, but also to the particular kind of pocket-book voting defined above. That that cannot be done is easily shown by quotations from the public-choice literature. One example about an altruistic voter, from Downs’ An Economic Theory of Democracy, is for instance given in the section “More about public choice” below. The conclusion is thus not tenable.
Regardless of this it is interesting to highlight Lewin’s way of presenting his conclusion about the public-choice school. In the very last and concluding sentence of the chapter on voting he drives home the point as follows (p 59): “The extensive empirical material that has been produced through modern research into economic voting makes it impossible to uphold the first assumption of the public-choice school that the voters are primarily guided by self-interest.” This is telling. In a chapter which, according to its title, is about whether voters follow their pocket-books or not, the final, underlined conclusion is that an assumption, which has suddenly gained the importance of “the first assumption of the public-choice school”, is wrong. This, I think, lends support to my contention that public-choice is a main target for Lewin.
Before finishing this section I would like to take up one more point about public choice. Following his ambition to show that voters are public-minded, and public-choice theorists wrong, Lewin quotes (p 39) a “philosopher” maintaining “that if one were to believe the view held by Downs, Buchanan, and other public-choice theorists concerning the voters’ egoism, the fact that the voters participated at all appeared completely inexplicable and irrational. There was no reason to vote, for the individual could not exert any appreciable impact on the outcome. The influence or exchange gained was not proportional to the effort made.” The conclusion is that voters are public-minded, since otherwise it is impossible to explain why they vote at all. The paper, from which Lewin has taken this argument, was published in 1977.
Now, it is interesting to note that Downs, in An Economic Theory of Democracy, which was published twenty years earlier, discusses exactly these matters in a rather elaborate way in the section “Participation in elections when voting is costly” (Downs, 1957, p 265 ff). I will not go into the details of this here. I just want to make clear that the public-choice theorists were well aware of the problems indicated long before Lewin’s “philosopher”, and that they discussed them openly and explicitly. Downs himself saw clearly that voting costs represented a challenge to his theory and that they therefore required, for their integration into the theory, a partial revision of the theory’s basic assumptions about rational voting. Downs exact way of making this integration, although very reasonable, may of course be discussed, but that is not the point here. The point, rather, is that Lewin’s attack on public choice for being inconsistent, with a philosopher as a striking-weapon, besides probably being erroneous, also has other deficiencies. It fails to give the public-choice theorists themselves credit for the so called “paradox of voting”, and it also severely underrates the richness and complexity of Downs’ theory.
Summing up this section on voters I will say that Lewin has succeeded in showing, by accounts of other scholars’ investigations, that voters probably do not, at least not to a large extent, vote according to their pocket-books in Lewin’s particular sense. Voters may however follow their pocket-books in other ways, and Lewin has not shown that they do not do so. Still less has he shown that the voters are public-minded. Finally, as we have seen, Lewin’s emphasized announcement about the fall of “the first assumption of the public-choice school” lacks substance.
The Chapter on the Politicians
The chapter on the politicians opens in the same vein as the preceding one on the voters ends. In the very first sentence Lewin states (p 60) that “(a)ccording to the second assumption of the public-choice school, politicians, just like voters, are primarily guided by their self-interest.” The public-choice school is thus again, and typically, focused. In this case Lewin intends to demolish what he has elevated to the school’s second assumption.
Now, in the same way as for the voters, the self-interest of the politicians has to be interpreted in some way, in order to be empirically manageable. In a first step, where he is in good company with Downs, Lewin says that self-interested politicians are vote-maximizing. When the properties of being self-interested, and being vote-maximizing, are linked in that way, the question is whether the politicians really are vote-maximizing or if they act in some other way. This is the question asked in the very title of the chapter: “Are Politicians Vote-Maximizers?” In a somewhat more elaborated form this basic question, as we have seen, becomes: “Do politicians mainly strive to maximize their votes or to realize their electoral platform?”
This last formulation also includes what effectively is Lewin’s idea about public-minded behavior. At the very end of the chapter on the politicians Lewin confirms this point by informing us (p 74) that the “opposite” of “the self-interest hypothesis” is “programme implementation”. Thus, in analogy with the hypotheses for the voters, we get:
- The self-interest hypothesis: Politicians are vote-maximizing.
- The public interest hypothesis: Politicians strive to realize their electoral platform, or, in another formulation, to implement their program.
In his discussion about the voters Lewin, as we saw, to a large extent used interview data directly from the voters about their motives. Here it is not so. There are, it seems, no investigations in which politicians directly are asked whether they are vote-maximizing or not, and perhaps for good reasons. Lewin therefore has to rely on some indicator of politicians’ motives, and he uses, in fact, two such indicators. The one is related to so called political business cycles, the other to the electoral strategies of socialist parties in Europe.
Lewin introduces the first of these indicators, that is the political business cycles, in this way (p 62):
If it were true, as Downs maintained, that politicians are first and foremost vote-maximizers, then we should expect the government to try and manipulate the economy in such a way as to make it as attractive as possible at election time and thus make the prospects of the government’s being reelected as favorable as possible. In other words, we should be able to count on the government’s stimulating the economy so that the favorable effects would be most noticeable in an election year whereas cuts and other measures with unfavorable consequences would not be taken until after election-day. The electoral tactics of the parties could thus be expected to give rise to fluctuations in economic activity. Instead of acting to mitigate the swing between recession and over-heating, which according to economic theory is an important task for politicians, it was quite likely that the politicians actually created exogenous movements based on their own interests – the political business cycle.
This reasoning, which Lewin calls the political business cycle theory, obviously has the form of a deduction. Stripped to its essentials this deduction may be described as leading from the assumption that ‘politicians maximize votes’ to the conclusion that ‘political business cycles will occur’. Now, this deduction, as any deduction, may be valid or not, and the consequences of that validity or non-validity are of great importance in this context. If the deduction is valid, then a falsification of the assumption that ‘politicians maximize votes’ is within reach. What is required, in addition, is an empirical demonstration that business cycles do not exist. The whole argument, in this case, follows the familiar pattern: If A implies B, then not-B implies not-A.
If, however, the deduction not is valid, matters are, on the contrary, fundamentally unclear. In that case we can only say that vote maximization is one thing, and political business cycles another, and that we do not know, so far, anything about the relations between the two.
The important implication of this is that, if one wishes to prove that politicians do not maximize votes, as Lewin does, one has to demonstrate two things quite clearly.
- First, that the deduction in the political business cycle theory really is valid – this is essentially a logical undertaking.
- Second, that political business cycles do not exist – this is essentially an empirical undertaking.
Starting with this last matter Lewin really makes considerable efforts to show that political business cycles do not exist. Again, as in the discussion about pocket-book voting, he presents a number of empirical investigations by various scholars, which, according to Lewin, show that there are no political business cycles. This case seems, on the whole, strong, and I will not dispute Lewin’s results here.
The problem, rather, is Lewin’s handling of the first matter, the validity of the deduction in the political business cycle theory. Lewin does not in fact realize, it seems, that his ambition to prove that politicians do not maximize votes requires that the deduction is made credible rather than the opposite. He argues all the time as if every possible flaw in the political business cycle theory, including its deduction, is to his own advantage. Consequently he quotes, approvingly, arguments which undermine the deduction.
When reporting on an investigation about political business cycles Lewin quotes and writes, for instance, as follows (p 68):
Despite popular belief to the contrary, we could find no conclusive evidence of a political business cycle in Germany or in the United States.’ There were plenty of reasons why governments should not give in to such temptations. The voters would not like it, for they dislike the budget deficits that are the inevitable result; if all parties acted in this way, the voters would look for new parties; in coalition governments the largest party would receive the severest punishment; no government has anything to win from an economy in chaos.
Similarly, when presenting one more study, Lewin writes that (p 69) “(a)nother scholar has analysed the economic policy of the United States, Great Britain, and Germany. His conclusion is that the government’s policy is more affected by ‘shocks’ that hit the economy. Furthermore, there are so many errors in attempts to achieve popularity with the help of changes in unemployment and inflation that the whole notion of trying to plan these changes rationally in order to maximize votes appears ridiculous.”
I agree with Lewin in his approval of these arguments. But the implication, of course, is that the deduction in the political business cycle theory is non-valid. This, in turn, means that Lewin’s whole long discussion about vote-maximizing and political business cycles boils down to nothing. In the end Lewin has shown, by means of others’ empirical investigations, that political business cycles probably do not exist, but he has also made it clear that that is exactly what is to be expected – with or without vote-maximization.
Lewin himself does not, however, seem to realize that this is the result of his discussion. He remains all through in the belief that all arguments against the political business cycle theory, including those against its deduction, support the conclusion that politicians do not maximize votes. In the concluding part of the chapter, a few pages later, he consequently writes (p 73f) that the basis for rejecting the self-interest hypothesis for politicians “is first and foremost the extensive research that might be gathered under the heading of the ‘political business cycle’.”
This, as we have now seen, does not hold. The extensive research on political business cycles, which Lewin is talking about, does not show that politicians do not maximize votes. What it shows is that manipulation of the economy, to the extent that discernible political business cycles appear, probably is a very bad and counterproductive way of maximizing votes – and therefore it is also unlikely to be used. Politicians may however maximize votes in other ways.
Having this in mind I will turn to Lewin’s second indicator of vote-maximization, the one related to the socialist parties in Europe. The idea is that the leaderships of these parties could get more votes, than in fact they have, if they tried to gain support from the middle class as well. Certainly such a strategy would mean the loss of some workers’ votes, but the gain in middle class votes would be greater. Thus, if politicians are vote-maximizing the socialist parties in Europe would make (p 72) “an optimum trade-off between the largest possible gains from the middle class and the least possible losses form the workers” and thereby grow further. Consequently, if they do not take advantage of that opportunity they are not vote-maximizing.
The basic idea here is that the socialist parties we are talking about have a potential for getting considerably more votes than, in reality, they have. In principle these extra votes are easy to get. What is required is just that they give up some of their socialist ideas and attributes in favor of pure vote-maximizing. But, one may ask, is this potential for more votes really there? Lewin says it is, but his arguments are meager.
He quotes an investigation according to which the percentage of workers, among those voting for the Swedish Social Democratic Party in the year of the investigation, was 78 %, whereas it would have been 21 % if the party had been vote-maximizing. He gives corresponding figures for Denmark and Norway and then states (p 73), without any further arguments or evidence, that “(c)oncern for other values (class, ideology, relations with the trade union movement) has outweighed the desire for votes.” This is hardly convincing. Many questions remain unanswered, as for instance: How are the very exact figures, which are quoted, arrived at? Which are the corresponding exact values, which are not quoted and which seem considerably more important, for the election results for the parties concerned, with and without vote-maximization? How is the trade union movement’s contribution to an election campaign evaluated? What is the power of an ideology as a magnet for votes? It is, I think, by no means evident that the easy way to more votes, which Lewin points to, is there.
Lewin then finishes his discussion about the socialist parties in Europe in a remarkable way. He writes (p 73) that the authors, whose investigations he uses, “formed their conclusions into a powerful polemic against Downs and the vote-maximization theorem. Parties cannot jump freely from one position to another in order to adopt the policy they believe the majority of voters accepts at any one time. To disregard such limitations on the parties’ room to manaeuvre is to commit the error of reducing the analysis of political science to ’empty formalism’.”
This deserves two comments. The first one is that the authors quoted, and with them Lewin, seem to argue against themselves. The parties unable to jump from one position to another must, reasonably, be the socialist parties contemplating to change policies in order to reach the middle class voters – otherwise the passage would be meaningless or without reference. But if that understanding is correct the authors, and Lewin, are in fact saying that votes cannot be maximized in the way they earlier said it could – the potential for more votes just is not there. Thus, in a similar way as in his discussion about the political business cycle, Lewin is again undermining his own argument. Furthermore, and this is more important, the implication of this undermining is that the politicians in the example in fact, and contrary to Lewin’s intentions, do maximize votes. Given all the conditions, and restrictions, inherent in the situation, they seem to do a good job in getting as many votes as possible.
The second comment is about the attack on Downs. As it happens Downs has, in fact, in An Economic Theory of Democracy, a whole section entitled “Ideological Coherence and Stability”, dealing exactly with the matters at issue here (Downs, 1957, p 109 ff). There he writes, among other tings, that (his italics) “(i)deological immobility is characteristic of every responsible party, because it cannot repudiate its past actions unless some radical change in conditions justify this”. Apparently thinking about vote-maximization he also calls the immobility “rational”. He moreover writes that “the party will not radically shift its policies and doctrines overnight, but will only slowly change their nature.” To accuse Downs of saying that parties freely can jump from one position to another, and thereby of empty formalism, is obviously wrong.
Summarizing we may now say that Lewin, in this chapter, essentially has made one thing probable, namely that there are no political business cycles. This, as we have seen, fails to prove that politicians do not maximize votes, which was Lewin’s one main ambition. It also fails to prove that politicians are not primarily guided by self-interest. Lewin thus fails to demolish his conception of the second assumption of the public-choice school, which was his other main ambition. Finally, and obviously, he does not prove the public interest hypothesis either, namely that politicians strive to realize their electoral platforms, or to implement their programs.
Having said this I will just add that the two goals, vote-maximizing and program implementation, may be interrelated in various ways. It may, for instance, be necessary to have at least some votes for being able to realize any program at all. It may also be easier for a party to realize its program the more votes it has. These matters are not discussed by Lewin at all.
The Chapter on the Bureaucrats
This chapter differs a bit from the two preceding ones, but there are also similarities. Starting with the latter the chapter is given a title containing the typical question, in this case: “Are Bureaucrats Budget-Maximizers?” And, as in the former chapters, this question is the short version of a basic question namely: “Do bureaucrats primarily try to maximize their budgets or to implement as loyal servants the political decisions of their masters?” In the same way as in the former chapters we thus get two hypotheses:
- The self-interest hypothesis: Bureaucrats are budget maximizers.
- The public interest hypothesis: Bureaucrats implement loyally the decisions of their masters.
Here the self-interest hypothesis becomes (p 75) “the third assumption of the public-choice school”, and again Lewin sets out to demolish it.
The differences between this chapter and the preceding ones are that the discussion about the public-choice literature is more important here, and that accounts of empirical investigations are less important. The reason is two-fold. The public-choice literature on bureaucracy is comparatively rich, and the empirical investigations are few. “The budget-maximization hypothesis has not”, as Lewin says (p 78), “been made the object of tests as coherent and systematic as those applied to the hypotheses concerning pocket-book voting and the political business cycle.”
The Public Choice Literature on Bureaucrats
In the first chapter of his book, as I have already noted, Lewin mentions five works in particular as main results from the public-choice school. Out of these five works no less than three are about bureaucracy. These are Tullock’s The Politics of Bureaucracy, Downs’ Inside Bureaucracy, and Niskanen’s Bureaucracy and Representative Government. Now, in the chapter on bureaucracy, Lewin says (p 75) that Niskanen’s “study Bureaucracy and Representative Government has come to be the most renowned and will be taken here to represent the position of the public-choice school regarding the third level – bureaucracy – in our analysis of self-interest and public interest in Western politics.” Since all the three books mentioned belong to the public-choice school, and thereby are relevant for its position, Lewin must be interpreted as holding that the books are quite similar. Otherwise the three could hardly be represented by one.
In order to be able to deal with this particular contention of Lewin, and with other aspects of his account of the books, I will start by characterizing the three books very summarily. They are, in fact, quite different. They are also rich in content and ideas, each one in its own way. But even so the two first books, that is Tullock’s book from 1965 and Downs’ book from 1967, are in many respects similar – Downs to a large extent continues Tullocks’ work, and builds on it – whereas Niskanen’s book from 1971 runs on another track.
The theories of Tullock and Downs deal foremost, and in considerable detail, with the inner workings of bureaucracies – in the case of Downs this is apparent from the very title of the book. The hierarchical structures, which constitute bureaucracies, are carefully described, and the interaction between these structures and the individuals working in them, with their specified incentives, is focused upon. The determinants of the efficiency of different types bureaucracies, in specified senses, is one, but not the only, object of interest. Within this framework the study of flows of information within bureaucracies, and distortion of information, becomes particularly important. Niskanen’s subject matter is quite different. He is primarily interested in the gross determinants of the physical output of a bureau, which, in his theory, is determined as an equilibrium between the incentives of the bureaucrats within the bureau, and the incentives of the politicians funding the bureau. Both groups, it should be added, are basically treated as single actors, in order to keep the theory as simple as possible. Thus, in Niskanen’s theory a bureaucracy’s internal structure is quite rudimentary. Or, in Niskanen’s own words (Niskanen, 1971, p 11): “The theory developed here does not address the management processes internal to bureaus … “.
Things being like this it is remarkable to select, as Lewin does, one of the books as representing all three, and choosing, in particular, Niskanen’s book as the representative one is still more remarkable. So let us see, by looking at a few specific subjects, which are especially relevant in this context, how Lewin carries out his presentation of public choice.
The Incentives of Bureaucrats
Lewin entertains no doubts about his enemy’s position. According to “the third assumption of the public-choice school” it is, he says (p 75), the bureaucrats “self-interest that best accounts for their actions.” To a large extent this is true, but only to a large extent – some shades are missing.
Let us start with Tullock, who in an interesting passage (Tullock, 1965, p 21) writes about two hypothetical, extreme bureaucrats. The one “is interested solely in his own career” and the other “only in the objectives of the organization”. In an ideal organization these two bureaucrats, says Tullock, would act in the same way in equivalent situations. In a real organization the career-minded bureaucrat is however likely to have at least some advantages over the idealistic one. And therefore “it appears probable that the higher ranks of most governmental bureaus are made up of people who are less interested in the ostensible objectives of the organization than in their own personal well-being.”
Here Tullock considers differently motivated bureaucrats, he discusses their interaction with the organization, and he concludes that the more selfish are likely to advance higher in the organization than the more idealistic ones. This may be right or wrong, that is not the point here. The point, rather, is that the bureaucrats are more diversified and complicated than in Lewin’s description of public choice, and that the differences are paid attention to in the theory.
Downs goes further along the same road. First he writes (Downs, 1967, p 2) that “(b)ureaucratic officials in general have a complex set of goals including power, income, prestige, security, convenience, loyalty (to an idea, an institution, or the nation), pride in excellent work, and desire to serve the public interest”. Then, trying to take all of this into account in his theory, Downs distinguishes explicitely, and throughout his book, between five types of bureaucrats. These are the climbers, the conservers, the zealots, the advocates and the statesmen (Downs, 1967, for instance p 88). The first two types are motivated almost exclusively, though in different ways, by self interest. The last three types are motivated by self-interest and altruism in varying types of combinations.
In spite of the fact that Downs uses five different types of bureaucrats in his theory, in an utterly clear and explicit way, Lewin does not say a word about that anywhere. Rather, Lewin’s only, and entire, account for Downs’ dealing with the incentives of the bureaucrats is this one (p 6): “In Inside Bureaucracy Downs … contended that bureaucrats tried to improve their own welfare in terms of salary, esteem, and influence by seeing to it that their office became as large as possible. Aggressive budget demands and expansive reorganization were part of this deliberate endeavor. Even in the exercise of official duties the bureaucrat acted more from considerations of personal advantage than from loyalty to the intentions of the legislature.” It is hard to believe that the author of these lines has read the book he is writing about.
Lewin also positively criticizes the public-choice theorists for not taking into account particular incentives which he considers important. Thus, he says (p 88) that they “seem to have forgotten the traditional critique of bureaucracy, not for being ambitious and energetic, but rather lethargic and without initiative.” One comment to this is that the kind of idleness, which Lewin describes, seems to be more typical for a self-interested bureaucrat than for one loyally serving his master. Another comment is that Downs’ conservers, which “consider convenience and security as nearly all important” (Downs, 1967, p 88), are close to exhibiting the kind of behavior Lewin has in mind. It is difficult to see that something is forgotten here.
Lewin also criticizes public-choice for not realizing that bureaucrats may be “mission-committed” (p 88) and therefore “determined to put the intentions of the politicians into practice in an efficient manner”. But Downs’ zealots, advocates and statesmen are endowed, in varying forms, with exactly this kind of incentives. Thus, and again, Lewin is mistaken.
In this section I have limited myself to comments related to Tullock’s and Downs’ books. Lewin’s account for the bureaucratic incentives in Niskanen’s book will be discussed in the next section.
The Budget Maximization Goal
Lewin is right in saying that budget maximization is the over all goal of Niskanen’s bureaucrats, but he fails in giving a proper understanding of the nature of that goal in Niskanen’s theory. The important thing here is Niskanen’s explicit declaration that the goal is a proxy.
Niskanen starts by considering several possible goals for individual bureaucrats. “Among the several variables”, he writes (Niskanen, 1971, p 38), “that may enter the bureaucrat’s utility function are the following: salary, perquisites of the office, public reputation, power, patronage, output of the bureau, ease of making changes, and ease of managing the bureau. All of these variables except the last two, I contend, are a positive monotonic function of the total budget of the bureau during the bureaucrat’s tenure in office.” Then, a few lines later, he continues by writing (my italics) that “(i)t is not necessary that a bureaucrat’s utility be strongly dependent on every one of the variables which increase with the budget, but only that it is positively and continuously associated with the level of the budget. For these reasons, budget maximization should be an adequate proxy even for those bureaucrats with a relatively low pecuniary motivation and a relatively high motivation for making changes in the public interest.” Again, and for my present argument, the point is not whether this is reasonable or not. The point is that Niskanen’s reasoning is far more subtle and comprehensive than Lewin tries to make us believe.
I have already mentioned Lewin’s critique of public choice for not taking incentives, which Lewin considers important, into account. Here is another example. He says (p 85) that the “prospect for promotion ranks very high on the bureaucrat’s scale of values”. Then, referring to his repeated (p 86) claim that “(b)ureaucrats wish to advance”, he tells us that doing “a good job is a more obvious strategy than to attempt something as speculative as maximizing the budget.” This is certainly true in many cases. It may just be added, first, that this whole idea is perfectly compatible with Niskanen’s notion of budget maximization as a proxy, and, second, that the importance of the promotion incentive, and the consequences of that importance, is a very visible main theme in Tullock’s book in particular, but also in Downs’.
Lewin’s allegation that (p 77), according Niskanen, “(b)udget-maximization is an instinct in every forward looking bureaucrat” is thus, at best, a gross caricature. Now, since Lewin has taken Niskanen as a representative for Tullock and Downs, we should take a look into the books of these scholars as well. The fact is that they do not say anything at all about budget maximization. Given the concept’s character of a proxy in Niskanen’s theory this should hardly come as a surprise.
The Growth of Bureaucracies
Lewin shows a lot of interest in public sector expansion. A first thing to note about that is that Lewin sometimes talks about public sector growth, sometimes about growing bureaucracy, as if they were the same thing. The difference is however great, in particular in a modern welfare state with huge transfers. Having said that I will however leave that topic. What is important here are rather the reasons for, and the consequences of, Lewin’s interest in the expansion.
The problem for Lewin is that the “the rapid expansion of the public sector in the last few decades”, as he himself puts it (p 75), may lend support to ideas about budget-maximization and the like. Lewin therefore explores the possibility that the expansion has other causes than self-interested bureaucrats. “To accept the self-interest hypothesis on the evidence of such administrative expansion”, he says (p 75), “is however to buy a pig in a poke. Before jumping to conclusions one should look to see if there are not other explanations of the growth of the public sector.”
The public choice theorists, according to Lewin, explains the expansion by means of two mechanisms. The first one, which we may consider the main one, is the self-interested budget maximizing behavior of the bureaucrats. The other one, which we may rather consider auxiliary, consists in the politicians’ passivity towards the bureaucrats.
A few quotations show that Lewin really believes that the public choice-theorists argue like this. He says, for instance (p 75), that “the hypothesis that the growth of the bureaucracy is best explained by the self-interest of the bureaucrats … has been formulated by such leading members of the public-choice school as Tullock, Downs, and Niskanen.” About Downs, in particular, he says (p 6) that he, in Inside Bureaucracy, “puts forward an elaborated deductive theory of how this bureaucratic growth could be explained with the hypothesis of self-interest.”
Going then to the second auxiliary mechanism Lewin writes (p 76) that Niskanen “considered politicians to be completely under the control of the bureaucrats”, and he also talks (p 79) about “(t)he passive role assigned to politicians by Niskanen”. Lewin furthermore quotes with sympathy some authors arguing (p 81) “that the most serious shortcoming of Niskanen’s model is that it blames budget-maximizing bureaucrats for the expansion of the public sector and forgets about the politicians who decide on requisite financing”, and another author declaring (p 82) that the first fundamental error in Niskanen’s model “is to believe that bureaucracies are allowed to grow in just any fashion without the politicians being able to do anything about it”.
Now, if this really was true, Lewin would have reasons for his concern. He also would have reasons for his efforts to demonstrate, in a rather lengthy discussion, that the mechanisms imagined are non-existent, and that the expansion can be explained by other means. All of this we can however neglect for the very simple reason that Lewin, from the very beginning, has misunderstood the public-choice theorists’ undertaking.
A first point is that the public-choice theorists’ interest in the expansion of public sectors and bureaucracies is not as dominating as Lewin will make us believe. His allegation that Downs “puts forward an elaborated deductive theory of how this bureaucratic growth could be explained with the hypothesis of self-interest” is, in its context, at best utterly misleading. The formulation gives the impression that explanation of bureaucratic growth is a major interest, or even the major interest, of Downs. This is certainly not the case. Still the interest is there, as an interest among other ones, and in particular, I would say, in Tullock’s book. But even so, the explanation differs from the one suggested by Lewin (Tullock, 1965, the chapter “Parkinson’s law”, p 134; the section “The Growth of Bureaucracy”, p 174). This being so we should not be surprised to discover that the very mechanisms described by Lewin are difficult to find in the public-choice theorists’ works. Thus, as we have already noticed, the self-interest motive hardly appears in the pure and dominating form suggested by Lewin. Furthermore, and this is more important, Tullock, Downs and Niskanen certainly do not say that politicians are completely unable to control bureaucrats. I will substantiate this last point with illustrations from Niskanen and Downs.
In Niskanen’s theory, as I have already mentioned, the total physical output of a bureau is determined in an equilibrium process, as a result of the activities of the bureaucrats in the bureau and the bureau’s sponsors, which means politicians. The sponsors are characterized by a “budget-output function”, which “represents the maximum budget the sponsor is willing to grant the bureau for a specific level of output” (Niskanen, 1971, p 25). The higher the level of output is, the more they are prepared to pay. Furthermore, for low levels this sum is greater than what is needed for the production, whereas, for high levels, it is not sufficient for the production. Consequently, there is a level in between, for which the sponsors’ accepted maximal amount of money is just about enough for financing the production. This is the equilibrium level in the model. Among the possible or permitted alternatives this level has the biggest budget, and thus represents budget-maximizing. It is not optimal but bigger than that. It is however neither unconstrained, which it would have been without the influence of the politicians. The politicians clearly limit the amount produced. It is therefore not correct to say that the politicians, in this theory, are “completely under the control of the bureaucrats”. That formulation, rather, reveals a complete lack of understanding of the most basic property of Niskanen’s theory – namely that it is an equilibrium theory in which bureaucrats and politicians balance each other.
As for Downs it should be enough to quote the following (Downs, 1967, p 258): “If the bureau as a whole were really excessive in size, some political party would advocate drastic reductions affecting a whole spectrum of minority-serving bureaus. This party would receive the vote of every citizen who believed he was paying more support to wasteful bureaus than he was receiving from those minority-serving bureaus that benefited him directly. If such citizens were in the majority, the bureau-wrecking party would be elected, and would presumably slash the size of bureaucracy as a whole.” The politicians here are neither passive, nor without means of control.
In his chapter on bureaucrats Lewin has failed to give a fair account of the public-choice theorists’ writings and ideas about bureaucracy and bureaucrats’ incentives. In particular he has failed to describe the real nature of the budget-maximization goal in a fair and correct way. Finally, he has by no means shown that bureaucrats, on the whole, tend to be public-minded rather than self-interested.
The Fifth Chapter
Lewin’s final chapter covers several topics, but here I will concentrate on one single aspect. On the very first page of his book he introduced the “specialists in the fields of rational choice, political philosophy, and electoral research” and declared “cross-fertilization” to be the aim of his undertaking. Now, my comment is that in the end there is not, and has not been, much of cross-fertilization. In the final chapter the philosophers and the electoral researchers are mentioned just once or twice each, whereas the public-choice school is mentioned again and again – about 20 times if I have counted correctly. Furthermore some of the remarks about them are quite aggressive, derogatory, or rallying. This, I think, substantiates my contention that Lewin’s book, in reality and foremost, is an attack on public choice.
My criticism of Lewin’s handling of the public-choice school has so far been related to the three chapters dealing specifically with voters, politicians and bureaucrats. A few more comments of a more general character are however still due.
First, I have so far said very little about Lewin’s opinion that political actors, according to the public-choice school, always and necessarily are just self-interested egoists. It should therefore be added that public choice, generally, not in any way depends on a self-interest assumption. Buchanan and Tullock write, for instance, in The Calculus of Consent, one of the five representative public-choice books mentioned by Lewin, that (Buchanan and Tullock, 1962, p 3f) “(t)he analysis does not depend for its elementary logical validity upon any narrowly hedonistic of self-interest motivation of individuals in their behavior in social-choice processes. The representative individual in our models may be egoist or altruist or any combination thereof.”
As an example of this possibility a passage from Downs’ An Economic Theory of Democracy – also one of Lewin’s five books – may be quoted. Downs writes (Downs, 1957, p 37) that “(i)t is possible for a citizen to receive utility from events that are only remotely connected to his own material income. For example, some citizens would regard their utility incomes as raised if the government increased taxes upon them in order to distribute free food to starving Chinese. There can be no simple identification of ‘acting for one’s own greatest benefit’ with selfishness in the narrow sense since because self-denying charity is often a great source of benefits to oneself. Thus our model leaves room for altruism in spite of its basic reliance upon the self-interest axiom.”
The very last sentence in this quotation makes a distinction between assumptions actually made because they seem reasonable, and the set of all different assumptions permitted by the theory, because of its nature. This distinction, in spite of its fundamental importance, is strangely absent in Lewin’s book. Having said that I have also begun my second comment in this section, which is about distinctions missed by Lewin. Another such distinction is perhaps still more important. What I am thinking about is the fundamental distinction between the social function of an individual’s actions, and his or her motives, or incentives, for carrying them out. This, of course, is the distinction made by Adam Smith in his keen statement that “It is not from benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest.” This distinction is fundamental in public-choice theory, it appears there repeatedly, for instance in the five books mentioned by Lewin, but Lewin himself does not mention it all. That makes his whole discussion about incentives much more obtuse than otherwise it would have been. It may, for example, be the case that vote-maximizing among politicians – if the relevant political institutions are properly designed – is a good thing enhancing general welfare.
A third, and final comment, is based on the observation that the public-choice school is engaged in theory building. This, indeed, I would say, is what really distinguishes it from political science in general. But theory should be judged as theory, on its own terms. For getting an idea about these terms the following quotation from The Calculus of Consent (Buchanan and Tullock, 1962, p 7f) may be helpful: “Any theorizing, be it about private or collective decision-making, must initially be based on simple models which define clearly the constraints within which the individual actor operates. In a preliminary analysis, simplification and abstraction are required. The institutional constraints on human action must be stripped of all but their essentials. … progress is made from by building from the ground up, and we do not propose to present a fully developed theoretical structure.” This suggests that theories, and in particular theories in their infancy, should be judged with regard to such criteria as general reasonableness, fruitfulness and logical consistency, rather than with regard to whether some individual assumptions are exactly true or not. This, also, seems to have escaped Lewin.
A Note on Lewin’s Technique of Presentation
I said above that Lewin, in his last chapter, mentions the public-choice school quite frequently, and much more than the specialists in political philosophy and electoral research. It should also be obvious by now that the discussion about the public-choice school is a main concern for Lewin throughout his book. This makes one expect that Lewin, to a reasonable extent, substantiates his presentation of the public-choice theorists with references to their works. The table below, which covers all foot-notes in Lewins book, shows his record in this respect.
|Chapter in Lewin’s book containing the notes|
|Works the notes are referring to||1||2|
|5||Total number of notes|
|Downs: An Economic Theory of Democracy||2||0||1||0||1||4|
|Downs: Inside Bureaucracy||1||0||0||0||0||1|
|Buchanan and Tullock: The Calculus of Consent||1||0||0||0||0||1|
|Tullock: The Politics of Bureaucracy||1||0||0||0||0||1|
|Niskanen: Bureaucracy and Representative Government||1||0||0||1||0||2|
|Total number of notes||44||63||25||54||20||206|
The result is surprising. We see that there are altogether only nine foot-notes referring to the five public choice works highlighted by Lewin. It may be added that among the remaining 197 foot-notes only about five refer to other works which may be considered as emanating from the public-choice school. This hardly a lot.
Going into detail we also see, for instance, that the chapter on bureaucracy contains only one footnote referring to public choice, in this case to Niskanen’s Bureaucracy and Representative Government. The discussion about the other two books on bureaucracy, that is Downs’ and Tullock’s books, in the chapter on bureaucracy, is carried out without any references at all in that chapter. The discussion, rather, relies on a few very general characterizations made in the first chapter. Hence the solitary references to these books indicated in the table’s first column.
It may be added that the pieces of text in Lewin’s book, to which the footnotes are attached, not necessarily are direct verbal quotations. They may very well be presentations of the works referred to in Lewin’s own words. Take, for instance, the one and only reference to Tullock’s The Politics of Bureaucracy. The piece of text to which it is attached is the following (p 6):
“The third group of actors – the bureaucrats – soon came to be included in public-choice theory as well. Tullock claimed in The Politics of Bureaucracy that governmental agencies, in contradistinction to private enterprises, expand without any attempt at minimizing their costs.”
This is the complete and only description of The Politics of Bureaucracy in Lewin’s book. When Lewin occasionally says anything more about Tullock’s book, for instance in the chapter on bureaucracy, it is built on this single characterization, which, it should be added, is remarkably crude and uninformed. I am quite confident that people who have read Tullock’s book seriously, whether they like it or not, will agree with me on this particular point.
Lewin starts his book by asking in the very first sentence: “Is it self-interest or public interest that predominates in public life?” This is an interesting question, and on the back of the book the publisher asks it again, and then goes on saying that “This question has been debated in many fields, and through the 1980s a consensus developed, supported by extensive research, that in political decisions and actions people are largely motivated by self-interest, and not by the common good.” Then, a few lines later, the publisher says that “In this book … Lewin … challenges the accepted wisdom”, and finishes by saying that Lewin, in short, shows “that the plausible and prevalent theory that egoism rules simply doesn’t match the facts”.
It is interesting to note the negative formulation of this conclusion. Lewin has shown, we learn, that it is not true that egoism rules. In spite of the formulation of the opening question the publisher, it seems, does not feel able to declare that Lewin has positively shown that public interest predominates public life. And if the publisher feels so, that may be for the very good reason that Lewin, in fact, has not shown anything like that. What he has shown about interests is, I would say, just one thing, namely that voters probably do not vote according to their pocket-books in Lewin’s particular sense. That does not, however, exclude that they behave selfishly in other ways, and therefore it does not mean anything for the balance between self-interest and public interest either. As for politicians and bureaucrats he has not, I submit, shown anything.
To this I will just add that Lewin, as I have already said, does not discuss interest organizations and the actors within them at all. Considering the nature of his undertaking this is clearly a deficiency. Similarly, he has no comments at all on the whole literature, and discussion, about rent-seeking.
To me the conclusion is therefore inescapable. Lewin has not contributed anything to the discussion about whether self-interest or public interest predominates in public life. The question, as quoted above, is still there, as interesting, and as unanswered, as before.
Furthermore, as I have tried to demonstrate, Lewin’s attack on public choice is a failure.
Buchanan, J M & Tullock, G, 1962, The Calculus of Consent, Logical Foundations of Constitutional Democracy. Ann Arbor: The University of Michigan Press.
Downs, A, 1957, An Economic Theory of Democracy. New York: Harper Collins Publishers.
Downs, A, 1967, Inside Bureaucracy. New York: Harper Collins Publishers.
Niskanen, W A, 1971, Bureaucracy and Representative Government. Chicago: Aldine Publishing Company.
Tullock, G, 1965, The Politics of Bureaucracy. New York: University Press of America.
Originally published on Erik Moberg’s website.