Democracy in Deficit: The Political Legacy of Lord Keynes

There is little mystery about Keynes’ own assumptions concerning the politics of economic policy. Personally, he was an elitist, and his idealized world embodied policy decisions being made by a small and enlightened group of wise people. But these “presuppositions of Harvey Road,”
*1 as they are called by his biographer, extended beyond idealization. Economic policy decisions must be made in the real world, and these Keynesian presuppositions also impinged on reality. Keynes not only envisaged government by an enlightened and small elite as his ideal; he also assumed that, at base, this model described government as it actually was observed to operate.

These political presuppositions of Keynes, as were and are those of many of his professional colleagues before and since, were probably influenced by a mixture of positive and normative elements. Keynes held important positions in the British government, and his ideas exerted notable influence on policy, especially during World War I. He probably exaggerated his own role in policy decisions. Furthermore, as so many who knew him personally have remarked, Keynes was an extremely persuasive man in argument, and confidence in his own ability to convince others may have led him to discount the potential importance of genuinely differing attitudes that might emerge in a collective decision process. His biographer, R. F. Harrod, suggests this when he questions the applicability of the presuppositions in a setting different from that assumed by Keynes.

If, owing to the needs of planning, the functions of government became very far-reaching and multifarious, would it be possible for the intellectual aristocracy to remain in essential control? Keynes tended till the end to think of the really important decisions being reached by a small group of intelligent people, like the group that fashioned the Bretton Woods plan. But would not a democratic government having a wide multiplicity of duties tend to get out of control and act in a way of which the intelligent would not approve? This is another dilemma—how to reconcile the functioning of a planning and interfering democracy with the requirement that in the last resort the best considered judgment should prevail. It may be that the presuppositions of Harvey Road were so much of a second nature to Keynes that he did not give this dilemma the full consideration which it deserves.
*2

Normatively, Keynes was no democrat, in any modern descriptive meaning of this term. To the extent that Keynes might have predicted interferences with rational policy to emerge from the sometimes grubby institutions of electoral and party politics, he would have been quite willing to jettison such institutions, regardless of their history and of their traditional role. Perhaps it is best simply to say that Keynes was not particularly concerned about institutions, as such. His emphasis was on results and not on rules or institutions through which such results might be reached. And if institutional barriers to what he considered rational policy planning should have worried him, Keynes would have been ready to set up a “national planning board” run by a committee of the wise.
*3

It is important to recognize explicitly what Keynes’ political presuppositions were because the policy implications of his “general theory” were developed within this framework. We observe, however, that, for the United States in particular, these policy precepts have been advanced as appropriate for a political framework that scarcely resembles that postulated by Keynes.
*4 If the form of government is relevant for economic policy, we must at least question this unchallenged extension.

The preliminary response of the orthodox economist would be to suggest that his role does not require that he pay attention to the institutional means through which his ideas might or might not be accepted and implemented. A widespread and often implicit conviction to this effect has prevented even the most elementary recognition of the possible linkage between politics and policy. The orthodox economist, whether Keynesian or non-Keynesian, whether macroeconomist or microeconomist, has remained essentially uninterested in the political setting within which economic ideas may or may not be translated into reality. To explain this long-observed and continuing insularity of the economic theorist from politics, from public choice, would require a separate treatise, replete with methodological argument. For our purposes, we take this insularity as fact. Knut Wicksell noted, as early as 1896, that economists had almost all proceeded on something resembling the Keynesian presuppositions, on the notion that their role was one of proffering advice to a benevolent despot.
*5

And this frame of mind has continued to the present, despite the observed presence of democratic choice-making processes. Herschel Grossman, in reviewing the contributions of James Tobin to macroeconomics, has described clearly this failure of economists to recognize that macroeconomic policy emerges through democratic politics, not from the board room of a committee of the wise. While particular references were to Tobin, Grossman was speaking of economists in general when he noted:

Tobin presumes that the historical record of monetary and fiscal policy involves a series of avoidable mistakes, rather than the predictable consequences of personal preferences and capabilities working through the existing constitutional process by which policy is formulated. Specifically, Tobin shows no interest in analysis of either the economically motivated behavior of private individuals in the political process or the behavior of the government agents who make and administer policy.
*6

The Economic Environment of the “General Theory”

Why should political institutions influence economic policy? Why should policy norms that are found to be applicable under one structure for collective decision making be inappropriate or inapplicable in an alternative structure? These questions might be disposed of readily, indeed in summary fashion, if the choice among separate policy options were always sharp and clear, if precepts of rationality suggested a dominant policy strategy. It is in this respect that the economic environment within which the Keynesian theory was developed becomes relevant for understanding the neglect, by both Keynes and the Keynesians, of the political framework. In the setting of the early and most elementary Keynesian model, no rational government, regardless of organization, could fail to carry forward the policy norms that seem to emerge so clearly from the “general theory”; this being so, the political framework for decision could be treated as largely irrelevant.

Let us summarize this elementary model. The aggregate supply curve runs horizontally to some point that is labeled “full-employment income,” beyond which the curve turns sharply upward. The economy in underemployment equilibrium is then depicted by the intersection of the aggregate demand curve with the aggregate supply curve along the horizontal portion of the latter. The policy implication becomes evident; aggregate demand should be increased. There is no other way to increase employment and output. But, from another set of diagrams or equations, we can also see that the economy is caught in a liquidity trap. Interest rates are at a floor, and additional money creation alone will merely add to hoards; private spending will not be affected. Additional government spending becomes the natural policy consequence, with deficit creation indirectly implied.

We are not interested here in examining either the logical coherence or the empirical validity of the description of the national economy that is embodied in this most basic of Keynesian models. Our emphasis is on the attitudes of those who accepted such a model of economic reality as the ground upon which to construct policy prescriptions. In the depression-stagnant economy depicted here, the creation of a deficit in the government’s budget seems clearly to be dictated by rational policy norms, requiring only the acknowledgment that full employment and expanded real output are appropriate objectives. Why would any person, placed in a responsible decision-making position, raise objections to the policy prescribed if he understood the argument? Consider the elected member of a legislative assembly, a congressman or a senator. Deficit creation involves either a reduction in taxes and/or an increase in public spending. Both of these actions, taken independently, are desired by citizens-constituents. There are no offsetting costs to balance against the benefits; more of everything may be secured. There seems here to be no possible conflict between “politics” on the one hand and the true “public interest” on the other. Good politics seems good policy, and opposition can be traced to the presence of outdated and inappropriate fiscal rules.

If the elementary Keynesian diagnosis-description of underemployment equilibrium is accepted, and, furthermore, if this is considered to be a possible permanent state of the economy in the absence of corrective governmental action, the economist will tend to stress the clear and distinct benefits that stem from the indicated policy action, and he will tend to be blind to attempts at rational refutation. He will, in addition, find it nearly impossible to imagine that the institutional barriers of practical politics can permanently frustrate the clearly defined vision that the revolutionary Keynesian model offers him. If we place ourselves in the stance of the early Keynesian enthusiasts, we should perhaps not be surprised at their neglect of political structure.

These enthusiasts did not critically question the logical coherence or the empirical validity of the model, nor did they examine carefully the applicability of the policy norms in other situations, despite the claims of having developed a general theory. The interest was concentrated on restoring prosperity; the prevention of inflation was not considered to be even so much as a potential problem. How could we have expected the early enthusiasts of the revolution in ideas to foresee the political impracticality of applying the Keynesian policy tools in such a converse setting?

Strings Can Be Pulled

The continued stance of those who called themselves “Keynesian” after the end of World War II should not be so sympathetically interpreted. Keynes died in 1946, and we have no means of judging how his own attitude might have been shifted by his observation of events. After 1946, our strictures apply to those who considered themselves to be the preachers of the Keynesian gospel, the textbook writers, the economic journalists, the governmental counselers.

The facts were soon clear in the immediate postwar years. The American economy was not settling down to a permanent underemployment equilibrium. Inflation, not unemployment, seemed to be the threat to prosperity. Aggregate demand was excessive, not deficient, and the Keynesian tools were, at best, awkward in their application. Under the political presuppositions of Harvey Road, these tools were symmetrical in application. Budget surpluses should have been created to mop up the excessive demand.

Instead, in early 1948, taxes were reduced in the United States. Should not this action have given early pause to those Keynesian economists who had hitherto paid almost no heed to the workings of democratic process? It should have been evident that the fiscal policy precepts emerging from the simple Keynesian analysis were one-sided in their practical application. The creation of surpluses would have been recognized as politically different in kind from the creation of budget deficits. In retrospect, it seems likely that the political biases of the whole Keynesian edifice would have been acknowledged in the early postwar period, had it not been for the simultaneous “rediscovery” that money matters. This allowed the responsibility for demand restriction to be shifted temporarily from fiscal policy to monetary policy instruments, which were allegedly adjusted under a set of institutions that fit the Harvey Road presuppositions.

We should recall that, in its early and elementary form, the Keynesian policy model embodied a total disregard of money and of monetary policy. In underemployment equilibrium, money creation alone was alleged to be ineffective in stimulating aggregate demand. The government’s budget became the central policy instrument, and, by implication, it was the only instrument required to maintain national economic prosperity. Under the presumption that fiscal policy application was fully symmetrical, monetary policy could be neglected whether stimulation or restriction in aggregate demand was needed. Monetary policy could, therefore, be relegated to a subsidiary role of facilitating credit markets, and it might be directed mainly toward keeping interest rates at or near the liquidity floor as a means of encouraging investment. But experience under this policy, in both the United States and Great Britain, soon suggested that, unless genuinely draconian fiscal measures were to be undertaken as offsets, inflation was not likely to be controlled. Monetary policy was quickly reincorporated into the sophisticated Keynesian’s set of instruments, and the efficacy of monetary restriction on reducing aggregate demand was acknowledged.

Indirectly, the Keynesian economists who modified their norms to allow for the use of monetary policy instruments when demand restriction seemed to be needed were acknowledging the political asymmetry of the whole Keynesian structure. They did not, however, do so consciously, and they continued to talk in terms of some idealized application of both fiscal and monetary policy tools. In one of his best-known policy statements, Paul Samuelson called for a policy mix of “easy money” and “tight budgets,” designed to ensure a high rate of investment and capital formation while holding down consumption spending to avoid inflationary pressures.
*7 In this idealized setting, budget surpluses, which could be financed by tax increases, well might be required in order to allow for the expansion in the demand for investment goods generated by the low interest rates. Samuelson’s position was widely accepted and discussed by economists, as if the postwar fiscal experience had not happened, as if the “presuppositions of Harvey Road,” or their American counterparts, remained descriptive of political reality.

Little or no sophisticated insight should have been required to suggest that, at best, the fiscal and monetary policy instruments would tend to be applied nonsymmetrically in political democracy. Budgets would rarely, if ever, be observed to be overbalanced as a result of purposeful efforts at demand restriction. “Easy budgets” promised to be the order of the day, and especially during periods of recession, regardless of magnitude. On the other hand, if demand was to be restricted at all during periods of threatened inflation, this task would fall to the monetary authorities, who did seem to possess some nominal independence from the political process, at least in the sense of direct constituency pressures. The best policy package that might have been predicted to emerge was clearly one of “easy budgets and tight money,” just the reverse of the Samuelson norm.
*8

The Great Phillips Trade-off

The incorporation of monetary policy instruments for the purpose of restricting spending might have occurred even if the form of the aggregate supply function assumed in the early Keynesian models had been proved to be an accurate representation of the underlying economic reality. If the aggregate demand curve cuts the aggregate supply curve to the right of the full-employment kink or corner, inflation control rather than employment stimulation becomes an objective, even in the most naive of Keynesian models. The postwar record soon revealed, however, that this simple functional form was far from descriptive. There was no horizontal portion of an aggregate supply curve, and models based on the presumption that such a portion exists were likely to produce biased results. There were no kinks; increases in aggregate demand did not exert an effect wholly, or even largely, on employment and output up to some magic point labeled “full employment,” and only thereafter commenced to exert upward pressures on prices in the economy. If an aggregate supply relationship was to be used at all, empirical reality seemed to dictate that this be shown diagrammatically as a curve that sloped upward throughout its range. This suggested the presence of a continuing trade-off between the two acknowledged objectives for national economic policy, between employment and price stability.

The converse of this relationship, that between the rate of unemployment and the rate of price inflation, became the central topic for discussion among macroeconomists for the better part of two decades. The “Phillips curve,” which depicts this alleged trade-off, found its way into the textbooks in elementary economics, replacing earlier Keynesian supply functions.
*9 We shall discuss some of this Phillips-curve analysis in more detail in Chapter 11. Our purpose here is limited to questions concerning how the recognition of such a trade-off might have modified economists’ political presuppositions and, in turn, how the political framework itself might have become more significant for policy outcomes in the presence of Phillips-curve trade-offs.

The relationship dramatically modified the setting for macroeconomic policy choice. Even during periods of recession, when aggregate demand might have seemed to be deficient by some standards, policies designed to increase total spending in the economy were not costless. Additional output and employment, acknowledged as desirable, could be attained only at the expense of some inflation, equally acknowledged as undesirable. There simply was no horizontal portion of some aggregate supply curve where employment might be increased without inflation, where increased employment was, in this sense, “costless” to achieve. The Phillips curve suggested that a trade-off between mutually desirable but mutually conflicting objectives was likely to be present, regardless of the state of the national economy.

No single set of policy actions dominated all others; rational, intelligent, and fully informed persons might differ as to the relative weights to be assigned to alternative objectives. Whether inflation should be stimulated as a means of securing more employment, or some unemployment should be accepted as the price for holding inflation within bounds, was a question that could be answered only in terms of basic social values, about which persons might differ. Even under the Keynesian presuppositions about policy making, even with governmental economic policy decisions in the hands of a few wise people, who could now predict their actions? No longer did the policy precepts emerge with clarity from the economists’ analytical model of the world.

The Phillips curve was alleged to depict the set of possible outcomes; the choice among these possible positions was to be made on the basis of the community’s preferred rate of trade-off between the components. Economists diagrammed all of this by introducing a set of community or social indifference curves, and they could then indicate some “optimal” policy choice as that which allowed the community to attain its highest level of utility.
*10 This construct was then utilized to interpret aspects of political reality; the shift in economic policy between the Eisenhower and Kennedy administrations was almost universally interpreted as a shift in the relative weights assigned to the inflation control and the employment objectives.

It should have been clear that the presence of a Phillips-curve trade-off between unemployment and inflation would make the institutions of decision making, the politics of policy, more important rather than less so. Despite this, economists continued to ignore this element in their diagnoses and prescriptions. Even for a group of enlightened people, decisions on relative weights could scarcely remain wholly immune from political feedbacks. Until and unless the public, acting upon and through their elected political representatives, generally could be depended on to understand and to acquiesce in the weights assigned, tension would be set up between the decision makers and the community at large. The New Economics, which now was Keynes amended by the Phillips-curve trade-off, was not so simple as before. Consider the potential for policy conflict when a public opinion and political attitude partially reflecting the norms of the earlier simplistic Keynesian model were imposed on a set of policy decisions emerging from the allegedly more sophisticated Phillips-curve analytics. Unemployment might be observed to exist; full-employment income, defined by some presumed Keynesian kink in an aggregate supply curve, was not being generated. The simple Keynesian precepts might then have called for increases in total spending, for increased budget deficits, without much if any recognition of the dangers of inflation. By contrast, the economic decision makers, whoever they might be, might recognize the existence of a Phillips trade-off, but how could they fail to remain unaffected by the prevailing public-political attitudes? Could we not predict an inflationary bias in the Phillips-curve world, even under the presuppositions of Harvey Road?

Post-Keynes, Post-Phillips

Perhaps we are lucky that the inflation-unemployment relationship discussed so widely under the Phillips-curve rubric (which has not yet disappeared from the elementary textbooks) was, like the simple Keynesian model that preceded it, doomed to be disavowed and discredited under the weight both of logical analysis and of accumulating empirical evidence. In a short-run context, additional employment and output may be stimulated by increases in total spending, accompanied by some unanticipated inflation in prices. If, however, explicit policy measures designed to add to total spending are continued over a sequence of periods, inflation will come to be anticipated, and inflationary expectations will be built into the whole structure of negotiated contracts. From this point in time, the generation of inflation that has been predicted will do nothing toward stimulating employment and output. If, in fact, it could be realized, a fully predicted and steady rate of inflation will ensure the same overall level of employment that would have been ensured with a zero rate; in this setting, there is no Phillips-curve trade-off. To secure a further short-run increase in employment, the rate of increase in total spending and the rate of inflation must be accelerated, and the acceleration itself must be unanticipated. Once the community finds itself saddled with inflationary expectations, however, attempts to reduce the rate of inflation will themselves have the same results as reductions in aggregate demand in a more stable setting. Unemployment may be generated by attempts to do nothing more than hold the rate of inflation within tolerable limits.

For long-run policy planning, the evidence as well as the logic suggests that there is, in effect, no sustainable trade-off between unemployment and inflation. There exists neither the Keynesian kink nor the modified Phillips slope. Aggregate demand increases cannot permanently stimulate employment. The rate of employment can be influenced by such things as minimum-wage legislation and the monopolistic practices of trade unions, but not by changes in aggregate demand.
*11

Our concern here is not with evaluating critically the evidence that has modified economists’ attitudes, which has caused many economists who earlier called themselves “Keynesians” to become increasingly skeptical of the selfsame policy norms they espoused in the 1960s. Our concern is with the effect that this change in attitudes and analysis might have had on economists’ presuppositions about the political process and on the feedback influences that this process itself exerts on economic policy.

We are lucky in that the creeping inflationary bias introduced by the Phillips-curve paradigm was exposed earlier than might have been the case if the underlying empirical realities had been as depicted. The Keynesian tool kit is bare in the world of the 1970s and 1980s. And this is independent of the structure of political decision making. Even if wise persons of Whitehall or Washington, as envisaged by Keynes and the Keynesians, should be empowered to make macroeconomic policy without influence from the grubby world of everyday politics, they could scarcely attain satisfactorily full employment simultaneously with an acceptable rate of inflation. To the extent that such persons were honest as well as wise, they would have to turn to non-Keynesian tools, to those that might attack the structure of labor markets, to those that might open up opportunities for investment and for employment.

The contrast between what an enlightened elite would impose as economic policy and the actions that we, in fact, observe was never so great as in the post-Keynes, post-Phillips era of the 1970s. As noted earlier, there could have been little or no such contrast in the deep depression of the 1930s, if only the politicians had learned their Keynesian lessons. And, during the Phillips-curve years, disagreements might have emerged only over the relative weights assigned to conflicting objectives. In the post-Phillips setting, however, we observe massive budget deficits, high rates of inflation, and high levels of unemployment. Those who accept the Keynesian political presuppositions in the normative sense, those who feel that a few wise people should be empowered to direct the lives of the rest of us, must now base their argument on the allegation that things would indeed be better if only politics did not intervene.

Reform through National Economic Planning

Perhaps this, in itself, represents progress in understanding. At least those who blame the workings of modern democratic processes for the sorry state of economic policy are indirectly acknowledging that the institutional structure does exert its influence. The distance is shortened between this acknowledgment and possible suggestion for institutional reform. But two separate and divergent routes to reform may be taken, one which we may label “democratic,” the other clearly as “nondemocratic.” Unfortunately, those who tend to be most critical of democratic politics tend to support structural changes that will, if implemented, remove economic policy decisions from democratic controls. These reformers seek to force upon us something like the institutions postulated under the presuppositions of Harvey Road. If we have not been, and are not yet, ruled by an elite and self-chosen small group, these critics say, so much the worse for us. Such a ruling group “should” be established, empowered with authority, and divorced from the feedbacks of electoral politics. “National economic planning,” done through some “National Economic Council” or “National Planning Board,” could, presumably, save the day. Not surprisingly, given the acknowledged loss of faith in the Keynesian gospel, articulate demands for “planning” surfaced with some fanfare in 1975, after a quiescence of thirty years. These demands took the form of discussion surrounding a proposed legislative program introduced by Senators Humphrey and Javits, a program supported by a few prominent economists.
*12

How could a group of planners genuinely divorced from politics, an “economic supreme court,” resolve the American dilemma of stagflation? As suggested above, they could accomplish little by even the most skillful use of the basic Keynesian tools. How could more careful manipulation of the federal budget or of monetary policy secure both full employment and a return to price-level stability after the orgies of the 1965-1976 period? Inflation could, of course, be brought within bounds; price-level stability can be accomplished. But how could this be done, save at the expense of unacceptably high levels of unemployment, higher even than those experienced in 1975?

An incomes policy, a euphemism for wage and price controls, could be imposed, only with more firmness and strictness than in 1971. Decisions about prices and wages would be orchestrated by a committee of experts; such decisions would no longer be left to agreement among free individuals over their terms of contract. Such controls have an ancient history, dating back at least to the Code of Hammurabi in the eighteenth century &bc; While such controls historically have invariably failed, they have created much damage in the process. The efforts to evade and avoid the controls come about through a diminution in the “socially” productive activities of individuals, so levels of economic well-being consequently decline.

Alternatively, a planning board, intent on achieving the dual objectives of full employment and price-level stability, could work on the structure of national labor markets. Such a board would find it necessary to intervene directly in the economic order, as it exists within politically imposed boundaries and constraints. An “economic supreme court” could, for example, declare minimum-wage regulation to be contrary to some implicit “economic constitution.” In so doing, a sizeable increase in employment, and notably among members of teenaged minority groups, could be secured. But this and other comparable steps which would possibly improve the working of labor markets seem clearly out of bounds through normal legislative processes. Does it seem likely that Congress would consciously delegate such powers of economic policy making to any appointed officials, whether these be designated as members of a national planning board or anything else?

In contrast to nondemocratic approaches to reform, a quite distinct avenue for reform lies in the prospect that the democratic political processes themselves can be improved. Is it not more consistent with American political tradition that the institutions of decision making impose upon politicians constraints that will ensure against the excesses that have emerged from the widespread political acceptance of the Keynesian policy norms? The prudent person acts wisely when he imposes behavioral rules upon himself, rules that may bind his actions over a series of unpredictable future steps. Is it impossible to expect that prudent members of democratic assemblies of governance could do likewise? Should we not look for genuine institutional reform within the structure of democratic decision making rather than for changes that replace this structure?

We shall discuss possibilities and prospects for democratic reform in some detail in Chapter 12. Before we can do so, however, we must first understand the problem that we confront. We must first analyze carefully the reasons democratic decision making, as it exists in the United States, has produced the economic policy results that we observe in the 1970s. We must drop all pretense that economic policy decisions are, or should be, made by a small and well-informed group of people seeking the “public interest.” We must escape the blinders imposed on us by all presuppositions akin to those of Harvey Road. We must look at the application and acceptance of Keynesian economics in a political setting where democracy is reality, where policy decisions are made by professional politicians who respond to demands, both of the public and of the bureaucracy itself.