Jan. 10, 1995


1995

THEORIES OF INSTITUTIONAL AND TECHNOLOGICAL INTERDEPENDENCE

by A. Allan Schmid

Michigan State University

Two bodies of theory have arisen to explain institutional change in agriculture in the context of technological change. One of these derives from neoclassical economics and the other from neoinstitutional economics. While there are variations within each school of thought, the neoclassical version is represented by Schultz and North and Thomas, but is best known currently in the work of Hayami and Ruttan (1985) under the label of "induced institutional change". The neoinstitutional version is found in Field, Hirschman, Myrdal, and a later work of North among others.

The neoclassical model of institutional change utilizes only the following variables; relative prices influenced by factor endowments, consumer preferences, and production functions (technology). The neoinstitutional model utilizes these variables plus such things as ideology, learning, and power. The attractiveness of the neoclassical model is fewer variables and variables which are easier to measure and quantify. This facilitates mathematical and statistical analysis instead of qualitative stories. But, efficiency in modeling is like any other, saving inputs is only relative to output, and the question is whether the model can explain and predict. In addition, some models while ostensibly for the purpose of prediction are primarily for the purpose of legitimation. That is to say, they are an argument for what is better to do.

This article will explore the two types of theories with respect to prediction and claimed legitimation of policy using new irrigation and crop variety technologies in the Philippines as a case in point. It will argue that there is a logical flaw in the neoclassical argument because there is no unique equilibrium associated with any institutional change. Rather there are many possible depending on whose interests count. There is nothing inefficient about any group of people being made owners of a large or small share of the gains from technological change. It will further be argued that ideology and culture have as much to do with the demand for institutions as with their supply.

Neoclassical Theory

Hayami and Ruttan regard institutional change as endogenous to the economic system modeled in terms of demand and supply. The demand for change results from exogenous change in factor prices influenced by technology and factor endowments (such as population changes). (1985, 95) Exogenous change causes disequilibrium between marginal value and marginal costs of labor and land, and institutions change to restore equilibrium again. The supply of institutional change is affected by the power structure, cultural tradition, ideology and education (96).

There are several corollaries. (1) If institutions don't change, they become obstacles to efficiency and economic growth. (2) Since institutional change is endogenous, political entrepreneurs are like any firm responding to consumer demand and have no independent preferences of their own (107). (Also see Downs, for an explicit development of this view.) (3) Culture and ideology don't affect the initiation and path of change (demand), only its speed and cost (supply). From the induced institutional change perspective, "land reform should not be viewed as a precondition for productivity growth. Rather, it should be viewed as induced by opportunities created by technical change and market development." (396) They further state that "ideological approaches to land tenure reform have often imposed great costs on both landowners and tenants with few gains in agricultural production or productivity." (398) Let us re-examine the evidence and logic for these claims.

The Philippine Case

Hayami and Ruttan (99-103) illustrate their theory with the history of a rice farming village in the Philippines between 1956 and 1976. The relevant parties are landlords, tenants, and landless laborers. With respect to tenants, the traditional system was share tenure. With respect to landless labor, the traditional property institution (hunusan) gave one-sixth of the rice to landless laborers of the village who participated in the harvest and threshing. In 1958, a publicly built irrigation system was extended to the village, and in the 1960's, publicly developed high-yielding rice varieties were available accompanied by fertilizer, pesticides and improved cultural practices. The effect of these new technologies was to raise farm income, and given the traditional rights, tenants and other laborers got some of this gain as well as the landlords. In the view of Hayami and Ruttan this created a disequilibrium and labor received more than its opportunity cost in other occupations, which was declining because of population growth.

Hayami and Ruttan argue that this situation induced a productive institutional change, namely a system called gamma where landless labor could get a share of the harvest only by giving otherwise uncompensated weeding labor. This lowered the marginal cost of labor to its opportunity cost. The resulting performance gave none of the benefits of technological change to the landless laborers.

Since the landless laborers received only a share and not an hourly wage, any putative restoration of equilibrium must take place via an institutional change. While the technological change may be the occasion for a struggle for the technological rents (rent seeking), the particular form it takes must be explained. The question is whether the institutional change is to be a change in harvest share, the amount of work, wage labor, or a grant of ownership in the residual rent. Any of these is consistent with marginal value product (MVP) of labor equal to cost.

For a theory to be predictive, it must identify a unique outcome among the several possible. Hayami and Ruttan (101) say that the "institutional innovation ... permitted farm operators to equate the harvesters' share of output to the marginal productivity of labor". But harvesters' share is not necessarily the same as the marginal cost of labor. The right of a part-owner to share in net returns is not a marginal cost. If the landlord is the sole decision maker and claimant on net firm income, then and only then is labor's share also a marginal cost to the landlord. But, if the present laborers are regarded as part owners (like any shareholder in a corporation), then their claim on net income is not a marginal cost -- only the wage payment is a marginal cost, not their "dividends". In other words, there can be an equilibrium of marginal cost and revenue with different distributions of ownership and "shares" to different parties. Equilibria between marginal cost and marginal revenue are not unique to one ownership system and thus cannot explain change in ownership. The particular equilibrium among many possible is rights dependent and can't explain rights change.

This point can be illustrated with events in the Philippine case. In 1963, a land reform act was passed which allowed tenants to shift from a traditional crop share to specified cash rents which were lower. Many tenants were quick to take advantage of this opportunity (Otsuka). Hayami and Ruttan (102) say the land reform gave tenants "the right to continue tilling the soil at a rent lower than the marginal product of land". This is a matter of selective perception. As noted above, an alternative perception is that the land reform made tenants a claimant on output which was not a function of the amount of tenant labor. Thus the tenant's decision to work on the land is still a matter of equating their marginal value product (MVP) and their opportunity cost considering the cost of landless laborers. With a greater endowment, the tenants may invest in their own human capital (which changes their MVP) or choose leisure. To say that landlords would have preferred a law making them the exclusive owner of rent residuals does not legitimate their right. If they had that right, then the tenant's labor is a marginal input, and if the price increases less will be employed.

The uniqueness of any particular equilibrium and its use in policy analysis is further brought into question by noting that the marginal product of the publicly provided irrigation and rice varieties was substantial. If the water and varieties had been fully priced, there would have been no rent to land. If the government revenues then had been remitted to tenants or landless laborers, there could be no objection that this upset the labor MVP-wage equilibrium.

In fact, the land reform act gave the tenants a share of the net gain from technological change. And this improvement in tenant income and the non-improvement of the income of landless laborers are both consistent with wage equilibrium. Thus, the establishment of equilibrium cannot explain either the land reform act (rent control) which benefited tenants or the development of the gamma system which reduced returns to landless labor.

Hayami and Ruttan emphasize that while the land reform act prohibited subtenancy, subtenancy became common as tenants assessed their opportunity costs (including value of leisure) and subleased the land at a rate that gave the landless laborers only low non-farm wages. Hayami and Ruttan regard this lawless subtenancy as proof that institutions change to restore equilibria. But, the particular equilibrium is not unique and would have occurred whether or not the rights had allowed the tenant to share in the gains of technological change.

The land reform law in effect made tenants part owners of land (or the associated technolgy if you will) and their income increased above that of landless laborers in rural or urban areas. Previously the income of both tenants and landless laborers was equivalent to the going non-farm wage and after only the landless laborers relied on the non-farm wage. The marginal labor unit is paid the same under the two sets of rights, but the landlord got less with land reform because the tenant got more. Ledesma (30) reports that tenants prepared the seedbed and took care of fertilizing, spraying and water control. Landless labor constituted 50 percent of total labor time and provided the transplanting, weeding, harvesting and threshing in exchange for one-sixth of the crop.

Hayami and Ruttan do not inquire into the reasons for the land reform act except that it may have been an attempt by the urban elite to reduce the power of the rural elite. Since ideology is to be examined only on the supply side, it is expected that any notions of social justice have little to do with the demand for institutional change. Since the rent control law interfered with efficient market determination, it is implied that it is an aberration which will be corrected in time. Hayami and Ruttan regard the subvention of subtenancy prohibition as evidence supporting their view. But, this outcome is efficient only if third party interests are not rightful.

"Land to the tiller" was part of the legitimation for shifting rights and income between landlord and tenants, i.e. the fixed rent law. In 1963, President Macapagal "saw land reform as a critical countermeasure against communist infiltration in the rural sector, a danger compounded by the growing tension in Southeast Asia with the spread of the Vietnam War." (Adriano, 46) Policy makers saw the shift from share to cash rent as the first step toward transferring landownership to tenants. The dominant political group may have wanted the gains of public investment in new technology to go to those they regarded as earning it or capable of bringing peace to the countryside, rather than to only substitute one rent collector for another.

But wishes are not neccessarily results. Hayami and Ruttan argue that prohibition of subtenancy is a non-Pareto-better disequilibrium requiring tenants to farm when they have something better to do. But, if legitimation and someone's desire to live in a society of owner-operators is rightful, then it can't be excluded from the marginal calculus. It is then part of the utility and products to be maximized and not a barrier to the realization of efficiency. Could anyone compensate those who derive utility from living in an environment of owner-operators if subtenancy occurs? It is value presumptive to assume that subtenancy prohibition prevents a Pareto-better equilibrium rather than defining a new one. There is an equilibrium either way, though not the same one.

The most that can be said about the prevalence of subtenancy is that its prohibition is difficult to enforce, not because of the asserted inefficiency, but because of high policing costs. When tenants are given control over the lease, they have the same incentives as the landlord did to use low cost labor. The tenants' new rights may have been obtained under the ideology of land to the tiller, but the tiller will in turn take advantage of cheap labor nevertheless. There is no reason to further legitimate this result by calling it Pareto-better trade by selectivly eliminating the desire of law givers from the trade. The disobedience of sub-tenancy prohibitions is not necessarily Pareto-better any more than selling my neighbors's car is Pareto-better. It all depends on who is truly the rightful owner.

Hayami and Ruttan assert that disobedience of the subtenancy prohibition restores equilibrium. Some tenants were also displaced by landlords which would also restore equilibrium (Otsuka 341). Is the glass half empty or half full? Is it remarkable that some broke the law to their advantage or that some did not? Perhaps if the ideological rationale for the fixed rent to operating tenants had been made explicit there would have been more self-control and community pressure on tenants to continue farming even if opportunistic subterfuge were readily available.

Neoclassical theory with its utilitarian foundations has no general suggestion as to what factors affect disadvantageous rule obedience except fear of punishment. J.R. Commons used the term "forbearance" to refer to behavior where people behaved according to rule even when there was the possibility of profitable opportunism. Alexander Field (186) says that "there is no reason to assume that there are unique overhead (enforcement) costs associated with every set of rules, overhead costs independent of the length of time that set has prevailed, and, if the individual rule is part of an overall set of rules, the permanence of other elements of the institutional structure." Field observes that obedience to rules depends on whether people expect them to continue. This creates a bit of dilemma for the neoclassical theory since change in institutions is necessary to restore equilibrium, but change makes it harder for the new rule to be obeyed and have the desired effect.

It can also be observed that people play by the rules if they regard them as fair. Thus, a theory which ignores distribution and just satisfies itself with potential Pareto improvement from any initial rights assumption can never fully explain institutional change. If people are not accepting of the actual distribution, the productivity gain may not be realized, as destructive civil wars and insurgency are witness. Without specification of forbearance, a model of institutional change cannot be closed.

Welfare Economics, Politics and Power

Even if models do not predict, they are often used to legitimate institutional change. Hayami and Ruttan (102) say, "It is important to recognize that subtenancy and gamma contracts were the institutional innovations to facilitate more efficient resource allocations through voluntary agreements by assigning more complete private property rights." How can these changes be regarded as "voluntary"? The landless laborer who must weed for free was not asked to sell his traditional rights. When faced with a choice between starvation and low wages equivalent to off-farm work, the laborer's choice is clear, but it in no way says that the changed agenda for choice was voluntarily arrived at. It is similar to the thug who gives you the choice between your wallet or your life. You make a choice, but your rights are violated. (For a discussion of this point in the context of yellow dog labor contracts see Michael Carter.) Depriving labor of its traditional hunusan rights was not Pareto-better. Neither was the deprivation of those who derived utility from living in a society of owner-operators when the subtenancy prohibitions were violated. What Hayami and Ruttan call a "more complete specification of property rights" is equivalent to saying that those who derive such utility are not to count.

An adequate history of Philippine technological/institutional change would have to document the struggle of landless laborers to maintain the traditional hunusan system and the power of the landlords to constrain any such collective action. While the urban elite may have had the political power to pass a land reform act to reduce the power of the rural aristocracy, it would be necessary to document who controlled the local officials and possibly blunt any effort by laborers to protest the gamma system or tenants to resist displacement by landlords (Adriano 50).

In addition to power variables, a theory of institutional change has to account for the formation of ideology. If people can be convinced that the trade-off is between efficiency (productivity) and distribution rather than different claims on the gains from technological change each with its own efficient equilibrium, the ensuing institutional change will be different. Consider the following alternatives for sharing the gain from technological change: (1) Complete dispossession of tenants. For example, in Java, the landlord has sold the standing crop to a middleman who then hires labor (Collier et. al.; Hayami and Hafid). (2) The Philippine case of tenants with leaseholds at a controlled price plus subtenancy. An alternative would be to increase the traditional tenant's share of the crop as had been done earlier by various efforts of Magsaysay and the Tenancy Act of 1954 (Akira 76-7). (3) The landless are declared part owners and receive a share of the crop regardless of labor participation, just as any capitalist owner. Or, landless workers could have been made the principle beneficiaries of land redistribution as was the case in China in the early 1950's (Ledesma) and was the objective of President Macapagal in the Philippines. (4) Government taxes the land rent and returns it to the landless laborers. This need not be a transfer payment, but rather the government acting as agent for the laborers as part-owners. (5) Government sells the new technologies and distributes the net proceeds to various groups. (6) Lower output prices to consumers. Herdt observes that in the Philippines from 1966 to 1981 "farm operators and hired laborers have both retained some portion of the benefits of technical change, but consumers have reaped the bulk of the benefits through lower real rice prices (348)."

All of the above are consistent with efficient labor wage equilibrium. A wholly endogenous theory fails to predict which of these outcomes occurs. Further, the MVP of land and labor is not alone a matter of production functions, demand, and endowments. Marginal productivity is itself a function of institutions including the pricing of publicly supplied inputs. Irrigation and new varieities are themselves the product of institutional innovation, not just spontaneous creators of new profit opportunities. Technology and land reform were both institutional responses to rural protest and urban insurgency. Institutions do not just follow endogenous profit-making opportunity; they help determine what is profitable.

An adequate theory of institutional change would have to explain why the landless laborers did not revolt more than they did when their traditional rights were denied. This occured in other countries. Clifton Wharton, Jr. reports that 43 persons were killed in India in December, 1968, in a clash between landlords and landless workers "who felt that they were not receiving their proper share of the increased prosperity brought by the Green Revolution". Perhaps there would have been more protest if the laborers regarded their traditional rights as being equivalent to part ownership of the fruits of technological change. Somehow the landlords' claim on these public investments seems tied naturally to the ownership of land. The very concept of induced institutional change with its focus on factor endowments and technology is part of the impression that it is natural forces that result in landlords receiving the lion's share. What is seen as natural is a powerful tool in the manipulation of symbols. A law making laborers explicit part-owners of some of the gain from technology would probably seem more revolutionary than the loss of their traditional rights. Such is the power of ideology. It is ironic that Pareto is not better remembered for his insights into the manipulation of symbols to shape political debate rather than for the Pareto-better logic which draws our attention away from the institutional base which determines what each person has to trade (see Samuels).

Feudalism

As noted in the introduction, there are other applications of the endogenous neoclassical theory of institutional change. Lest the above appear as a special case, a brief review of the North and Thomas explanation of the decline of feudalism in Europe and the critique of Alexander Field will be useful. North and Thomas give no great causal role for technology as do Hayami and Ruttan, but emphasize population change. From the 11th to the 13th Century, population grew and agricultural output also grew both due to the cessation of warfare and raiding. Field (180) observes that this was a political development which North and Thomas do not explain. North and Thomas argue that growth in population and spread of the market reduced contractual enforcement costs and this made the commutation of direct labor services to cash rents efficient.

But, during the 13th Century there was a reversion to direct labor services for a period. North and Thomas explain this by arguing that the decline in the land/labor ratio increased the value of land relative to labor. This is similar to the Hayami and Ruttan case of wage disequilibrium. But again, several alternative institutional changes are possible and a good theory must chose among them. An alternative way to reduce wages would have been to reduce the fixed cash rent. North and Thomas argue that direct labor services were reinstituted rather than lowering the cash rent because "a complete break with tradition, the governing law of the manor...to avoid costly complications, the alteration in payment had to be set within the framework of existing customs (60)."

This is a big hole in an endogenous theory of institutional change. As Field (190) puts it, the theory becomes "institutions change in response to underlying parameter shifts, except when they do not, that is, except when conservative principles prevail. This is, to put it mildly, a less general model." Some variables on ideological and power changes are needed to close the model. In Hayami and Ruttan, Philippine landless labor could have been excluded from sharing in technological gains by the gamma system, sale of the standing crop to middlemen, or straight-forward rejection of the rights of landless labor to participate in the harvest as anything but wage labor. Since these all result in wage equilibrium, some other explanatory variables such as ideological acceptability are needed.

Field (190) also observes that two different institutional changes occurred in western and eastern Europe when the explanatory variables of the North and Thomas model were similar. Both areas experienced population decline in the 14th and 15th Centuries, but feudalism continued its decline in Western Europe while reenserfment occurred in Eastern Europe. The eastern lords were able to collude and defeat the rising scarcity value of labor. An adequate theory of institutional change would have to explain why some types of collective action are successful in one area and not another.

Neoinstitutional Theory

Douglass North in a later work (1981) provides both a critique of his and neoclassical theory and a guideline to the needed model specification. He says, "The simple fact is that a dynamic theory of institutional change limited to the strictly neoclassical constraint of individualistic, rational purposeful activity would never allow us to explain most secular change ranging from the stubborn struggle of the Jews in antiquity to the passage of the Social Security Act in 1935. Secular economic change has occurred not only because of the changing relative prices stressed in neoclassical models but also because of evolving ideological perspectives that have led individuals and groups to have contrasting views of the fairness of their situation and to act upon those views" (1981, 58). North (7-8) argues that an adequate theory of institutional change would contain "a theory of property rights that describes the individual and group incentives in the system"; "a theory of the state, since it is the state that specifies and enforces property rights"; and "a theory of ideology that explains how different perceptions of reality affect the reaction of individuals to the changing 'objective' situation". These themes are repeated in North's subsequent work (1990).

For the Philippine case, Hayami and Ruttan (99) in a footnote provide an explanation which reflects North above. They say that the Land Reform Code of 1963 could be interpreted as as effort by the emerging urban elite to break the power of the landed elite and to enhance food supplies for urban growth. In a single sentence Hayami and Ruttan assert this is just an extension of their neoclassical theory since all of this contributes to economic growth. But, surely, there is conflict over the content and distribution of that economic growth which divide the new and old elites. The content of the outcome can't be predicted without some attention to the political rules of the Philippines (a theory of the state) and to the manipulation of symbols (ideology), i.e. what rights are seen as productive and natural.

Reference has already been made above in the reference to Alexander Field reflecting a major institutional theme of cumulative causation (also see Myrdal and Nelson and Winter). An adequate theory must include some attention to the sequence of institutional change (path dependence). What is possible is influenced by what institutions came before and the environment of other institutions. (North, 1990, 76)

It is not possible here to detail all of the variables which have been identified in the neoinstitutional literature, but one more element must be noted. An adequate theory of institutional change (both demand and supply) must include a theory of collective action in the face of high exclusion cost goods. In neoclassical economics most collective action is bad since the examples are drawn from collusions leading to monopoly. Collective action seems to be the natural result of greed and rent seeking and no further attention is given to it except to explore ways to defeat it. Since it is bad, there is no need to explore how to get more of it where it is lacking. But, in explaining why landless labor did not revolt more than it did or why the lords in Eastern Europe successfully colluded to keep their serfs, one needs better answers. One approach is that of Albert Hirschman who has observed a number of development efforts in Latin America. He emphasizes that experience with collective action is cumulative. Even past failures contribute to future success as people learn to trust one another and overcome the temptation of be free riders.

In conclusion, attempts to make institutions endogenous have failed to predict. It cannot explain such things as the passage of the land reform act in the Philippines. While it explains the subvention of subtenancy prohibition, it cannot explain why some tenants were allowed to stay on the land at reduced rents and not entirely displaced. Neither can it explain why the returns to landless labor were reduced by involuntary addition of work requirements (weeding) and not a reduction in share of the harvest, or other methods. The neoclassical model variables of factor endowments and technology (relative prices) in the context of simple profit maximization need to be extended to include neoinstitutional variables of history, power (including rules of political choice), and ideology (including the learning of forbearance). These variables must be added to explanations of both the demand and supply of institutional change in the context of technological change.

References

Adriano, Lourdes S. "Agrarian Reform and Rural Development," in Philippine Rural Development: Problems, Issues and Directions, Los Banos: University of the Philippines, 1991.

Akira, Takahashi. Land and Peasants in Central Luzon, Tokyo: Institute of Developing Economies, Occasional Paper No. 4, 1969.

Balisacan, Arsenio M. "Agricultural Growth, Landlessness, Off-Farm Employment and Rural Poverty in the Philippines," Economic Development and Cultural Change, 41(1993):533-62.

Carter, Michael. "A Wisconsin Institutionalist Perspective on Microeconomic Theory of Institutions: The Insufficiency of Pareto Efficiency." Journal of Economic Issues 19(1985):797-813.

Collier, William et. al. "Agricultural Technology and Institutional Change in Java," Food Research Institute Studies 13(1974): 169-94.

Commons, John R. Institutional Economics. New York: Macmillan, 1934.

Downs, Anthony. An Economic Theory of Democracy. New York: Harper and Row, 1957.

Ezkekiel, Mordecai. "Distribution of Gains from Rising Technical Efficiency in Progressing Economies," American Economic Review 47(1957):361-75.

Field, Alexander J. "The Problem with Neoclassical Institutional Economics: A Critique with Special Reference to the North/Thomas Model of Pre-1500 Europe," Explorations In Economic History 18(1981):174-198.

Hayami, Yujiro and Anwar Hafid. Some Observations On The Changes In Rice Harvesting Systems In Java. Working Paper No. 1, Agro-Economic Survey, Bogor, Indonesia.

Hayami, Yujiro and Vernon W. Ruttan. Agricultural Development. 2nd ed. Baltimore: Johns Hopkins University Press, 1985.

Herdt, Robert. "A Retrospective View of Technological and Other Changes in Philippine Rice Farming, 1965-82," Economic Development and Cultural Change, 35(1087):329-49.

Hirschman, Albert. Getting Along Collectively: Grassroots Experience in Latin America. New York: Pergamon Press, 1984.

Ledesma, Antonio J. Landless Workers and Rice Farmers: Peasant Subclasses Under Agrarian Reform In Two Phillipine Villages. Manilla: International Rice Research Institute, 1982.

Myrdal, Gunnar. "What Is Development?" Journal of Economics Issues 8(1974):72936.

Nelson, R.R. and S.G. Winter. "Neoclassical vs. Evolutionary Theories of Economic Growth: Critique." Economic Journal 84(1974): 886-905.

North, Douglass. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press, 1990.

North, Douglass. Structure and Change in Economic History. New York: Norton 1981.

North, Douglass, and R.P. Thomas. The Rise of the Western World. Cambridge: Cambridge University Press, 1973.

Otsuka, Keijiro. "Determinants and Consequences of Land Reform Implementation in the Philippines," Journal of Development Economics, 35(19991):339-55.

Samuels, Warren J. Pareto on Policy. Amsterdam: Elsevier, 1974.

Schultz, T.W. "Institutions and the Rising Economic Value of Man." American Journal Agricultural Economics 50(1968):1113-22.

Thirtle, C.G. "The Microeconomic Approach to Induced Innovation: A Reformulation of the Hayami and Ruttan Model," Manchester School of Economic and Social Structure, 53(1985):3.

Wharton, Clifton Jr. "The Green Revolution: Cornucopia or Pandora's Box." Foreign Affairs 47(1969): 464-76.