Rev Austrian Econ DOI 10.1007/s11138-012-0194-4
The theory of interventionism as an Austrian theory of slowdowns Pál Czeglédi
# Springer Science+Business Media New York 2012
Abstract This paper addresses the question of whether the Austrian theory of interventionism helps us understand what growth economists call decelerations or slowdowns. It is proposed that the theory of stagnation based on non-productive entrepreneurship Coyne et al. (Journal of Austrian Economics 23(4), 333–346, 2010) complements the theory of interventionism, and when they are combined, a theory of slowdowns is the result. The consequences of this theory are used in an empirical investigation whose strategy is based on two hypotheses. One is that the interventionist process has institutional determinants which can be derived from the theory. The other is that interventionism is the mechanism of the slowdown of economic growth not only for those countries that have experienced relatively high growth rates in the past but also for those with normal growth rates. Using logit regressions it is shown that reductions in any area of economic freedom can be seen as one of the causes of slowdowns. Keywords Interventionism . Slowdowns . Economic freedom . Entrepreneurship JEL B53 . D72 . O43
1 Introduction The Austrian economics interpretation of economic development has always been that it is economic freedom that makes entrepreneurs able to discover productive profit opportunities, the constant discovery and exploitation of which constitutes what can be called economic development.1 Over the past decades this interpretation has received more and more affirmation. On the theoretical side our understanding of Mises (1927/2000, p. 10) writes that “… all that has created the wealth of our time can be traced back to capitalist institutions.”
1
P. Czeglédi (*) Department of Economics, University of Debrecen, 4028 Debrecen, Kassai str. 26, Hungary e-mail:
[email protected]
P. Czeglédi
entrepreneurship (Kirzner 1973; Holcombe 2007) and its relations to institutions (Baumol 1990; Boettke and Coyne 2003) has been greatly improved. On the empirical side, the development of the economic freedom database (Gwartney and Lawson 2003) and the empirical investigations based on it (e.g., Chauffour 2011; Faria and Montesinos 2009)2 has reaffirmed this view. In addition, it has also been affirmed that entrepreneurship is the channel through which economic freedom turns into development (Bjørnskov and Foss 2008; Kreft and Sobel 2005; Nyström 2008; Sobel 2008; Sobel et al. 2007). In sum, it turns out, that the facts of economic development are in line with the theory of economic freedom and entrepreneurship. To understand economic growth, however, it may not be enough to look at longrun averages of growth rates and their explanatory factors. As the ambitious and fastgrowing literature on “accelerations” and “decelerations” shows,3 it would pay off to examine the ups and downs of the growth process instead of examining long-run cross sections. The facts uncovered by this literature pose a question for those looking at development through the lenses of entrepreneurship and economic freedom: Do these concepts help us understand accelerations and decelerations, too, just as they have helped us understand the long waves of development? This paper can be seen as a first step in answering this question as much as it concentrates on slowdowns in economic growth. The main hypothesis of this paper is that Austrian economics does have a theory of slowdowns. My proposition is that the Austrian theory of interventionism and non-productive entrepreneurship (Coyne et al. 2010) combined result in a theory of slowdowns in which economic freedom, political institutions, ideology and culture can be seen as elements of the same process. Based on this theory I propose that the interventionist process can be seen as a distinct cause of the slowdowns in economic growth. I will argue in four steps. After reviewing the literature on accelerations and decelerations in Section 2, I will present the theory of interventionism combined with the theory of stagnation (Coyne et al. 2010) as a theory of slowdowns in Section 3. This makes it possible to derive different hypotheses concerning the effects of institutional factors. Section 4 is an empirical evaluation of the propositions derived. Section 5 concludes.
2 What can we learn from the literature of accelerations and decelerations? The analysis of accelerations and decelerations is a relatively new segment of the literature in the field of growth theory, the pioneering works of which are those of Pritchett (2000, 2003). With these works he turned the attention of many economists to the process of growth instead of long-run averages4 by emphasizing that growth
2
See the annually published Economic Freedom of the World Reports (Gwartney et al. 2011) for an extensive list of publications that use this measure. 3 For references and detailed discussion of this literature see Section 2. 4 It must be noted that Pritchett’s (2000) argument and the follow-up literature concerned with accelerations can be seen (and is intended to be seen) as a serious and very convincing attack on the widespread methods used in growth econometrics. Cross country regressions are pointless because the growth rates are very volatile, particularly in developing countries. Average growth rates hide rapid upswings and stagnations.
The theory of interventionism as an Austrian theory of slowdowns
patterns are diverse (Prichett 2000) requiring a group of theories explaining “states” of growth and transitions between these states (Pritchett 2003). Following this line of research Hausman et al. (2005) study “growth accelerations” by which they mean an increase of at least two percentage points in the growth rate which is sustained at over 3.5 % for at least 8 years and come to the striking conclusion that it is extremely difficult to find good predictors for these periods. More importantly, sustained and unsustained accelerations are different in terms of their causes.5 A political change towards more democracy is a feature of sustained growth accelerations but not of unsustained ones.6 Economic liberalizations have a positive impact on sustained accelerations, but financial liberalizations do not (although they have a positive effect on unsustained accelerations). Positive terms of trade shocks also have an effect in the unsustained case only. However, all these factors are very poor at predicting accelerations. They might be even poorer than one might think seeing that Jong-A-Pin and De Haan (2008) show that Hausman et al. (2005) are wrong in evaluating the effect of regime changes.7 After correcting for this error Jong-A-Pin and De Haan (2008) conclude that political regime changes have no predictive power while economic liberalizations have a slightly stronger effect than what is shown in the original paper. This is the conclusion they arrived at in Jong-APin and De Haan (2011), applying a slightly different filter to find the starting year of accelerations. These institutions might not matter at all, as the results of Jones and Olken (2008) suggest. They show that accelerations do not result from investment; rather they are caused by a change in TFP, although it is also true that accelerations are not accompanied by an improvement in institutions measured by democracy, the rule of law, and corruption. What seems to be important in triggering growth accelerations is openness to trade, which contributes to growth not by increasing aggregate demand but by reallocating resources. The conclusion that good institutions are not required to switch to a “growth” regime but are required to sustain them is reaffirmed by Jerzmanowski (2006). When looking at the different institutional measures as possible determinants of transition probabilities between different growth regimes he finds that voice and accountability and democracy are very important determinants of a stable growth regime. Taking his analysis a step further, Jerzmanowski (2011) concludes that the effect of policies— inflation, exchange rate overvaluation, government size—depends on the quality of institutions. In addition he finds that democracy lowers the probability of a crisis period but also that of a miracle growth period. Additional insights can be drawn from Kerekes (2012) who applies a more general model and concludes that the role of the quality of institutions is special inasmuch as “successful” countries have better institutions than “moderately successful” ones, but the quality of institutions is not better in “moderately successful” countries than in “failing” ones.
5
A sustained acceleration is an acceleration after which the growth rate remains above 2 % over a 10 year horizon (the growth between the yeart+7 and t+17 is higher than 2 %). Their results also show that autocratic government increases the probability of an (unsustained) acceleration. This prediction is used by Dovern and Nunnenkamp (2007) to explain the fact that aid is more productive in “bad states” when identifying the criterion of success with acceleration. 7 As a result of a misreading of the Polity IV data their variables do not reflect what they want them to. 6
P. Czeglédi
In parallel with the literature on accelerations there seems to be a renewed interest in economic convergence, and the results of this line of research come to interesting conclusions concerning countries that are on their way to becoming developed. One of these is Eichengreen et al. (2011) who apply the same methodology as Hausman et al. (2005) to investigate decelerations. As they show, the slowdowns they identify are primarily productivity slowdowns which will occur at a higher level of per capita income in more open economies. However, they do not find any role for financial openness, terms of trade shocks and political regime changes.8 The results seems to support a view that the growth process is more complicated than the standard institutionalist view (Acemoglu et al. 2005) would imply when it suggests that political institutions will determine economic ones, which will finally lead to economic growth. This view is also challenged by Paldam and Gundlach (2008), who propose that this “grand transition” view is simply not in accord with the facts. Rather, as the results reviewed above also indicate, the growth process represents an interrelated process of culture, formal institutions and development. Thus, when looking for the determinants of catch-up the fundamental question is not what factors create fast growth, but what factors interrelate in a way that constitutes a growth regime.
3 Slowdown as a process of interventionism and unproductive entrepreneurship The literature just reviewed can be seen through the lenses of Austrian economics. I will explain that a possible understanding of slowdowns can be built on the theory of interventionism resulting in an Austrian theory of slowdowns that will be presented in the three following steps. First, I will examine those institutions that play a role in the process of interventionism. Second, I will discuss the mechanisms that carry on the interventionist process. Third, as an explanation for the interrelationship between the interventionist process and economic development, I will propose that these mechanisms are the same as those which Coyne et al. (2010) identify; they are just seen from a different viewpoint. 3.1 The institutional determinants of interventionism The theory of interventionism that was first developed by Mises (1929/1996, 1940/ 1998) is summarized by Ikeda (1997, pp. 192–193) as a process in which [a] set of initial relative-price and ideological conditions encourages the state to intervene at the margin in a particular way, which in turn further distorts relative prices (i.e., explicit prices and implicit opportunity costs), generates negative unintended consequences, and causes a revision in ideological preferences. The result is a new set of conditions that is even more favorable for further intervention. This process will continue unless exogenous ideological forces bring it to a halt. 8 Although it seems that they commit the same mistake Jong-a-Pin and de Haan (2008) identify in Hausman et al. (2005)
The theory of interventionism as an Austrian theory of slowdowns
There are at least two features of this theory that makes it different from other theories of interventions or market regulation. First, a great emphasis is given to the knowledge (coordination) problem. Just as the knowledge problem is the problem (Hayek 1945) of the market process, it is also the main driving force behind interventions. Second, interventionism is seen as a process, too, that reacts to the unintended consequences of previous interventions and the unpredicted reactions of the entrepreneurs on the market. As the main three factors that drive the interventionist process are discoordination (originating from the knowledge problem), interest group logic and ideology, I propose that the three underlying institutional factors that must be taken into consideration are (1) the level of economic freedom, which I associate with the security of property rights broadly understood, (2) the political constraints on executives, and (3) culture and political ideology. Bureaucrats (regulators) facing a knowledge problem (Ikeda 1997, pp. 93–99) are simply ignorant of some of the consequences of their regulatory actions. To put it differently, their actions will always have certain consequences which they did not predict before the intervention. The more interventionist the initial system is, the more severe this knowledge problem becomes. Simply put, if central planners of a socialist economy are “groping in the dark” (Mises 1920/1990, p. 17), regulators of a mixed economy are groping in the dusk, and the more they intervene, the darker it becomes. They do not miss prices as a central planner would but they must rely on distorted prices implying that even if their intentions are good, they will miscalculate the effect. That implies that the level of economic freedom is a determinant of the interventionist process. The reason is that the less economic freedom there is, the more ignorant the regulators will be concerning the effects, costs and benefits of their actions, since less economic freedom means more distorted prices, which means a higher level of ignorance. Without the true costs in mind they will be more inclined to intervene. For similar reasons, the knowledge problem of private agents is more serious, too, in cases of less economic freedom.9 This ignorance of the regulators is further deepened by the reduced error-correction capacity of the market (Ikeda 2005, p. 36). As Ikeda (1998) explains, when the distortion of prices is larger, the possibility of coordinating entrepreneurial discoveries is smaller. That is, as private property rights are eroded, the possibility that entrepreneurs discover new productive opportunities is also curtailed. The role of formal constraints on executives seems to be more straightforward, since these constraints represent direct limitations of the interventionist’s hands. Although there are serious questions as to whether such limitations can ever be permanent (Kurrild-Klitgaard 2005), they increase the cost politicians incur on themselves when intervening. The conclusion is not as obvious as it first seems for several reasons. One is the Olsonian claim (Olson 1982), that the long-run stability of the political system is the best environment for distributional coalitions to grow roots. Second, firm constraints on executives, by reducing the risk of expropriation, mean a more secure system of property rights. All in all, there are good reasons to be
9
As Leijonhufvud (1981, p. 250) emphasized in the case of inflation, because people cannot rely on private contracts to the same extent as with no (or less) intervention, they will seek “political compacts” instead.
P. Czeglédi
ambiguous when thinking about the role democratic checks and balances play in slowdowns. As a third important determinant of the interventionist process I identified two different but interrelated factors. I understand political ideology in the same way as Ikeda (1997, p. 113) namely, the “public chooser’s attitude towards state expansion”. This refers to the preferences of those players who decide whether to introduce another measure to constrain market forces. Culture is a broader concept and it does not only include the preferences of the political actors, but the electorate, too. Of the many aspects of culture, here it is fruitful to emphasize the one that reflects the attitude towards individual freedom (Gorodnichenko and Roland 2011) and the idea that people’s motivations (specifically the willingness to be altruistic) can be of the “pure” kind, that is, they may originate from ideology (Colombatto 2011, pp. 49–52). In this light the relevant aspect of culture and ideology is a non-consequentionalist attitude towards the free market and against interventionism. This view of culture and ideology leads us to propose three channels through which it has an effect on the interventionist process. On the side of the political decision-maker a freedom-biased ideology—understood as a view of economic freedom as an end in itself—is of course an important reason not to remedy the result of one intervention with another. A decision-maker with an ideology that is biased towards freedom to a larger extent will have a higher reservation price for an intervention. This increases the costs of rent-seeking through lobbying for intervention. Another channel through which the culture which is more inclined to respect individual freedom will increase the payoff of rent seeking is the culture shared by the general public. Sharing the ethics of individualism is to think of the market process as a positive-sum game (De Soto 2009), and see the difference between productive and unproductive activities. In a more individualist society people will dislike the redistributive actions of the government to a larger extent. As a result, those activities that promise a return from redistribution will be condemned to a larger extent, too, which reduces the (subjective) payoff of such activities. That is, resource owners with a more individualistic culture will be less inclined to employ their resources in rent-seeking activities. A third effect of culture and ideology in my framework comes from the hypothesis according to which culture is a fundamental institution of a market economy (Boettke 2001). The same rules of the market will work better in a cultural environment which is more against rent seeking. This role of culture in the interventionist process is similar to that of economic freedom since it is a determinant on whether the formal institutions of economic freedom work in such a way as to create knowledge through entrepreneurial discovery. 3.2 The mechanism of interventionism Generally, there are two kinds of argument that can explain how the interventionist process moves forward. One is based on the knowledge problem; the other is based on rent seeking and distorted incentives. Ikeda’s (1997) argument is completely based on the knowledge problem argument and he argues that interventionism can even be explained by supposing a benevolent
The theory of interventionism as an Austrian theory of slowdowns
government. When such a social planner faces a knowledge problem its actions will always have unintended consequences which—depending on the constraints described above—it will wish to cure with newer interventions. 10 The government is, however, not a well-intentioned social planner and one may use the theories of rent seeking to understand the interventionist process. One can find at least two reasons why lobbying for intervention is much more in line with political incentives than is lobbying for dis-intervention. One is the general claim that “good politics concentrates benefits on well-organized and well-informed interest groups in the short run, while dispersing costs on the ill-organized and ill-informed mass of voters (both rationally ignorant and rational abstainers) in the long run” (Boettke 2012, p. 11). In addition those well-informed and well-organized groups are small for reasons explained by Olson (1982), and the small groups will prefer to see a redistribution of income toward themselves rather than an increase in total income. The second reason is the fact that the gains these small groups receive are transitional (Tullock 1975). An intervention will grant privileges to a group but the “successors” to the initially privileged group will have to bear higher entry costs and, as a result, will make a profit that is just as ‘normal’ as they would make without the interventionist measure they inherited. Yet, a dis-intervention would cause them a transitional loss and it would be completely against the “rule of good politics” just cited to compensate them for their consent to a dis-intervention. Ideology and ideological change is another factor in addition to the knowledge problem and rent-seeking which partially explains the steady expansion of interventions. Interventions make ideological preferences change marginally towards a state that is more and more benign to newer interventions. Ikeda (1997, pp. 181–186) shows how enacting interventionist measures leads to insecurity and thus shifts preferences towards more security (interventions) on the margin. As Klein (2005) explains, it is possible to use theories to describe as rent-seeking the actions of those who are seemingly (and honestly) driven by good intentions. This is because when people enter government organizations the most rational act is to accept the culture of the organization. Using Yandle’s (1983) famous terms, bureaucrats become “Baptists”: they become ideologically committed to the policy of the bureau. In this way they create possibilities for political entrepreneurs (Simmons et al. 2011) to discover newer groups of “bootleggers” to match them and reap the profit. Explaining dis-interventionism receives less attention in the theory of interventionism, and as the Ikeda-quote above shows, remains mainly exogenous. One obvious reason to dis-intervene is an exogenous cultural or ideological change. A second is the depletion of the “reserve fund” (Mises 1949/1998, p. 851–854). Since the aim of interventionist rent-seeking is to redistribute wealth, when there is more wealth to redistribute, there are more incentives to seek rent. A third reason to reverse the tendency to interventionism is that during the interventionist period errors
As Friedman and Kraus (2011, p. 113) argue concerning the financial crisis of past years, “… the original policies may be recognized as mistakes and repeals. But this rarely happens in the real world… its unintended effects will not be attributed to the regulation in the future, given a general continuity in human psychology and in the history of ideas. …subsequent regulators will tend to assume that the problem with which they are grappling is a new excess of capitalism, rather than being an unintended consequence of previous regulators’ mistakes in the regulation of capitalism”.
10
P. Czeglédi
committed by entrepreneurs as well as public choosers accumulate and will lead to a systematic crisis which may make the public choosers reverse the process. Ikeda (1997, p. 137–144) explains that once the process of dis-interventionism gets started, some elements of the process can be paralleled with interventionism, but because of bottlenecks and intervention-biased ideology this process may be more discontinuous. 3.3 Interventionism as stagnation The economic forces that move interventionism forward, I claim, are the same forces that can be observed as those which slow down development. Indeed, they are the forces that Coyne et al. (2010) uncover as the forces of stagnation. As we have seen above, interventionism provides (1) possibilities for political entrepreneurship, (2) incentives to employ resources in rent-seeking activities, (3) and even reasons to believe that further interventions are desirable. These three mechanisms that make interventionism a process are paralleled with those of stagnation as explained by Coyne et al. (2010) in Table 1. Coyne et al. (2010) argue that “non-productive” entrepreneurship can be just as self-generating as productive entrepreneurship (Hayek 1968/2002; Holcombe 1998). They call those mechanisms ‘multiplier effects’, through which the realization of previous entrepreneurial activities create opportunities for further entrepreneurship. According to their theory there are three channels through which the multiplier effect of non-productive entrepreneurship becomes prevalent: (1) the reverse side of creative destruction (which one might call destructive creativity) which means that nonproductive entrepreneurship disrupts the old (good) equilibrium; (2) non-productive entrepreneurship opens up the way for new market niches for further non-productive entrepreneurship. In effect this refers to the mechanism through which a new method of rent-seeking becomes attractive to other rent-seeking groups or individuals; (3) nonproductive entrepreneurship creates social capital that can be transferred to other nonproductive activities. Political lobbying creates a network which makes further lobbying less costly by eliminating the fixed costs of constructing a network. This makes the process path-dependent because of the specific investments made in social capital. These three channels of the multiplier effect of non-productive entrepreneurship identified by Coyne et al. (2010) (listed in the right-hand column of Table 1) are virtually the same as the three channels I identified as mechanisms of the interventionist process (listed in the left-hand column of Table 1). Destructive creativity is the Table 1 The “Balance Sheet” of Interventionist Slowdowns Mechanisms of interventionism
Mechanisms of stagnation (Coyne et al. (2010))
1. Bureaucratic and political entrepreneurial discoveries
1. Destructive creativity
2. New incentives for employing resources in rent-seeking activities
2. New, unproductive market niches
3. Becoming “Baptists” through ideological commitment
3. Creating social capital transferable to other non-productive activities
The theory of interventionism as an Austrian theory of slowdowns
exercise of destructive entrepreneurship, that is, the discovery of new ways of redistribution in bureaucracy and polity. The exploitation of these new opportunities is a waste of resources from the point of view of economic development. Finally ideological commitment can be seen as a resource that is transferable between different “problems” that need solving. All in all, what is an interventionist process from the bureaucratic and political entrepreneur’s point of view is an accumulation of unproductive entrepreneurship from the market entrepreneur’s point of view. That is the theoretical reason to claim that interventionism leads to the slowdown of economic development. Using this framework one can conclude that the interventionist process slows down economic growth by (1) distorting the incentives of resource owners, (2) by providing more opportunities for political entrepreneurship to be successfully exploited, and (3) by providing less opportunities for market (productive) entrepreneurship to be successfully exploited. By distorted incentives I mean that interventionist measures provide incentives for resources to be employed in rent-seeking activities, and since rent-seeking means expropriation of wealth created by market entrepreneurship it creates disincentives to exploit productive profit opportunities. Political entrepreneurship becomes more successful, because interventionism creates opportunities for further political and bureaucratic entrepreneurship. Finally the intervention that is enacted to exploit the political or bureaucratic opportunity will prevent market entrepreneurs from discovering those profit opportunities that they would have discovered without the intervention. The first consequence reflects static costs, while the latter two are dynamic costs whose identification is unique to the Austrian view, as also argued by Benson (2005). It is important to notice that the costs of interventionism for economic development go beyond the waste of resources caused by rent seeking. This is also reflected by the argument in sections 3.1 and 3.2, where it was shown that the interventionist process is usually modeled as driven by political incentives and the knowledge problem (discoordination). Even if the government were well-intentioned and interventionism were not propelled by the rent-seeking logic, the interventionist process would still be costly by discoordinating the market and impeding the discovery of productive profit opportunities.
4 Some empirics of interventionist slowdowns The theoretical considerations above suggest that the roots of slowdowns must be found in the process of interventionism, which is affected by the level of market coordination and property rights protection, constraints on those who have political power to intervene, and political ideology and culture. In the following I will try to investigate this proposition empirically. That is, I will examine two determinants (or groups of determinants) of whether a slowdown occurs. One is the fact that an interventionist process can or cannot be seen, and the other is the three institutional determinants of the interventionist process mentioned above. 4.1 Determinants of slowdowns and interventionisms To begin an empirical investigation there is a need for an empirical definition of a slowdown and of interventionisms. When it comes to slowdowns I will need a
P. Czeglédi
different definition of a slowdown than the one in Eichengreen et al. (2011) for two reasons. One is that Eichengreen et al. (2011) reverse the definition of Hausman et al. (2005) which itself has been improved by Jong-A-Pin and De Haan (2011). In addition, slowdowns as identified by Eichengreen et al. (2011) seem to be devised to describe an episode during which a relatively fast (3.5 per cent averaged over 8 years) growth dies off. But there is nothing specific about the interventionist process which makes it a characteristic of fast growing countries. Rather they may characterize even countries with “normal” growth rates. What I claim is that interventionism is a way of slowing down. Since in the long run 2 % growth is generally considered a sustainable rate I modify the definition by setting this 2 % average growth as a threshold for a slowdown. Bearing in mind the modified definition of Jong-A-Pin and de Haan (2011, p. 96) I define a stagnation or a slowdown beginning in yeart as follows. A slowdown starts in yeart if gt+1
2.0 percent per annum, gt,t+7 −gt−7,t <−2.0 percent per annum, gt,t+7 <2 percent per annum, where gt is the growth rate in yeart, while gt,t+7 is the time trend of log per capita GDP between yeart and t+7. I apply this screening on per capita GDP measured at 2005 international dollar (chain series)11 from Heston et al. (2011) for all the data available. This includes data for 189 countries between the years 1950 and 2009 although not every country has data for all the years. This exercise leads to 525 slowdowns. If we then apply a Chow test (as in Eichengreen et al. (2011, p. 6) and in Hausman et al. (2005, p. 306)) to single out just one slowdown year within a 5-year interval in cases when slowdowns are found in consecutive years the number of slowdowns reduces to 216. Following again the “tradition” of the literature on accelerations and decelerations (Jong-A-Pin and De Haan (2011, p. 100); Hausmann et al. (2005, p. 321)) a slowdown dummy is created by introducing a dummy that equals one in yeart-1, t and t+1 if yeart is identified as a slowdown year by the above method. Beside slowdowns I have to identify interventionisms and dis-interventionisms. To define these concepts empirically I will use the Economic Freedom of the World (EFW) index of the Fraser Institute (Gwartney et al. 2011).12 This index provides data 11
The variable is called rgdpch in Heston et al. (2011). This is the generally used GDP variable of the many provided by their database to calculate growth rates and the one they also recommended for long-run comparisons (Summers and Heston 1991, p. 344) because the chain-linked method reduces the problem related to changing relative prices that arises when the Laspeyres method is used. 12 There are two widely used indexes of economic freedom. One is that of the Fraser Institute (Gwartney et al. 2011), the other is that of the Heritage Foundation (Miller et al. 2012). I have three reasons to use the Fraser Institute’s index: it provides data over a longer period of time (since 1970 as opposed to 1995), it is more widely used in the academic literature (see Rode and Coll (2012), for example, and the literature cited therein, or Gwartney et al. (2011, pp. 1–4)), and finally, as we will see, this is the index for which there exists a criterion as to what should be considered a significant change in the index. De Haan and Sturm (2000) conduct an explicit comparison of the two indexes and conclude that they are similar to such an extent that the main statistical interferences are not disturbed by whichever index is chosen for the analysis. Note, however, that the composition of both indexes has changed somewhat since then.
The theory of interventionism as an Austrian theory of slowdowns
on five areas of economic freedom between 1970 and 2009 for every 5 years before 2000 and for every year after 2000 and evaluates economic freedom on a scale between 0 and 10, with a higher number meaning a higher level of economic freedom.13 The data base includes data for 140 countries. An important question is what to consider a significant increase or decrease in economic freedom. Based on Paldam’s (2003, p. 464) guess, according to which the measurement error in the EFW index is in the range of 0.25 to 0.5, and which concludes that “if two countries differ by less than 1/2 point, the difference should be disregarded, but differences of more than 1 point are probably significant”, I will use both critical numbers (0.5 and 1) to create an interventionist dummy, but my preferred choice is 0.5 for two reasons. One is that my definition is not based on the overall index, but on its areas, and second, when identifying slowdowns I do not compare different countries, but the same countries for different years. Accordingly I will call a quinquennium interventionist if any of the areas (on which data are available) reduces by at least 0.5 index points and none of them increases by more than 0.5 index points. Each year in this interventionist quinquennium is an interventionist year. Dis-interventionist quinqennia and dis-interventionist years are defined accordingly. Alternatively, I will use another definition for interventionism, and dis-interventionism, in which case the crucial change is 1 index point. I will refer to these definitions as interventionism or dis-interventionism with a narrow margin (defined as a 0.5 index point margin), and those with a wide margin (those defined as a one index point margin). We must note, however, that interventionism and dis-interventionism dummies are not independent because a country cannot have an interventionist and a disinterventionist year at the same time. There are only three possibilities: disinterventionism, interventionism, and none. This provides the reason to create a variable to describe interventions or the lack thereof in the following way: a variable called interventionism/dis-interventionism is equal to 1 if we are in a dis-interventionist quinquennium; it is minus 1 if we are in an interventionist quniquennium, and zero if neither is true. In addition to data on slowdowns, interventionism, and dis-interventionism, three other kinds of data are needed: ideology, economic freedom, and constraint on government executives. Data on “government ideology” come from the Database of Political Institutions (Beck et al. 2001). Particularly, government ideology is identified with the political orientation of the largest government party, being assigned a value of 1 if it is right-wing, -1 if it is left-wing and 0 if it is in the center of the political spectrum. 14 Another source of data on ideology is Bjørnskov and Paldam (2012) who create a capitalism-socialism (CS) score which is based on data from the World Values Survey 13
The five areas are (1) size of government, (2) legal structure and the security of property rights, (3) access to sound money, (4) freedom to trade internationally, and (5) regulation of credit, labor and business. See Table 12 in the Appendix (or Gwartney et al. 2011) for a more detailed exposition of these areas. 14 The coding follows the method of Bjørnskov (2005, 2008). In these papers he shows that the ideology of governments does have an effect on economic growth. In his brief review article on voting behavior and ideology Rubin (2001) concludes that thinking of ideology as a one-dimensional space is consistent with “what we have now come to understand is the nature of ideology” (ibid., p. 328). It may be even more interesting that he adds (ibid., p. 333) that ideology is even more important in explaining the policies that are actually selected than in explaining voting per se.
P. Czeglédi
and is meant to express the preference for private against public ownership of the population in general. The index—in theory—runs between −100 and 100 with a higher number being in favor of private ownership to a larger extent. The CS-score is available for 200 observations in the years 1990, 1995, 2000 and 2005. To fill in the missing years between these dates I use the following method: I use the value of the score in yeart for yeart+1 and t+2, then I use the value in t+5 and t to interpolate a score for t+3 and I also associate this score with t+4. By this process I probably get a score that is closer to reality, since it is likely that this value does not change abruptly. As the theory proposes that the level of economic freedom is itself a determinant of the interventionist process I will use the overall chain-linked index of Economic Freedom of the World (EFW) which was described above, as an independent variable. Given that before 2000 this index is also only available for 5 year intervals, I use the same interpolating technique as above to get the values within 5 year intervals before 2000. The variable that is meant to describe the formal constraints interventionists face is the “executive constraints” (Xconst) variable from Marshall et al. (2011a). Standardized authority scores are decoded following the methods Marshall et al. (2011b, p. 17) apply in the case of the polity variable to derive the polity2 variable. The value of this variable runs between 0 and 7 with 0 being the least and 7 the most constrained executives. Overall I have a panel database covering the years between 1956 and 200315 with the different data being differently available for these years. Table 13 in the Appendix gives descriptive statistics on these variables. In what follows I will use logit regressions16 to examine the claim that interventionism (or the absence of dis-interventionism) on the one hand, and the presence of institutional conditions I identified in section 3.1 on the other, raise the probability of having a slowdown year. Results in Tables 2 and 4 are those derived when pooled logit regressions are run and the coefficients are the changes in the log odds ratio. Results in Tables 3 and 5 are the marginal effects from the regressions in Tables 2 and 4. As the interpretation of the marginal effects is more straightforward, I will concentrate on Tables 3 and 5. The conclusions drawn from the theory can more or less be confirmed. Most importantly, the conclusion that the higher the interventionism/dis-interventionism variable the lower the probability of experiencing a slowdown (Table 3, column 2 and 3).17 The results in Table 3, column 2 show that the probability slowdown is reduced by 0.036 by a one unit change in the interventionist/dis-interventionist variable. This implies that having an interventionist year as opposed to a dis-interventionist one will reduce the probability of a slowdown by roughly seven percentage
15
Since the GDP data runs between 1950 and 2009, 1957 is the first and 2002 the last possible year to be identified as the starting year of a slowdown. By the definition of the slowdown dummy, the next and the previous year receives the value of one if these are slowdown years. 16 I chose logit (and not probit) regressions so that I could add country fixed effects without losing the consistency of the estimate, because as Wooldrige (2002, p. 484) shows, probit regressions with unobserved effects will not yield consistent estimates of the parameters in question. 17 Throughout the paper I refer to columns of tables as columns of the whole table not those of the results. That is, the first column with results is referred to as column 2.
The theory of interventionism as an Austrian theory of slowdowns
points, holding the other explanatory variables at their means. Seeing that the average probability of a slowdown is 10 % (see Table 13 in the Appendix), this is a substantial effect. In columns 4–6 in Table 3, instead of the slowdown dummy I added those factors that I proposed to be the main blocks of the institutional environments conducive to interventionism, namely the reserve fund (measured by per capita GDP), constraints on executives, and economic freedom (EFW summary index). The directions of these effects are in line with the theory of interventionism with the exception of the ideology of the government’s main party, which is not significant statistically in any of the specifications. The results in column 4 Table 3 predict, for example, that a country with the same level of per capita GDP and a one unit higher EFW level (the difference between the UK and Italy in 2009 (Gwartney et al. 2011)) will have a 1.4 percentage point lower probability of a slowdown in a certain year compared to the other with a one unit less EFW level and the same GDP level. Similarly the stronger constraints on executives and a culture with a more benign view of private property will reduce the probability of a slowdown. When the CS-score is included, too (column 6, Table 3), the sample size is drastically reduced because, as I explained above, Bjørnskov and Paldam’s (2012) data are only available for some of the years for which I have other data. The problem is, as can be seen from the results in columns 5–6, Table 3, that these variables are not significant when included together. What is more, the coefficient of the EFW index becomes positive in column 6, Table 3. This might be attributed to the fact that economic freedom, executive constraints, and the CSscore correlate with each other as shown in Table 15 in the Appendix. It is only government ideology that does not correlate significantly and strongly with the CSscore, although it correlates with the other two institutional determinants of slowdowns. The reason for their correlation is, I think, that these variables include the same factor: respect for private property which, with the help of the theory in section 3, might be thought of as the main fundamental determinant of interventionisms and slowdowns. As a confirmation of this idea I included these variables separately among the independent variables in Table 5. As shown, each variable is significant (with the exception of the government’s ideology), once included singly with the same sign, suggesting that a higher respect for private property makes a slowdown less probable. Both in Tables 2 and 3 and in Tables 4 and 5 I included regime change variables18 to account for a possible Olsonian explanation of slowdowns. One can argue that a regime change creates an environment that makes a slowdown less probable by breaking up the established system of distributive coalitions. The results argue slightly against this hypothesis. Democratic change is statistically significant in four cases (column 2, 3 and 4 in Table 3, and column 2 in Table 5) while an autocratic change is also significant in three cases (column 5 and 6 in Table 3 and column 4 in
18 Here I follow Jong-A-Pin and De Haan (2011, p. 101) who “use a dummy equal to one for the first 5 years after a political regime change and zero otherwise”. I also “differentiate between positive regime changes (i.e., more democracy) and negative regime changes (i.e., more autocracy).” (ibid)
−0.395 (−2.09)
**
0.599 (2.01)** −0.117 (−0.48)
−0.109 (−1.50) −0.147 (−4.33)***
114
number of countries
114
3256
0.010
114
2825
0.008
−0.003 (−1.07)
−0.364 (−1.14)
114
2825
0.019
−0.000 (−0.38)
−0.552 (−1.65)*
−1.109 (−1.40)
53
378
0.144
−0.015 (−2.68)***
3.015 (1.83)*
Heteroskedasticity robust z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
3256
pseudo R
number of obs.
0.013
−0.004 (−2.09)**
−0.004 (−1.99)**
regime duration
2
0.018 (0.07)
−0.327 (−1.87) 0.025 (0.10)
democratic regime change
autocratic regime change
−0.253 (−1.33)
1.230 (3.66)***
0.311 (4.26)***
−16.192 (−5.23)***
−0.052 (−3.50)*** −0.320 (−1.83) *
**
−0.152 (−2.06)
0.200 (2.99)***
−3.444 (−6.83)***
0.107 (0.55) *
0.121 (2.37)**
−2.843 (−5.68)***
CS-score
0.113 (2.22)
−0.289 (−3.66)***
−2.932 (−6.93)***
government ideology
constraints on executives
EFW summary index
ln(GDP)
**
−0.384 (−4.69)
interventionism/dis-interventionism with wide margin
interventionism/dis-interventionism with low margin
***
−2.865 (−6.76)***
Constant
dependent variable: dummy for slowdown years
Table 2 The Probability of Slowdowns: Pooled Logit Regressions
P. Czeglédi
114
3256
−0.000 (−2.00) 114
3256
114
2825
114
2825
−0.000 (−0.38) 0.000 (−1.07)
−0.000 (−2.10) **
−0.022 (−1.33) −0.049 (−1.67)*
−0.033 (−1.14)
−0.013 (−4.45)***
−0.009 (−1.51)
0.028 (4.36)***
0.002 (0.07)
−0.014 (−2.07) **
0.018 (3.01)***
−0.036 (−2.10)**
−0.030 (−1.83)*
0.011 (2.38)**
−0.027 (−3.70)***
53
378
−0.001 (−2.52)**
0.144 (1.91)*
−0.053 (−1.48)
−0.003 (−3.61)***
0.005 (0.54)
−0.006 (−0.48)
0.029 (2.11)**
0.059 (3.32)**
Heteroskedasticity robust z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
number of countries
number of obs.
regime duration
0.002 (0.10)
autocratic regime change **
−0.031 (−1.87)*
0.011 (2.23)
**
−0.036 (−4.77)***
democratic regime change
CS-score
government ideology
constraints on executives
EFW summary index
ln(GDP)
interventionism/dis-interventionism with wide margin
interventionism/dis-interventionism with low margin
dependent variable: dummy for slowdown years
Table 3 The Probability of Slowdowns: Pooled Logit Regressions, Marginal Effects at the Means
The theory of interventionism as an Austrian theory of slowdowns
P. Czeglédi Table 4 The Probability of Slowdowns: Pooled Logit Regressions dependent variable: dummy for slowdown years −2.843 (−5.68)***
constant
***
ln(GDP)
0.200 (2.99)
EFW summary index
−0.152 (−2.06)
−3.275 (−9.23)*** ***
0.182 (3.78)
−15.961 (−5.27)*** 1.534 (4.77)***
**
−0.087 (−3.47)***
constraints on executives government ideology
0.180 (0.93)
CS-score
−0.051 (−3.76)***
democratic regime change
−0.395 (−2.09)**
0.055 (−0.36)
−1.040 (−1.35)
autocratic regime change
−0.364 (−1.14)
−0.204 (−1.01)
3.378 (2.37)**
regime duration
−0.003 (−1.07)
−0.002 (−1.24)
−0.010 (−2.22)**
2
pseudo R
0.008
0.006
0.133
number of obs.
2825
4875
388
number of countries
114
156
56
Heteroskedasticity robust z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
Table 5) but usually it is only one of the two that is significant. The effect of the democratic change is consistently negative while that of the autocratic one is positive in some cases (column 6 in Table 3 and column 4 in Table 5). In addition, the regime durability variable of Polity IV as an independent variable is significant, too, predicting that a more durable regime lessens the probability of a slowdown in four cases (column 2, 3, and 6 Table 3 and column 4 in Table 5) although its effect is sometimes not higher than a half of one tenth of a percentage point.
Table 5 The Probability of Slowdowns: Pooled Logit Regressions, Marginal Effects at the Means dependent variable: dummy for slowdown years ln(GDP)
0.018 (3.01)***
EFW summary index
−0.014 (−2.07)
0.016 (3.82)***
0.077 (4.82)***
**
−0.008 (−3.51)***
constraints on executives government ideology
0.009 (0.92) −0.003 (−3.79)***
CS-score democratic regime change
−0.036 (−2.10)
−0.005 (−0.36)
−0.052 (−1.41)
autocratic regime change
−0.033 (−1.14)
−0.018 (−1.01)
0.169 (2.54)**
regime duration
−0.000 (−1.07)
−0.000 (−1.24)
−0.000 (−2.04)**
number of obs.
2825
4875
388
number of countries
114
156
56
**
Heteroskedasticity robust z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
The theory of interventionism as an Austrian theory of slowdowns
In Tables 6 and 7 I repeated the same regressions with country fixed effects, that is, I ran conditional fixed effects logistic regressions. Applying this method reduces the sample size because those countries experiencing no slowdown at all (in the sample) have to be dropped. That is, the sample now includes only those countries that have experienced a slowdown as defined above. This narrower sample is of course further reduced by the different degrees of availability of the dependent variables. Another disadvantage of the method is that the marginal effects can only be calculated if one knows the fixed effects (Wooldridge 2002, p. 492). Since I cannot formulate a plausible assumption concerning the fixed effects I have to make do with, and try to interpret, the effects on the log odds ratios. Table 6 shows the results when the regressions in Table 2 are run with this method. The main results remained roughly the same. More importantly, the coefficients of the interventionism/dis-interventionism variables are again negative with a magnitude similar to that in Table 2. Constraints on executives and the EFW index also have the sign that is expected, but the CS-score is not, possibly because of the fact that the sample size is even further reduced in this case, including only 14 countries (column 6, Table 6). I think that the oddly enlarged coefficient on the log GDP per capita here, as well as in column 4 in Tables 7 and 8, can be attributed to the small size of the observations and to the fact that within this small sample the standard deviation of GDP is much smaller, too. Table 7 is a reproduction of the regressions of Table 4 with the conditional fixed effects method. Again, the results do not change much except for the coefficient of the CS-score. In addition, regime change seems to matter, but so does the durability of the regime; in fact durability seems to be more important. The pseudo R-squares are still low, but they are comparable to what Jong-A-Pin and De Haan (2011) derives for the cases of accelerations with the same regression method. The pseudo R-squares in Tables 6 and 7 are, however, larger than in Tables 2, 3, 4 and 5, reflecting that the model is better at explaining within-country differences than between-country ones. What Eichengreen et al. (2011) find an important factor in a slowdown is spectacular here, too; no matter which kind of specification is considered, higher income increases the probability of a slowdown. They argue (ibid., pp. 8–9) that the reason is that above a certain level of income it is no more possible to raise productivity by reallocating labour from agriculture to industry and by applying foreign technology. Another reason which the theory of interventionism provides us with is that the higher the GDP the higher the reserve fund the depletion of which could constrain the interventionist process. All in all, with a method used in the literature reviewed in section 2 I have shown that interventionism—defined in two ways—and those institutional determinants that were hypothesized to affect the process can be shown to statistically matter for whether a growth slowdown occurs in a certain year. 4.2 Sensitivity of the results and the role of entrepreneurship Since I used relatively new definitions to identify both slowdowns and interventionisms, the question can be raised as to whether the results depend on the choice of
−0.473 (−3.29)
−0.532 (−7.73)***
***
5.388 (9.98)***
−15.410 (−0.01)
0.508 (0.18)
50.618 (3.79)***
86
2672
0.073 86
2672
0.071
−0.027 (−3.23)***
0.189 (0.70)
80
2196
0.070
80
2196
0.118
−0.059 (−5.12)***
−1.230 (−3.06)***
−0.220 (−0.61) −0.032 (−3.14)***
−0.285 (−1.15)
−0.605 (−2.58)**
2.789 (0.55)
14
114
0.579
−1.366 (−3.73)***
−2.706 (−0.50)
Z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
number of countries
number of obs.
pseudo R
2
−0.027 (−3.13)***
0.153 (0.56)
autocratic regime change
regime duration
−0.436 (−2.08)**
democratic regime change
−0.249 (−1.03)
***
−0.691 (−5.04)
4.352 (8.66)***
−0.494 (−0.79) −0.421 (−2.01)**
3.412 (9.29)***
−0.357 (−3.93)***
government ideology
3.365 (9.20)
***
−0.413 (−4.37)***
CS-score
constraints on executives
EFW summary index
ln(GDP)
interventionism/dis-interventionism with wide margin
interventionism/dis-interventionism with low margin
dependent variable: dummy for slowdowns
Table 6 The Probability of Slowdowns: Conditional Fixed-Effects Logistic Regressions
P. Czeglédi
The theory of interventionism as an Austrian theory of slowdowns Table 7 The Probability of Slowdowns: Conditional Fixed-Effects Logistic Regressions dependent variable: dummy for slowdown years ln(GDP)
4.352 (8.66)***
EFW summary index
−0.691 (−5.04)
3.521 (12.29)***
41.878 (3.81)***
***
−0.355 (−7.41)***
constraints on executives government ideology
−0.041 (−0.09)
CS-score
−0.003 (−0.04)
democratic regime change
−0.605 (−2.58)**
−0.162 (−0.91)
−4.45 (−1.52)
autocratic regime change
−0.220 (−0.61)
−0.681 (−2.92)
2.63 (0.96)
regime duration
−0.032 (−3.14)***
−0.051 (−6.68)***
−0.954 (−3.89) ***
2
***
pseudo R
0.070
0.087
0.438
number of obs.
2196
4078
114
number of countries
80
110
14
Z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
these definitions. The sensitivity of the results to the definition of interventionism/disinterventionism was tackled to some extent by the fact that I used two different interventionism/dis-interventionism measures above. Now I turn to the definition of slowdown years. To assess the sensitivity of the results to the definition of slowdowns I will apply the following definition to identify slowdown years in an alternative way: gt+1 3.5 percent per annum, gt,t+7−gt−7,t <−2.0 percent per annum, where the notations are the same as before. This definition is almost the one Eichengreen et al. (2011) use with the exception that their requirement for a high GDP per capita is dropped and I maintain the requirement for a reduction in the yearly growth rate. In sum, this definition requires a higher than 3.5 % 8-year average growth rate to slow down by at least 2 percentage points, but it is not necessary for it to remain below 2 % for another 8 years. This method identifies 212 slowdowns in the sample between 1957 and 2002, 128 of which are found in the same year as before. To check the sensitivity of the results I only repeat the analysis of Table 7 but I also include the regressions with the interventionist/dis-interventionist variables as determinants in column 5 and 6 in Table 8. Generally speaking, the results are not really sensitive as far as the interventionism/dis-interventionism dummies are concerned but they become more so when one looks at the institutional variables. More precisely, in this case the economic freedom index and the CS-score are not significant, though still negative. However, the coefficients of the two interventionism/dis-interventionism measures are still significant at the 1 % level, and their magnitudes are only slightly larger than in Table 6.
63
number of countries
96
3511
0.089
−0.087 (−9.89) ***
13
104
0.268
−0.646 (−3.02)
72
2205
0.039
−0.050 (−5.05)
***
−0.363 (−1.10)
−0.128 (−0.07) ***
−0.664 (−3.04)***
1.227 (4.66)
***
0.524 (0.46)
−0.096 (−1.43)
1.156 (2.38)**
16.134 (3.00)
***
−0.501 (−4.87)***
72
2205
0.042
−0.049 (−4.94)***
−0.285 (−0.86)
−0.649 (−2.97)***
1.291 (4.85)***
−0.520 (−5.36)***
Z-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. Z-values without an index mean that the coefficient is not significant even at the 10 % level
1716
pseudo R
number of obs.
0.028
−0.059 (−4.79)
regime duration
2
−0.629 (−2.48)
***
−0.121 (−0.29)
autocratic regime change
***
−0.557 (−2.81)***
−0.456 (−8.30)***
2.724 (11.14)
***
−0.590 (−2.49)**
−0.076 s(0.50)
1.462 (4.06)
***
democratic regime change
CS-score
government ideology
constraints on executives
EFW summary index
ln(GDP)
interventionism/dis-interventionism with wide margin
interventionism/dis-interventionism with low margin
dependent variable: dummy for slowdown years
Table 8 The Probability of Slowdowns: Conditional Fixed-Effects Logistic Regressions on an Alternative Slowdown Dummy
P. Czeglédi
The theory of interventionism as an Austrian theory of slowdowns
The reason for the insignificance of the EFW and the CS-score index must lie in the construction of the alternative slowdown dummy. A slowdown in this interpretation is not necessarily followed by slow growth. To put it simply, in this case the reason for a slowdown can be fast growth before a slowdown year, while in my original definition the slowdown is rather characterized by slow growth after the slowdown year. It seems that economic freedom and the attitude towards private property are more important to explain a slowdown’s transformation into a stagnation, or a decline, than they are to explain a slowdown’s emergence from a “miracle”. Another concern that can be raised regards the question as to whether the causality of the effects analyzed above run through the entrepreneurship channel. This is a question which would deserve to be addressed in a separate paper. Since entrepreneurship, in the sense I use it in this paper, “does not describe a distinct group of individuals, but rather, is an omnipresent aspect of human action” (Boettke and Coyne 2003, p. 68), it is a very difficult task to have a quantitative measure of entrepreneurship. In addition, even if one accepts some proximate measure of the entrepreneurial function of human action, cross-country measures are much less available than different institutional measures. Despite these difficulties of concept and measurement, as we saw in the introduction, there are a considerable number of papers that address the issue empirically, and estimate the effect institutional quality has on the rate of productive or non-productive entrepreneurship. They usually conclude that economic freedom promotes productive entrepreneurship and discourages non-productive entrepreneurship (as do Kreft and Sobel (2005) and Sobel (2008) for US States, and Sobel et al. (2007) for OECD countries), or that some areas of economic freedom lead to a higher level of productive entrepreneurship (Nyström for OECD countries, and Bjørnskov and Foss (2008) for another cross-country sample), although they use different measures of entrepreneurship.19 Some further confirmation of this claim is given in Table 9. Here I use a crosssection measure of entrepreneurship from the World Bank Group Entrepreneurship Snapshots (World Bank 2010), which defines entrepreneurship (ibid., p. 7) as “the activities of an individual or a group aimed at initiating economic enterprise in the formal sector under a legal form of businesses”. The measure I use as a dependent variable in Table 9 is the main output of this study, namely the number of newly registered limited liability firms per 1,000 working age individuals during a year. This measure is considered here as one possible rough proxy of productive entrepreneurship. Beside the idea that the creation of a new firm requires some creativity and innovativeness, the number of entries of new firms can also be thought of as an antithesis of nonproductive entrepreneurship, considering that lobbying to prevent new competitors from entering a market is a typical nonproductive activity. The measure is provided for the years between 2004 and 2009, which makes it possible to run some simple OLS regressions with a cross section of countries, with the 19
Kreft and Sobel (2005) use sole proprietorship, venture capital investment and patent activity, Sobel (2008) creates an index for productive and another for nonproductive entrepreneurship, Sobel et al. (2007) and Bjørnskov and Foss (2008) choose the “total entrepreneurial activity” index from the Global Entrepreneurship Monitor as their dependent variable, while Nyström (2008) focuses on self-employment
P. Czeglédi Table 9 The Institutional Determinants of Slowdowns as Determinants of Entrepreneurship – Cross Section Regressions dependent variable: average entry density, 2005–2009 constant
−18.062 (−5.02)*** −10.395 (−5.09)*** −10.026 (−2.46)*** −2.605 (−0.49)
ln(GDP)
0.766 (2.31)**
EFW summary index
2.152 (2.70)***
1.288 (5.40)***
1.524 (2.86)***
0.722 (1.15)
0.111 (2.39)**
0.144 (1.67)
0.339 (2.50)***
constraints on executives CS-score government ideology
0.099 (0.05)
R2
0.284
0.226
0.262
adjusted R2
0.267
0.209
0.218
number of obs.
84
94
37
0.200 0.080 24
Heteroskedasticity robust t-statistics are in parentheses. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. T-values without an index mean that the coefficient is not significant even at the 10 % level
dependent variable being the average of this entry measure over the years 2005 and 2009, while the dependent variables are those I identified as institutional determinants of the interventionist process. To avoid the simultaneity of the dependent and the independent variables the independent variables are measured in 2004,20 with the exception of culture, whose numbers refer to the year 2005, since, as explained above, this data is provided only every 5 years and the imprecision of interpolation in this case seems to be too high a price to pay to avoid a 1-year simultaneity. Log GDP as defined above is also added to account for the multiplier effect of previous productive entrepreneurship. The results of Table 9 show that the coefficients of the institutional measures are those that can be expected from the theory: those institutional factors the presence of which reduces the probability of slowdowns also increase the rate of entrepreneurship measured in the way that has just been explained. Government ideology, again, does not seem to matter, although the sample is small when both ideology variables are included (column 5, Table 9). 4.3 Interventionist and non-interventionist slowdowns Even if one accepts the results that interventionisms occur in step with slowdowns in economic growth, one may still not believe that interventionisms cause slowdowns. Indeed, the sample I used above includes interventionist,21 as well as non-interventionist, slowdowns. Looking at the 8-year growth rates and the difference between the pre- and post-slowdown rates does not reveal much difference between interventionist and non-interventionist slowdowns. This is shown in Table 10.
20 21
This is the same structure of cross-section data as in Klapper and Love (2010). In this subsection I define interventionism as ‘interventionism with the narrow margin’.
The theory of interventionism as an Austrian theory of slowdowns Table 10 Descriptive Statistics of the 8-Year Average Growth Rate After Slowdowns and Their Differences from the Same Growth Rate Before the Slowdowns for Interventionist and Non-Interventionist Slowdowns Variable
mean
std. dev.
min.
max.
no. of obs.
interventionist slowdowns gt,t+7
−0.005
0.034
−0.135
0.020
24
gt,t+7 −gt−7,t
−0.050
0.0311
−0.159
−0.022
24
non-interventionist slowdowns gt,t+7
−0.009
0.026
−0.121
0.020
110
gt,t+7 −gt−7,t
−0.052
0.034
−0.215
−0.020
110
Considering the limited availability of the EFW data (as compared to GDP) there are in total 134 slowdowns for which we have enough data to see whether or not they can be classified as an interventionist slowdown. Of these 134 there are 24 cases that meet the criteria of an interventionist slowdown if the definition with the narrow margin is applied
As there are non-interventionist slowdowns, too, one may develop an argument, which I will call proposition 1, which is different from that of this paper. One may propose that interventionisms can be seen as successful reactions to coming slowdowns, but since interventionism achieves its purpose, on average what is experienced is a slowdown22 which is possibly milder that it would have been without the intervention. The proposition that can be derived from the argument of this paper, which I will call proposition 2, is that interventionism is one of the causes of slowdowns. I will argue that proposition 2 is also the one which the facts support.23 These two propositions are illustrated in Fig. 1. The difference between them can be understood by what they suggest regarding the significance of the “curing effect” and the “stagnation effect”. According to proposition 1, the reduction of economic freedom is a cure because it helps avoid an even more serious slowdown. That is, there is a systematic curing effect, while there is no stagnation effect, or at least this effect is accidental. According to proposition 2, the situation is the reverse. Interventionism has a stagnation effect—with no interventionism (other things being equal) one would not see a slowdown, while the curing effect is accidental—there is no systematic difference between an interventionist slowdown and a non-interventionist one. To test these two alternative hypotheses I will run panel regressions with the independent variables in Table 7 on the difference in the 8-year growth rates after yeart and before yeart (gt,t+7 −gt−7,t). The question is whether interventionism/disinterventionism and its institutional determinants have an effect on the extent to which the growth rate changes after yeart. If interventionism has only a curing effect 22 The theoretical reasons why different kinds of interventionism – protectionism, exchange rate undervaluation, export subsidies, or the rationing of cheap credit by government – may lead to faster growth in developing countries can be derived from, for example, Acemoglu et al. (2006); Hausman and Rodrik (2003) or Rodrik (2011a, b) 23 At first sight, it seems that the results in subsection 4.1 have already falsified proposition 1 by showing that a slowdown is generally predicted by an interventionism. But proposition 1 does not say that a slowdown does not happen in the absence of interventionism; it only says that the slowdown will be smaller than it would be without an intervention. Provided that the regression results in subsection 4.1 are not seen as a description of causality, the case for proposition 1 is not excluded.
P. Czeglédi ln (GDP per capita) path without interventionism – proposition 2
“stagnation effect” “curing effect”
actual, interventionist path
path without interventionism – proposition 1
time
Fig. 1 Difference between Proposition 1 and 2: “Stagnation Effect” and “Curing Effect”.
as suggested by proposition 1, one expects interventionism and a more interventionist institutional environment to increase this difference. If it is only the stagnating effect that one assumes to exist, then one would expect them to decrease it. The results in Table 11 show that in most of the cases the coefficient of the variables are of the sign which could be expected under proposition 2. Most importantly, the interventionism/dis-interventionism variable has a significantly positive coefficient (Table 11, column 2 and 3), reflecting a negative effect of interventionism for the difference in the growth rate. Having a non-interventionist year as opposed to an interventionist one, or having a dis-interventionist year as opposed to a noninterventionist one, is predicted to increase the difference in the growth rate by one percentage point. When it comes to the institutional environment and GDP, they generally have an opposite sign to what they had in Table 7 (or in Tables 4 and 5) with the exception of the ideology of the government, which now is shown to have a statistically significant, but economically insignificant, effect. In column 3 of Table 11 for example, it is predicted that a one-point increase in the overall EFW score will increase the difference in the growth rate by 1.1 percentage points. There is no sign of interventionism having a curing effect.
5 Conclusions The literature on growth accelerations and decelerations suggests that understanding the growth process needs different theories of transition between growth regimes. I have argued that Austrian economics provides us with such a theory—the theory of interventionism. The theory of interventionism in itself is however not enough for a theory to explain the slowing down of economic development. The theory of interventionism was for this reason combined with the theory of nonproductive entrepreneurship. I put these two theories together to explain how economic freedom deteriorates, fuelling nonproductive entrepreneurship and leading to slow growth. I applied this theory to what growth economists called decelerations or slowdowns and I argued that (1) a deterioration of economic freedom must be among the causes of slowdowns and (2) those institutions that are good or bad for the interventionist process must also be good or bad for slowdowns. The conclusions and the results are somewhat against what Hausman et al. (2005) or Pritchett (2003) suggest, proposing
0.0004 (1.68) 0.148 0.001
regime duration
R2 within
R2 between 114
*
*
114
3256
0.001
0.154
0.0004 (1.67)
0.003 (0.38)
0.009 (1.80)*
−0.057 (−5.27)***
0.011 (5.92)***
114
2825
0.000
0.140
156
4875
0.014
0.173
0.001 (3.20)***
0.016 (2.95)***
0.015 (1.78)* 0.0002 (1.02)
0.007 (1.71)*
0.007 (5.52)***
0.006 (1.17)
0.011 (4.26)
***
−0.061 (−5.04)*** −0.059 (−6.75)***
56
388
0.201
0.117
0.0007 (1.42)
−0.005 (−1.06)
−0.003 (−0.34)
−0.001 (−0.53)
−0.0007 (−1.91)***
−0.094 (−3.16)***
T-statistics are in parentheses, standard errors are clustered by country. Letters in the upper index refer to significance: *** : significance at 1 %, ** : 5 %, * : 10 %. T-values without an index mean that the coefficient is not significant even at the 10 % level
number of countries
3256
0.004 (0.45)
Autocratic regime change
number of obs.
0.009 (1.87)*
−0.056 (−5.21) ***
0.010 (6.03)***
Democratic regime change
CS-score
government ideology
constraints on executives
EFW summary index
ln(GDP)
interventionism/dis-interventionism with wide margin
interventionism/dis-interventionism with low margin
dependent variable: gt,t+7 −gt−7,t
Table 11 Curing Effect Versus Stagnation Effect: Fixed-Effects Panel Regressions
The theory of interventionism as an Austrian theory of slowdowns
P. Czeglédi
that we need different, new theories to explain the ups and downs of the growth process. At least when it comes to the “downs” of the process, I identified the same factors that are usually found as determinants of long-run growth. This paper can thus be seen as affirming the proposition that “… we know more of what it takes to create an “economic miracle” than we often admit as economic scientists” (Boettke 2001, p. 251, emphases in original) or that “[g]ood policies are boringly similar: rule of law, property rights, and above all dignity and liberty for the bourgeoisie” (McCloskey 2010. p. 122). Perhaps more importantly in the present economic policy situation in the developed countries, I have found no sign that some smart intervention can lead to higher economic growth. On the contrary, and in line with the theory I presented, I have found that the interventionism defined as a significant reduction of any one of the five areas of the Economic Freedom of the World index usually goes in step with a substantial reduction of economic development. This shows that understanding how the interventionist process works and how it leads to slow growth is an important and relevant task. This paper is meant to be one small step in this understanding process. Acknowledgments This paper was supported by the János Bolyai Research Scholarship of the Hungarian Academy of Sciences. I thank two anonymous referees of this journal for their really helpful comments on numerous points of the paper and for their suggestions that substantially improved the paper, with all the remaining errors being mine.
Appendix
Table 12 Areas, Components and Sub-Components of the Economic Freedom of the World Index Area 1: Size of government: expenditures, taxes, and enterprises A General government consumption spending as a percentage of total consumption B Transfers and subsidies as a percentage of GDP C Government enterprises and investment D Top marginal tax rate i Top marginal income tax rate ii Top marginal income and payroll tax rates Area 2: Legal structure and security of property rights A Judicial independence (GCR) B Impartial courts (GCR) C Protection of property rights (GCR) D Military interference in rule of law and the political process (ICRG) E Integrity of the legal system (ICRG) F Legal enforcement of contracts (DB) G Regulatory restrictions on the sale of real property (DB) Area 3: Access to sound money A Money growth B Standard deviation of inflation C Inflation: most recent year D Freedom to own foreign currency bank accounts
The theory of interventionism as an Austrian theory of slowdowns Table 12 (continued) Area 4: Freedom to trade internationally A Taxes on international trade i Revenues from trade taxes (% of trade sector) ii Mean tariff rate iii Standard deviation of tariff rates B Regulatory trade barriers i Non-tariff trade barriers (GCR) ii Compliance cost of importing and exporting (DB) C Size of the trade sector relative to expected D Black-market exchange rates E International capital market controls i Foreign ownership/investment restrictions (GCR) ii Capital controls Area 5: Regulation of credit, labor, and business A Credit market regulations i Ownership of banks ii Foreign bank competition iii Private sector credit iv Interest rate controls/negative real interest rates B Labor market regulations i Minimum wage (DB) ii Hiring and firing regulations (GCR) iii Centralized collective bargaining (GCR) iv Mandated cost of hiring (DB) v Mandated cost of worker dismissal (DB) vi Conscription C Business regulations i Price controls ii Administrative requirements (GCR) iii Bureaucracy costs (GCR) iv Starting a business (DB) v Extra payments/bribes (GCR) vi Licensing restrictions (DB) vii Cost of tax compliance (DB) Gwartney et al. (2011, p. 5). Abbreviations in parentheses refer to source of the particular data. GCR: Global Competitiveness Report; ICRG: International Country Risk Guide, DB: Doing Business Report
P. Czeglédi Table 13 Descriptive Statistics of the Variables Used in Panel Logit Regressions Variable
mean
std. dev. min
max
No. of no. of obs. countr.
dummy for slowdowns (as defined originally)
0.104
0.306
0
1
6034
187
dummy for slowdowns (as defined for sensitivity analysis)
0.105
0.307
0
1
6034
187
interventionism/dis-interventionism with narrow margin
0.203
0.658
−1
1
3928
122
interventionism/dis-interventionism wide margin
0.211
0.686
−1
1
3928
122
EFW summary index
5.915
1.214
9.21
3318
122
3.761
2.275
159
constraints on executives CS-score
11.652 20.061
government ideology
−0.248
2.11
0.911
0
7
6936
−34.8
51.6
603
79
−1
1
3119
138
democratic regime change
0.132
0.338
0
1
6722
159
autocratic regime change
0.073
0.261
0
1
6722
159
27.388
0
194
6987
159
regime duration
22.45
ln(GDP)
8.228
1.265
4.764
11.634 7160
188
gt,t+7 −gt−7,t
0.011
0.045
−0.318
0.548 6034
187
The time span of the different data are different, as explained in the text. The widest range is between 1956 and 2003 for slowdowns and GDP per capita. See also footnote 15
Table 14 Descriptive Statistics of the Variables Used in the Cross-Section Regressions in Table 9 Variable
mean
std. dev.
average entry density, 2005–2009
min
35.618
334.739
0.004
ln(GDP) in 2004
8.905
1.207
6.143
EFW summary index in 2004
6.833
0.823
constraints on executives in 2004
5.537
1.832
government ideology in 2004
0.043 −1.040
CS-score in 2005
max 3545.8
no. of obs. 112
11.125
105
5
8.75
84
1
7
95
0.939
−1
1
70
17.485
−36.7
30
38
Table 15 Pairwise Correlations between the Institutional Explanatory Variables Used in Panel Logit Regressions EFW summary index
constraints on executives
CS-score
EFW summary index
1.000 (3318)
constraints on executives
0.478 (3065)***
1.000 (6936)
CS-score
0.339 (490)***
0.331 (577) ***
1.000 (603)
0.334 (2835)***
0.062 (487)
government ideology
***
0.190 (2039)
government ideology
1.000 (3119)
Numbers of observations are in parentheses. *** : significance at 1 %, an observation number without an index means that the coefficient is not significant even at the 10 % level
The theory of interventionism as an Austrian theory of slowdowns
References Acemoglu, D., Johnson, S., & Robinson, J. A. (2005). Institutions as the Fundamental Cause of Long-Run Growth. In P. Aghion & S. N. Durlauf (Eds.), Handbook of Economic Growth (Vol. 1A, pp. 385–472). Amsterdam: Elsevier. Acemoglu, D., Aghion, P., & Zilibotti, F. (2006). Distance to Frontier, Selection, and Economic Growth. Journal of the European Economic Association, 4(1), 37–74. Baumol, W. J. (1990). Entrepreneurship: Productive, Unproductive and Destructive. Journal of Political Economy, 98(5), 893–919. Beck, T., Clarke, G., Groff, A., Keefer, P., & Walsh, P. (2001). New Tools in Comparative Political Economy: The Database of Political Institutions. World Bank Economic Review, 15(1), 165–176. Benson, B. L. (2005). Regulation, More Regulation, Partial Deregulation, and Reregulation: the Disequilibrating Nature of a Rent-seeking Society. In P. Kurrild-Klitgaard (Ed.), The Dynamics of Intervention: Regulation and Redistribution in the Mixed Economy. Advances in Austrian Economics (Vol. 8, pp. 105–143). Bingley: Emerald Group Publishing Limited. Bjørnskov, C. (2005). Does Political Ideology Affect Economic Growth? Public Choice, 123(1), 133–146. Bjørnskov, C. (2008). The Growth–Inequality Association: Government Ideology Matters. Journal of Development Economics, 87(2), 300–308. Bjørnskov, C., & Foss, N. J. (2008). Economic Freedom and Entrepreneurial Activity: Some Cross-Country Evidence. Public Choice, 134(3–4), 307–328. Bjørnskov, C., & Paldam, M. (2012). The Spirits of Capitalism and Socialism. A Cross-country Study of Ideology. Public Choice, 150(3–4), 469–498. Boettke, P. J. (2001). Why Culture Matters? In: Calculation and Coordination: Essays on Socialism and Transitional Political Economy (pp. 248–265). New York: Routledge. Boettke, P. J. (2012). Living Economics: Yesterday, Today, and Tomorrow. Oakland: The Independent Institute in Cooperation with Universidad Francisco Marroquin. Boettke, P. J., & Coyne, C. J. (2003). Entrepreneurship and Development: Cause or Consequence? In R. Koppl, J. Birner, & P. Kurrild-Klitgaard (Eds.), Austrian Economics and Entrepreneurial Studies. Advances in Austrian Economics 6 (pp. 67–87). Bingley: Emerald Group Publishing Limited. Chauffour, J.-P. (2011). On the Relevance of Freedom and Entitlement in Development. New Empirical Evidence (1975–2007). World Bank Policy Research Working Paper, 5660. Colombatto, E. (2011). Markets, Morals and Policy-Making. A New Defence of Free-Market Economics. London: Routledge. Coyne, C. J., Sobel, R. S., & Dove, J. A. (2010). The Non-productive Entrepreneurial Process. Journal of Austrian Economics, 23(4), 333–346. De Haan, J., & Sturm, J. E. (2000). On the Relationship Between Economic Freedom and Economic Growth. European Journal of Political Economy, 16(2), 215–241. De Soto, J. H. (2009). The Ethics of Capitalism. In J. H. De Soto (Ed.), The Theory of Dynamic Efficiency (pp. 169–181). London: Routledge. Dovern, J., & Nunnenkamp, P. (2007). Aid and Growth Accelerations: An Alternative Approach to Assessing the Effectiveness of Aid. Kyklos, 60(3), 359–383. Eichengreen, B., Park, D., Shin, K. (2011). When Fast Growing Economies Slow Down: International Evidence and Implications for China. NBER WP 16919. Faria, H. J., & Montesinos, H. M. (2009). Does economic freedom cause prosperity? An IV approach. Public Choice, 141(1), 103–127. Friedman, J., & Kraus, W. (2011). Engineering the Financial Crisis. Systemic Risk and the Failure of Regulation. Philadelphia: University of Pennsylvania Press. Gorodnichenko, Y., & Roland, G. (2011). Which Dimensions of Culture Matter for Long-Run Growth? American Economic Review, 101(3), 492–498. Gwartney, J., & Lawson, R. (2003). The Concept and Measurement of Economic Freedom. European Journal of Political Economy, 19(3), 405–430. Gwartney, J., Lawson, R., Hall, J. (2011). 2011 Economic Freedom Dataset. Published in Economic Freedom of the World: 2010 Annual Report. Economic Freedom Network. http://www.freetheworld. com/2011/2011/Dataset.xls. accessed December 28, 2011. Hausman, R., & Rodrik, D. (2003). Economic Development as Self-Discovery. Journal of Development Economics, 72(2), 603–633. Hausman, R., Pritchett, L., & Rodrik, D. (2005). Growth Accelerations. Journal of Economic Growth, 10 (4), 303–329.
P. Czeglédi Hayek, F. A. (1945). The Use of Knowledge in Society. American Economic Review, 35(4), 519–530. Hayek, F. A. (1968/2002). Competition as a Discovery Procedure. Quarterly Journal of Austrian Economics, 5(3), 9–23. Heston, A., Summers, R., & Aten, B. (2011). Penn World Table Version 7.0. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. Holcombe, R. G. (1998). Entrepreneurship and Economic Growth. Quarterly Journal of Austrian Economics, 1(2), 45–62. Holcombe, R. G. (2007). Entrepreneurship and Economic Progress. New York: Routledge. Ikeda, S. (1997). Dynamics of the Mixed Economy: Toward a Theory of Interventionism. New York: Routledge. Ikeda, S. (1998). Interventionism and the Progressive Discoordination of the Mixed Economy. In P. J. Boettke & S. Ikeda (Eds.), Advances in Austrian Economics 5 (pp. 37–50). London: JAI Press. Ikeda, S. (2005). The Dynamics of Interventionism. In P. Kurrild-Klitgaard (Ed.), The Dynamics of Intervention: Regulation and Redistribution in the Mixed Economy. Advances in Austrian Economics, 8 (pp. 21–57). Bingley: Emerald Group Publishing Limited. Jerzmanowski, M. (2006). Empirics of Hills, Plateaus, Mountains, and Plains: A Markov-Switching Approach to Growth. Journal of Development Economics, 81(2), 357–385. Jerzmanowski, M. (2011). Acceleration, Stagnation, and Crisis: the Role of Policies and Institutions. Working Paper, Clemson University. Jones, B. F., & Olken, B. A. (2008). The Anatomy of Start-Stop Growth. The Review of Economics and Statistics, 9(3), 582–587. Jong-A-Pin, R., & De Haan, J. (2008). Growth Accelerations and Regime Changes: A Correction. Economic Journal Watch, 5(1), 51–58. Jong-A-Pin, R., & De Haan, J. (2011). Political Regime Change, Economic Liberalization and Growth Accelerations. Public Choice, 146(1), 93–115. Kerekes, M. (2012). Growth Miracles and Failures in a Markov Switching Classification Model of Growth. Journal of Development Economics, 98(2), 167–177. Kirzner, I. M. (1973). Competition and Entrepreneurship. Chicago: University of Chicago Press. Klapper, L., Love, I. (2010). The Impact of the Financial Crisis on New Firm Registration. World Bank Policy Research Working Paper 5444. Klein, D. B. (2005). If Government is So Villainous, How Come Government Officials Don’t Seem Like Villains? With a New Postscript. In P. Kurrild-Klitgaard (Ed.), The Dynamics of Intervention: Regulation and Redistribution in the Mixed Economy. Advances in Austrian Economics, 8 (pp. 223– 244). Bingley: Emerald Group Publishing Limited. Kreft, S. F., & Sobel, R. S. (2005). Public Policy, Entrepreneurship, and Economic Freedom. Cato Journal, 25(3), 595–616. Kurrild-Klitgaard, P. (2005). Ulysses and the Rent-seekers: The Benefits and Challenges of the Constitutional Constraints on Leviathan. In P. Kurrild-Klitgaard (Ed.), The Dynamics of Intervention: Regulation and Redistribution in the Mixed Economy. Advances in Austrian Economics, 8 (pp. 245–278). Bingley: Emerald Group Publishing Limited. Leijonhufvud, A. (1981). Costs and Consequences of Inflation. In A. Leijonhufvud (Ed.), Information and Coordination: Essays on Macroeconomic Theory (pp. 227–289). USA: Oxford University Press. Marshall, M. G., Jaggers, K., & Gurr, T. R. (2011a). Polity IV Project, Political Regime Characteristics and Transitions, 1800–2010. Center for Global Policy, School of Public Policy, George Mason University and Center for Systemic Peace. http://www.systemicpeace.org/inscr/inscr.htm accessed December 28, 2011. Marshall, M. G., Jaggers, K., & Gurr, T. R. (2011b). Polity IV Project: Dataset Users’ Manual. Center for Global Policy, School of Public Policy, George Mason University and Center for Systemic Peace. McCloskey, D. N. (2010). Bourgeois Dignity. Why Economics Can’t Explain the Modern World. Chicago: The University of Chicago Press. Miller, T., Holmes, K. R., & Feulner, E. J. (2012). 2012 Index of Economic Freedom. Washington and New York: The Heritage Foundation and Dow Jones & Company, Inc. Mises, L. von (1920/1990). Economic Calculation in the Socialist Commonwealth. Auburn: The Ludwig von Mises Institute. Mises, L. von (1927/2000). Liberalism In The Classical Tradition. Online edition. Auburn: The Ludwig von Mises Institute. Mises, L. von (1929/1996]). A Critique of Interventionism. Irvington-on-Hudson, New York: Foundation for Economic Education.
The theory of interventionism as an Austrian theory of slowdowns Mises, L. von (1940/1998). Interventionism: An Economic Analysis. Irvington-on-Hudson, New York: Foundation for Economic Education. Mises. L. von (1949/1998). Human Action. A Treatise On Economics. The Scholar’s Edition. Auburn: The Ludwig von Mises Institute. Nyström, K. (2008). The Institutions of Economic Freedom and Entrepreneurship: Evidence from Panel Data. Public Choice, 136(3–4), 269–282. Olson, M. (1982). The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. New Haven: Yale University Press. Paldam, M. (2003). Economic Freedom and the Success of the Asian Tigers: an Essay on Controversy. European Journal of Political Economy, 19(3), 453–477. Paldam, M., & Gundlach, E. (2008). Two Views on Institutions and Development: The Grand Transition vs. the Primacy of Institutions. Kyklos, 61(1), 65–100. Pritchett, L. (2000). Understanding Patterns of Economic Growth: Searching for Hills among Plateaus, Mountains, and Plains. World Bank Economic Review, 14(2), 221–250. Pritchett, L. (2003). A Toy Collection, a Socialist Star and a Democratic Dud? In D. Rodrik (Ed.), In Search of Prosperity: Analytic Narratives on Economic Growth (pp. 123–151). Princeton University Press: Princeton. Rode, M., & Coll, S. (2012). Economic Freedom and Growth. Which Policies Matter the Most? Constitutional Political Economy, 23(2), 95–133. Rodrik, D. (2011a). Unconditional Convergence. NBER Working Paper No. 17546. Rodrik, D. (2011b). The Future of Economic Convergence. NBER Working Paper No. 17400. Rubin, P. H. (2001). Ideology. In W. F. Shughart II & L. Razzaloni (Eds.), The Elgar Companion to Public Choice (pp. 328–336). Cheltenham: Edward Elgar. Simmons, R. T., Yonk, R. M., & Thomas, D. W. (2011). Bootleggers, Baptists, and Political Entrepreneurs. Key Players in the Rational Game and Morality Play of Regulatory Politics. Independent Review, 15 (3), 367–381. Sobel, R. S. (2008). Testing Baumol: Institutional Quality and the Productivity of Entrepreneurship. Journal of Business Venturing, 23(6), 641–655. Sobel, R. S., Clark, J. R., & Lee, D. R. (2007). Freedom, Barriers to Entry, Entrepreneurship, and Economic Progress. Review of Austrian Economics, 20(4), 221–236. Summers, R., & Heston, A. (1991). The Penn World Table (Mark 5). An Expanded Set of International Comparisons, 1950–1988. Quarterly Journal of Economics, 106(2), 327–368. Tullock, G. (1975). The Transitional Gains Trap. Bell Journal of Economics, 6(2), 671–678. Wooldridge, J. M. (2002). Econometric Analysis of Cross section and Panel Data. Cambridge: The MIT Press. World Bank (2010). Entrepreneurship Snapshots 2010. Measuring the Impact of the Financial Crisis on New Business Registration. Washington, D.C.: The World Bank. http://econ.worldbank.org/WBSITE/ EXTERNAL/EXTDEC/EXTRESEARCH/EXTPROGRAMS/EXTFINRES/0,,contentMDK:21454009~ pagePK:64168182~piPK:64168060~theSitePK:478060,00.html accessed July 10, 2012. Yandle, B. (1983). Bootleggers and Baptists—the Education of a Regulatory Economist. Regulation, May/June, 12–16.