The Soft Budget Constraint(1)

Janos Kornai

In many segments of contemporary economies a remarkable trend can be discerned: the budget constraints of economic units become "soft." The phenomenon appears in mixed economies and is conspicuously apparent in socialist systems. The "soft budget constraint syndrome" is usually associated with the paternalistic role of the state towards economic organizations and private firms, non-profit institutions and households.

I. Conceptual Clarifications

The term budget constraint is of course taken over from microtheory of the household. The assumption that the decision maker has a budget constraint is equivalent to the assumption that Say's(2) principle prevails. In agreement with Clower(3) the budget constraint is not a bookkeeping identity nor a technical relation, but a rational planning postulate. Two important properties must be underlined. First, the softening of the budget constraint occurs when the strict relationship between expenditure and earnings has been relaxed because excess of expenditure over earnings will be paid by some other institution typically by the State. A further condition of softening is that the decision maker expects such external financial assistance with high probability, and this probability is built firmly into his behavior. . . .

There are different ways and means to soften the budget constraint of the firm.

1. Soft subsidies granted by national or local governments. The subsidy is soft if it is negotiable, subject to bargaining, lobbying, etc. The subsidy is adjusted to past, present or future cost overruns.

2. Soft Taxation. The attribute soft does not refer to the rate of taxation. Even with a low tax rate the taxation system can be hard, if rules are uniform, fixed for a long period and the payment of taxes rigorously enforced. In contrast taxation is soft, even with a high tax rate, if the rules are negotiable, subject to bargaining, political pressures. The tax rates are not uniform, but almost tailor-made according to the financial situation of different sectors or different regions or different forms of ownership. The fulfillment of tax obligations is not strict; there are leaks, ad hoc exemptions, postponements, etc.

3. Soft Credit. Again softness does not refer to the magnitude of the interest rate. The credit system can be hard even with a low interest rate (provided that the credit market generates a low rate), if the fulfillment of credit contracts is strictly enforced. The creditor lends money expecting discipline in debt service and not for the sake of assistance to an ailing firm which will not be able to service its debt. Enforcement of the credit contract continues to the bitter end; harsh sanctions in the case of insolvency, including receivership, bankruptcy, forced merger, sell-out or other similar legal means. In contrast, the credit system can be soft even with high interest rates, if the fulfillment of a credit contract is not enforced, unreliable debt service is tolerated and postponement and rescheduling are in order. Soft credit is used to assist firms in great and chronic financial trouble, without real hope of repayment of the debt.

4. Soft administrative prices. This can be applied in the case when the price is not set by a free contract seller and buyer, but by some bureaucratic institution. The administrative price is hard if, once set, it restricts expenditure and does not automatically adjust to cost increases. An administrative price is soft if it is set according to some permissive "cost plus" principle that automatically adjusts prices to costs.

These four means of softening the budget constraint are not mutually exclusive; they can be applied simultaneously or successively. The list is not exhaustive, there are other means as well. . . .

All four means of softening the budget constraint of the firm refer to dynamic processes; assistance fills up the gap between the flow of expenditures and the flow of sales-generated revenues of the firm. It is meaningless to talk of the hardness or the softness of the budget constraint of one individual firm, looking at the history of that firm. As mentioned in the general definitions, the subjective probability distribution of external assistance will depend on the collective experience. The decisive question in this respect is this: what was the regular experience of a large number of firms over a longer period in the past? And can it be expected that similar experiences will occur in the future?

"Hard" and "soft" are two extreme positions on a scale of stringency. In a deterministic maximizing model an upper constraint either holds or does not hold. But here we are facing a stochastic problem: subjective expectations concerning external assistance and the enforcement of financial discipline. Therefore, intermediate positions between a strictly binding and a totally redundant constraint may exist. Consider the speed limit on highways. Some people will observe it, some others not, exceeding the permitted limit more or less frequently, to a larger or smaller extent. The distribution of violations will depend on the enforcement of the limit. But even with soft enforcement, the mere fact that there is a limit may have some influence on speed. That is, the constraint is not completely redundant.

There is one more reason to think in terms of a stringency scale rather than in a "yes" or "no" framework, in which a completely binding or a completely ineffective constraint are mutually exclusive. External assistance is usually not granted automatically, as some effort is needed to obtain it. The firm's managers (and in the case of the private firm, also the owners) must resort to political pressure groups and lobbies, or to personal connections. Explicit bribery may be frequent or rare with experience varying from country to country. Some hidden corruption in [the] form of reciprocal favors is more wide-spread. In any case, rent seeking(4) and budget constraint-softening is not without costs. Therefore, even if it might be softened, the budget constraint has at least some influence on the behavior of the firm or of other microunits.

Hardness of the budget constraint is not a synonym for profit maximization. A profit maximizing firm, even if it is in the red will try to cut its losses. A hard budget constraint means that even if the firm tries hard to cut its losses, the environment will not tolerate a protracted deficit. The emphasis is on punishment. The budget constraint is hard if persistent loss is a matter of life and death; the more the loss maker is spared from tragic consequences, the softer is the constraint. What is really important is the psychological effect of the constraint; with a hard budget constraint, a deficit causes fear because it may lead to extremely serious consequences. Profit maximization refers to the internal goal-setting of the decision maker of the firm; the softness-hardness of the budget constraint refers to the external tolerance-limits to losses.

It follows from this line of reasoning that the stringency of the budget constraint is not simply a financial matter. It reflects in financial form a deeper socio-economic phenomenon. Using a Marxian term: it reflects a certain social relationship between the state and the economic micro-organization. Clower and Due(5) wrote about Say's principle that it "constitutes an implicit definition of the transactor as distinguished from the concept of thief or a philanthropist." In the case of the soft budget constraint, the State and the firm are neither merely transactors nor is the firm a thief and the State a philanthropist. We are faced with a new kind of relationship. Different analogies come to mind: the State as a protective father and the firm as a child, the State as patron and the firm as client, the State as insurance company and the firm as the insured party. The soft budget constraint syndrome is the manifestation of the paternalistic role of the modern state.

The economic theory of the market concentrates on the horizontal relationship between seller and market. The sociological theory of bureaucracy, from its beginning with Max Weber up to now, studies the vertical relationship of superiors and subordinates within a hierarchy. The firm with a soft budget constraint is an issue at the intersection of these two disciplines. Our firm has a horizontal relationship with his customers and suppliers, and at the same time a very special relationship with the State. . . .

II. The Impact on the Firm's Conduct

The trend toward the softening of the budget constraints has many interrelated consequences. Here only three of them will be surveyed: the impact on price responsiveness, on efficiency and on the creation of excess demand. As in the second part of the previous section we still focus on the behavior of the firm.

The first issue is the effect of prices on the decision-making of the firm. The trivial case of a downward sloping demand curve by the firm for its inputs presupposes the existence of a hard budget constraint. The softer the budget constraint, the weaker the compulsion to adjust demand to relative prices. In the extreme position of a perfectly soft budget constraint, the own-price elasticity of demand is zero, the demand curve is vertical i.e. determined by other explanatory variables and not by the price. . . . [T]he exact slope of the original budget line does not matter too much if cost increases can be easily compensated by external assistance, so that the budget line is replaced by a fuzzy strip.

The softness of the budget constraint decreases the elasticity of demand of all alternative inputs, of all factors; diminishes the firm's sensitivity toward the interest rate, exchange rate and so on. Similarly, the multiproduct firm will be less sensitive to changes in relative output prices. Summing up: the general price responsiveness of the firm declines. . . .

A second issue worthy of attention is the impact on efficiency of the trend toward a softer budget constraint. Allocative efficiency cannot be achieved when input-output combinations do not adjust to price signals. Within the firm there is not a sufficiently strong stimulus to maximum effort; weaker performance is tolerated. The attention of the firm's leaders is distracted from the shop floor and from the market and to the offices of the bureaucracy where they may apply for help in the case of financial trouble.

The most important issue is dynamic adjustment. If the budget constraint is hard, the firm has no other option but to adjust to unfavorable circumstances by improving quality, cutting costs, introducing new products or processes, i.e. it must behave in an entrepreneurial manner. If, however, the budget constraint is soft such productive efforts are no longer imperative. Instead, the firm is likely to seek external assistance, asking for compensation for unfavorable external circumstances. The State is acting like an overall insurance company taking over all the moral hazards with the usual well-known consequences; the insured will be less careful in protecting his wealth. Schumpeter(6) emphasized the significance of "constructive destruction": the elimination of old products, technologies, organizations which were surpassed by the more efficient new ones. The soft budget constraint protects the old production line, the inefficient firm against constructive destruction and thus impedes innovation and development.

A third consequence of the soft budget constraint may show up in the formation of excess demand. Whatever goals the managers of firms have (maximizing short- or long-term profits, sales, growth of sales, size of firm, discretion and power) these objectives or any combination of them will be associated with expansion. And whatever specific input-output combination may serve expansion, the drive to achieve the goals listed above generates an ever-increasing demand for at least some inputs over time. If the budget constraint is hard, this demand is constrained. Expenditures on purchasing inputs is conditional on past, present, and future revenues generated by the sale of output, which again is constrained by the firm's demand for output. If, however, the budget constraint of many firms is soft, their demand for inputs becomes unconstrained (or at least unconstrained from the point of view of financing). Run-away demand will appear. These firms feel that when they cannot pay the bills, someone else will step in and bail them out. Therefore there is no compulsory limit on demand for inputs, and particularly on investment. If the share of economic units with a soft budget constraint and a tendency to run-away demand for inputs is large enough to have a strong effect on total demand, the system becomes a "shortage economy."

Here we arrive at some theoretical conclusions. As emphasized before, the existence of a (hard) budget constraint is equivalent to Say's principle being in force. If, however, the budget constraint is soft in sufficiently large segments of the economy, then Say's principle does not hold and, as a consequence, Walras' law(7) does not hold either. Say's principle assumes that the firm is ready to start the project only if it seriously believes that the flow of revenues from the sale of the output generated by the new project will cover the flow of expenditures needed to accomplish the project. True, in a world of uncertainty different decision-makers might exhibit different degrees of risk-aversion. But given the distribution of risk-aversion over all investment decision makers, total demand for investment resources (investment credits, investment goods, etc.) will be constrained, because of the genuine fear of financial failure, that is, because the budget constraint is hard. There will be self-restraint in the capital formation decision. This symmetric relationship between demand for investment resources and the supply generated by the same investment resources underlies the idea of Walras' law, i.e. the sum of the (positive and negative) values of excess demands will be zero.

This kind of symmetry gets lost in the case of a sufficiently large number of decision makers with soft budget constraints. The symmetry breaks down if financial support can appear like manna. The firm can start a project even though it may have the subconscious suspicion that the cost will be more than planned and the revenue less. In case of financial failure it will be bailed out. Under such circumstances there is no self-restraint in investment intentions; the demand is not counter-balanced by a "dead-serious" consideration of revenues and ultimately of supply.

There are identities in all economies: stock-flow balances of real inputs and outputs and of money. These identities self-evidently hold also in [an] economy with soft-budget constraints. But Walras' law is not an identity but a certain relationship between buying and selling intentions. Intentions can be inconsistent. In case of a soft budget constraint they are inconsistent. Subsidies, soft tax-exemptions, soft credits, etc., will be financed through the redistribution of income via taxation or inflation. Everyone takes into account the usual tax burden, inflation rate and so on when planning his finances. The expectation that the firm can spend more than its "earnings" because in case of failure it will be bailed-out, comes in on top of that. Here is the source of asymmetry: the possibility of run-away demand of the firm with soft budget constraints. The individual expectations can be incompatible with each other. The softening of the budget constraint is an inducement to such incompatibility: the softer the budget constraint and the larger the sphere of the economy where the syndrome prevails, the more incompatibility appears.

Another important aspect is the effectiveness of monetary policy. A monetary ceiling is a necessary condition of financial discipline, but is not sufficient to ensure it. The transmission between a tighter monetary policy and the micro-response becomes unreliable in the case of a soft budget constraint. The latter is like a cog-wheel in this transmission. The microunit will not react to monetary restraint by restricting its demands when it is not convinced of the dangers of financial failure. In the sphere of microunits with a soft budget constraint money is more or less passive(8). . . .

III. The Experience of Socialist Hungary

We now turn to empirical observations. . . . The case of "classical socialism," i.e. the highly centralized pre-reform command economy, is rather straight forward. It is officially acknowledged that profitability must not play a decisive role: entry, exit, expansion and contraction of the firm does not depend on profitability, but is decided by higher authorities applying other criteria. A loss-making firm or a whole sector can survive indefinitely, provided that the higher organs of the State want it. . . .

In Hungary a research team is studying the financial situation of all State-owned enterprises (1,755 firms in 1982) which produce the bulk of total output. The balance sheets of all these firms have been processed and several special indicators have been computed for cross sectional and dynamic analysis. Here only a few examples of the numerical results can be presented.

Some explanation of terminology is needed. We distinguish four categories of profit.

1. Original profit. This is a hypothetical number: profit before receiving any kind of subsidies from the State and before paying any kind of taxes to the State. . . .

2. Corrected original profit. This is profit number 1 plus subsidies granted for the purpose of keeping certain consumer prices down minus turnover taxes levied for the sake of keeping certain consumer prices up. The rationale for this correction is as follows. We want to filter out the component of fiscal redistribution which aims at subsidizing or taxing the consumer households, not the firms.

3. Reported profit. This is the profit reported in the balance sheets and later on, in all sectoral and national statistics on profits. They reflect already a large degree of fiscal redistribution: most of the subsidies and most of the taxes are subtracted from original profit at this stage.

4. Final profit. After the reported profit is determined a few more subsidies

are added and a few more taxes are subtracted.

The first observation is that the size of the fiscal redistribution is very large. . . . The State-owned sector as a whole is a net taxpayer. But the final net outcome is preceded by a far reaching reshuffling of profits crisscrossing among the original firms. The State takes away money from a firm with one hand and then gives money to another firm (or perhaps to the same firm, but with another "entitlement") with its other hand. Or more precisely, the State has not only two hands but it is a Shiva with many more hands: there are a total of 276 types of taxes and subsidies used by different tax-levying or subsidy-granting authorities.

At this point a word of caution is in order. We do not suggest that profitability No. 1 is the indicator of genuine efficiency. With the given distorted relative price system that cannot be the case. Therefore it is not legitimate to draw the simple normative conclusion to stop differentiated financial distribution, and apply a kind of flat tax while maintaining the present price structure. We do not want to draw any normative conclusion here, only to point out the characteristic feature of the present situation. When fiscal redistributions are so widespread and so complex then "profitability" does not have and cannot have any reasonable meaning. Reported and final profitability depend at least as much on the generosity or tight-fistedness of different subsidy-granting authorities, as they depend on the success or failure in production and on the market.

The fiscal redistribution of profits shows a conspicuous tendency to give financial assistance to the losers. We computed the following indicator: the total subsidy given to a firm over total taxes paid by the same firm. We call it the "ratio of redistribution." The correlation coefficient between original profitability and the ratio of redistribution for the whole population of State-owned firms is -0.99 for 1980, -0.97 for 1981, and -0.92 for 1982. The very strong negative correlation demonstrates that the lower is original profitability, the higher is the probability of getting a larger subsidy and paying a smaller tax.

The redistribution pattern, therefore, is to redistribute profits from the winners to the losers. For the sake of demonstration, firms are classified in four categories: "loss-making" means profitability less than -2%; "low profitability" is between -2% and +65% "medium profitability" is between +6% and +20%; and "high profitability" is greater than +20%.

Firms with high original profitability have only an 11% chance to end up in the same category after redistribution; almost every second one will be downgraded to low profitability. In contrast 9 out of every 10 loss-making firms will be upgraded. This is a rather paradoxical form of "egalitarian" redistribution: profit incentives dampened by the leveling of profits.

Every year a few Hungarian State-owned firms go out of business. They are liquidated or merged into a larger firm. Our analysis as well as other studies, have shown that exit is not related to profitability . The relationship between profitability and the growth of the firm is also worthy of attention. For the sake of cross-sectional and dynamic comparison we defined an indicator of "investment activity": expenditure on real capital formation divided by the value of physical assets. . . . [We found] that investment activity is not correlated with profitability at all. . . .

IV. Experience in Mixed Economies

Socialist economies exhibit a rather extreme degree of budget constraint softness. To a lesser degree and in more restricted segments of the system, similar phenomena can be observed in mixed economies as well.

It is impossible to make general propositions about the degree of softness or hardness of the budget constraint in mixed economies. The variance is large; there are great differences between countries, and within a particular country the situation may change as parties and political currents in power change. [A] few words must be said about the forces which create the phenomenon. As a first approximation we consider the arguments of the organizations which are asking for and expecting external assistance. The variety of specific arguments is of course very large, but we can try to find out their most important common ingredients.

(a) The most frequently quoted reason for external help is the protection of jobs. In a system of perfectly hard budget constraints of firms and households, all adaptation both cyclical macro-adjustments and structural micro-adjustments would be associated with large lay-offs and with wages fluctuating in both directions according to the situation in the labor market. Exit of the firm hurts owners, managers and employees; they try to get State assistance to avoid shut-downs. During recessions, the demand for State intervention is supported by great masses. But also in upswings there are sectors or single firms which are still ailing. The employees feel that it is not fair that they are left out of the benefits of growth.

(b) Another rather frequent argument is the protection of domestic production against foreign competition. This frequently coincides with (a), i.e. with defending jobs. Not all protectionist measures imply the softening of the budget constraint, but quite a few have such implications. The most important measures in this respect are the subsidies to firms or whole sectors which because of high domestic costs have troubles in competition with foreign firms selling at lower prices.

(c) In many instances the softening of the budget constraint is related to redistributive policies in favor of the poor, the handicapped, the sick, the elderly. Redistributive objectives in the name of fairness, social justice and solidarity can motivate non-profit institutions, local governments or certain branches of the national governments in their demands for additional financial assistance.

(d) An important argument, closely related to (a) and (c), in favor of softening the budget constraint is the demand for security and stability: to protect the individual and ultimately society as a whole from fluctuations and uncertainties. We already used the analogy of the State as a general insurance company. This desire for security and stability is the motivation for impeding the "natural selection" executed by the market, for guaranteeing the survival of malfunctioning banks and producing firms.

(e) Each organization serves almost by definition a certain purpose: an important argument is to refer to the social importance of that particular purpose when arguing for external additional support. As mentioned before, the leaders of an organization "fight" for the survival and for the expansion of their unit, usually supported by their staff. In this fight, military leaders will refer to the importance of national defense, the top administrators of the police to the importance of public security, the top administrators of the heath-system to the importance of health care and so on. All these requirements and claims are, of course, plausible and legitimate. Since they serve objectives which have no "market value," it is unavoidable that their relative value is determined by a political process.

Ultimately, the soft budget constraint phenomenon is a joint outcome of two closely related socio-political trends. First, the increasing, and often overloading demand of society on the State to become a "protector," responsible for the welfare, growth and the national economic interest, and second, the self-reinforcing tendency of bureaucratization. The softening of the budget constraint is an indicator of the fact that many basic allocative and selective processes are not left to the market, but are highly influenced or taken over by bureaucracies and by political forces. This trend proceeds with uneven speed in different countries; there are also reversals for some time. In any case, there is no contemporary mixed economy where the paternalistic role of the state and of political forces is not much stronger than, say, half a century ago.

A final remark on political and ethical implications. There will surely be readers who draw extreme conservative conclusions from the ideas outlined here. This is far from the intentions of the paper, which does not suggest that the hard budget constraint is "good" and the soft is "bad."

A system based on a perfectly hard budget constraint for every decision-making unit is a terribly cruel one. The symbols of such a system are the debtor's prison, the bailiff bringing under the hammer the home and the household goods of the insolvent family, mass lay-offs in bankrupt firms and so on. All changes departing from these brutal extremes contain some elements of a softer budget constraint. It can be hardly denied that the majority of the population in all countries wanted to move away from that extreme point.

Careful case-by-case considerations are needed if we turn to policy suggestions. Sometimes these are relatively easy. The budget constraint can be hardened for the sake of efficiency without (or with little) painful human consequences. In many other cases, however, the choice is much more difficult. There can be a trade-off between the two kinds of consequences of softening or hardening the budget constraints: the impact on efficiency and the impact on human well-being and suffering. The hardness of the budget constraint is based on fear of a financial catastrophe; the softness eliminates this fear. A hard budget constraint includes competition: the winner gains, the loser will be ruined. A soft budget constraint has mercy on the loser. It is not the purpose of this paper to "solve" the ethical dilemmas. There is no general solution; one has to search for acceptable compromises in each case. Here we want to emphasize only that there is a deep dilemma. Efficiency and security-stability are to a large extent conflicting goals.

1. -

2. - -

3. -

4. - - -