The neoliberal approach to economic growth of the past forty years has been to drive down real wages, deregulate the financial sector, privatize various government functions, cut social expenditures and reduce taxes on large corporations and the wealthy. The period has been characterized by sluggish growth and widening inequalities. In contrast, the markedly higher rates of growth achieved from the end of the Second World War until the early 1970s were sustained in a regime of steadily rising real wages, which were consciously linked through policy or collective bargaining to improvements in productivity. This approach to industrial relations was played out within a context of substantial direct public-sector involvement in the economy, tight financial regulation, high and steeply progressive taxation and a more equal distribution of income and wealth.
A similar pattern is discernible in communist and former communist countries. The Soviet Union achieved strong rates of growth in the period 1950-73 and then fared increasingly poorly over the period of “reform”, especially from 1985 onwards (Mazat and Serrano provide a recent study). The subsequent transition to capitalism and extreme application of neoliberal policy prescriptions resulted in an economic disaster that persists to this day in many parts of Eastern Europe. In Asia, of course, we see the strong growth record of state-dominated China, which is yet to embrace the neoliberal dogma.
With this broad historical comparison in mind, it is perhaps worth considering the role of government spending in fostering a high-wage, demand-led growth regime.
A Point of Clarification – Impending Ecological Calamity and All That
In case it is not clear, the intention, in what follows, is not to champion a continuation of our present environmentally and socially destructive course. Output and growth, as they are currently defined, are measured in monetary terms and essentially involve adding up monetary expenditures. ‘Real’ measures of output and growth are then derived by deflating nominal figures to take account of changes in the level of prices. Precisely what we choose our monetary expenditures to be directed toward – whether ecologically regenerative activities or environmentally damaging ones – is an open question. If we pay people to learn, or to clean up the environment, or to develop new eco-friendly technologies, to take a few examples, these will all count toward output and growth. An expansion of these and similar activities counts just as much in the official statistics as an expansion of environmentally damaging activities. Hopefully we choose a sane path. Governments, especially when currency issuers, have the capacity not only to foster growth but to foster healthier forms of growth.
It is clear that so long as we persist with capitalism, profitability will remain the overriding objective of private-sector investment. Even so, in principle, it seems conceivable that private investment can be induced within a regime of generally low or generally high profitability. If demand is kept sufficient to create a “high pressure economy”, in which firms who fail to invest risk losing market share, a low average rate of profit (corresponding to higher real wages) will not necessarily harm growth. At the micro level, capitalists are likely to seek out the most profitable opportunities among those available, irrespective of whether, at the macro level, the average rate of profit is intentionally kept high or low. Capitalists, of course, continually agitate for a high-profit-rate “low-pressure economy”. When they get their wish, the likely result, as recent history suggests, is slower growth (since there is less reason to invest if productive capacity is already ample to meet consumption demand) and widening inequality. It is a matter of class war who wins out on the distributive front – the masses or the wealthy minority.
If it is true that capitalist investment behavior at the micro level is guided by relative profitability, then it is necessary for government to reshape private market incentives dramatically, with the explicit purpose of altering the kinds of activities that are profitable. Otherwise, ecological disaster appears certain. A judicious mix of subsidies and tax breaks for desirable investments, punitive taxes and penalties on undesirable ones, the outright banning of socially and ecologically destructive production and the nationalization of strategically critical roles might all help to make a dint in our present reckless developmental path, though in my view there can be no substitute for ending capitalism altogether. Although the present focus is on the macro effects of government spending, this is not meant to downplay the importance of other aspects of policy, the institutional framework, or irredeemable features of the system itself.
Private Investment in Relation to Capacity Utilization and the Rate of Profit
Since we are, for now, dealing with capitalism, it is perhaps worth starting with a consideration of profit. The rate of profit r is sometimes expressed as the product of three macro ratios: (i) the profit share in total income π; (ii) the actual rate of capacity utilization u (defined here as actual output Y divided by ‘normal’ or planned average output Yn); and (iii) the ratio of ‘normal’ output to the capital stock K, denoted Θ:
In this expression, P is total profit before its distribution into retained earnings, dividends, rent, interest and so on.
Suppose, initially, that the normal output-to-capital ratio Θ and the profit share in income π are constant for some duration of time. Under these circumstances, it is clear that the actual rate of profit will rise and fall with capacity utilization u.
The ‘accelerator principle’ suggests that extra private investment tends to be induced when demand-determined levels of economic activity test the limits of productive capacity. In envisaging this process, it is assumed that there are untapped resources, including an excess supply of unemployed and underemployed labor, that makes it possible to adjust capacity to demand in this way. Capitalism itself tends to reproduce this excess supply of labor through technical innovations.
More specifically, investment will be induced via an accelerator mechanism when demand pushes output and the rate of capacity utilization above ‘normal’. By ‘normal’ is meant the average level at which firms expect to operate their capacity over the life of plant and equipment, allowing for fluctuations in demand. For present purposes, the normal rate of capacity utilization can be considered to occur when Y = Yn or, equivalently, u = 1.
Firms normally intend to operate with a planned degree of spare capacity. This enables them to respond to unexpected peaks in demand without losing market share. This means that Yn will be somewhat below the maximum level of output made possible with the existing capital stock. For the US, utilization rates, as a percentage of absolute capacity, have ranged from a low of 66.7 percent to a high of 89.4 percent since records began to be kept in 1967. This reinforces the view that the economy is normally constrained by demand, well inside ultimate supply limits.
In light of the US experience, it may be helpful, by way of a mental picture, to think of normal utilization, where u = 1, as occurring somewhere around 80 percent utilization of absolute capacity. That would mean absolute capacity corresponded to a utilization rate of around 1.25. In other words, u (in the way it is defined here) will frequently exceed 1.
In accordance with the accelerator principle, when demand persistently maintains output above the normal planned level (implying u > 1), this will tend to induce additional investment by firms in an effort to adjust normal capacity output Yn to demand, or, in other words, to restore the utilization rate approximately to 1.
Although under the accelerator principle investment is induced by demand, this behavioral assumption is, in a restricted sense, consistent with the idea of a positive relationship between investment and the actual rate of profit applying to the economy on average. It is just that this positive relationship holds only within the context of a particular ‘distribution’ (profit share in income) and ‘technology’ (normal output-to-capital ratio). Also, it is the rise in demand and subsequent higher rate of capacity utilization that is depicted as causing the increase in the actual rate of profit. This works through an expansion of employment. Causation runs from (a) expected final demand to (b) planned output to (c) a derived demand for labor power and inputs to (d) the performance of living labor and creation of value (as well as the transfer to final output of the preexisting value of inputs) to (e) an adjustment of actual output to demand to (f) a change in capacity utilization to (g) a change in the actual rate of profit.
The relationship between investment driven by an accelerator mechanism and the expected rate of profit can be somewhat different, depending on what time frame we have in mind. At one extreme, short-term expectations are likely to follow recent behavior of the actual rate of profit. At the other extreme, firms, in the very long run, will expect the average rate of utilization to be normal (u = 1). The expected rate of profit will then reduce to re = πΘ, with the expected rate of profit reflecting the anticipated long-run behavior of distribution (π) and technology (Θ). Within a stable institutional setting, the expected rate of profit over a long time frame might change only slowly in line with gradual, anticipated changes in the profit share and normal output-to-capital ratio.
Between these two extremes is an intermediate time frame in which firms might expect utilization to remain persistently above or persistently below normal for quite some time, perhaps because the economy is perceived to be transitioning from a low to high trend growth rate, or vice versa. If so, the actual rate of profit will fluctuate during the transitional phase around a medium-term average that differs from the rate associated with u = 1.
In terms of investment, the appropriate time frame is the expected life of the plant and equipment to be installed. This suggests a fairly long time horizon for most physical items and perhaps a medium-term outlook for software.
Suppose, still holding technology Θ constant, that there is an exogenous reduction in the profit share π. One way to think of the effect of this is that it will lower the range of profit rates to be delivered by variations in capacity utilization u. So long as the new distribution holds, rates of profit and capacity utilization will rise and fall together. But, for any given rate of utilization, the rate of profit will be lower than it would have been before the change in distribution.
This suggests a question: will capitalists be as likely to invest under the new distributional regime as under the old? We can break this down into two smaller questions:
- What is the effect of the distributional change on demand and capacity utilization?
- Will a persistently high rate of capacity utilization induce investment just as reliably in a low-profit regime as it would in a high-profit regime?
From the perspective of Marx, the second question could similarly be posed in relation to the normal output-to-capital ratio Θ. If, as Marx argued, there is a tendency for the organic composition of capital to rise during an extended period of accumulation, this would imply (for a given rate of surplus value) a falling normal output-to-capital ratio over a persistent growth phase. A study by Andrew Kliman (2011, The Failure of Capitalist Production, Pluto Press, London, p. 133) suggests that the organic composition has risen in the US economy by about 1.5-1.7 percent a year since the Second World War. If so, this will have had a dampening effect on the rate of profit. Here, too, it is relevant to ask whether high rates of capacity utilization (and so relatively high rates of profit within a low profit-rate regime) will be enough to induce private investment.
Distribution and Capacity Utilization
Kalecki offered one possible framework for thinking about the impact of distribution on income growth and hence rates of capacity utilization. Assume that workers, in aggregate, consume an amount equal to their wages and that demand can be fully met at current capacity due to the intentional maintenance of spare capacity and presence of underemployed and unemployed workers. Suppose, also, that the economy comprises three ‘departments’. Department 1 produces ‘investment goods’. Department 2 produces ‘luxury goods’ for capitalist consumption. And department 3 produces consumption goods (or ‘wage goods’) for workers.
Kalecki pointed out that if we imagine an all-round (i.e. global) increase in wages with no initial change in prices or investment, the effect would be temporarily to raise the profits of firms supplying wage goods while causing an exactly offsetting fall in the profits of firms producing investment and luxury goods. The reason for this is that the profit of the wage-goods sector is equal to the consumption expenditure of workers employed in the other two sectors. But what capitalists in the wage-goods sector gain in profit is offset, in aggregate, by what capitalists in the other two sectors have to pay in higher wages. For capitalists as a whole, the initial effects cancel out.
Note, however, that there is an assumption that the distributional change in favor of workers does not provoke an immediate negative investment response. As Kalecki points out, if capitalists did cut back investment, this would alter the initial impact of the wage increase. Kalecki, for his part, contends that an initial reduction in investment is unlikely. He reasons that the level of investment planned for the current period will, on the whole, already be set, based on orders for investment goods that have been placed prior to the distributional change. This relates to the lags between initial investment decisions, the placement of investment orders and the time it takes to produce the investment goods and arrange final delivery. In Kalecki’s view, the level of (current-period) production in the investment-goods sector will, to a large degree, already have been determined by prior decisions to invest and place investment orders.
Even if there is time for investment orders to be canceled in response to the wage increase, it is not entirely clear that they will be. If demand for investment goods is understood to be a derived demand (derived, at least ultimately, from the production requirements of firms in the consumption-goods sectors, of which the wage-goods sector is the largest), then it seems at least plausible that investment will not be negatively affected by the distributional change. The stronger demand for worker-consumption goods created by the wage increase may, in due course, induce a rise in demand for investment goods, and, as profits increase, a subsequent additional demand for luxury goods. The ultimate effect may be an increase in both consumption and investment. Of course, relative prices would likely change as well to spread the impact on profitability more broadly across the three departments of the economy.
The notion of investment as a derived demand may seem more in line with Keynes than Marx. But it is not necessarily clear that Marx would disagree. A previous post, Is Demand for Investment Goods a Derived Demand in Marx, considers a passage from Capital, Vol. 3, Ch. 18, on this question:
[A]s we have seen (Book 2, Part III), continuous circulation takes place between constant capital and constant capital (even regardless of accelerated accumulation). It is at first independent of individual consumption because it never enters the latter. But this consumption definitely limits it nevertheless, since constant capital is never produced for its own sake but solely because more of it is needed in spheres of production whose products go into individual consumption.
Whatever Marx’s view, if investment is understood to be a derived demand, it seems at least possible to conceive of a process in which the ongoing adjustment of wage-goods production to the consumption demand of workers requires additional investment sufficient to maintain planned margins of spare capacity.
Government Expenditure, Utilization and the Inducement of Private Investment
In the argument developed so far, the possibility of growth alongside high wages has been depicted as resting crucially on the initial response of private investment to the distributional change. Although Kalecki’s temporal argument, highlighting the effect of lags, is plausible, it might be that capitalists anticipate the distributional change in advance, or simply fear such a change and have these fears confirmed. If so, their response might be other than benign, and the impact of the global wage increase indeterminate.
The role of government expenditure, in this context, seems key. By adding to demand without directly adding to private-sector productive capacity, government expenditure promotes fuller utilization of capacity. A currency-issuing government can always ensure that its expenditure consistently grows at a pace sufficient to support income growth appropriate to the economy’s capacity to respond at more or less stable prices. This, combined with lower marginal propensities to leak to taxes and saving associated with the distributional shift in favor of workers, can encourage growth. Provided government persistently plays this role, a period of weak private investment (perhaps due to capitalists attempting an ‘investment strike’ in protest over the wage increase) will result in gradually mounting rates of capacity utilization until, at some point, adjusting output to demand will require installation of additional capacity.
Although the rate of profit associated with any given rate of capacity utilization will be lower than before the global wage increase, the capitalists’ hand may actually be forced at this point. Failing to expand capacity when demand warrants it will mean missing out on potential profit. At least within a competitive setting – in which firms face active competitors as well as the threat of new entrants joining the industry – it can also mean losing market share to rivals.
It may be that capitalists try very hard to prevent a distributional shift in favor of workers, but once a universal wage increase goes through, will not cut off their noses to spite their faces. If low profit rates are all that are on offer, then low profit rates are what capitalists will compete over. No doubt, in the meantime, they will do everything in their power to undo the distributional change through political and industrial means.
The period from the end of the Second World War until the present may offer some historical support for this idea. Although workers won some important concessions after the Great Depression and again after the Second World War, growth rates in many countries were relatively strong in the immediate postwar period. Gradually – and increasingly so in the neoliberal period to follow – capitalists were able to wrest back their ascendancy in the class struggle and, with the aid of compliant governments, dismantle some of the concessions previously granted to workers.
Government as Autonomous Spender
The foregoing discussion suggests a significant role for fiscal policy in promoting economic growth. In particular, it highlights the relevance of government expenditure as a major source of ongoing autonomous demand that does not directly create private-sector productive capacity. A government that spends a fairly stable and sizable proportion of GDP, and steps up this spending in absolute terms over time broadly in line with a feasible rate of economic growth, will do a lot to induce other expenditure, both private consumption and private investment.
Fiscal policy, conducted in this fashion, serves a somewhat stabilizing function in the economy provided the growth of government expenditure is not excessive in relation to the capacity of the economy to respond at low, stable rates of inflation. More generally, stability, to the degree it is experienced under capitalism, appears to come from four main sources, three of which are exogenous to the circular flow of income:
- The government’s role as persistent autonomous spender;
- The automatic-stabilizing effects of taxation in which tax revenues rise and fall endogenously with income;
- The government’s role as regulator of wages and working conditions as well as enforcer of the ‘rules of the game’;
- Other institutions promoting stability in social relations.
Since consumption out of wages is relatively stable, whereas private investment behavior is more volatile, institutions and regulations that prevent a free-fall in wages help to provide a somewhat stable demand support to the system.
Despite the presence of exogenous stabilizing influences, however, capitalism as a whole is always prone to instability. Cyclical behavior, even if around a rising trend, is to a degree unavoidable because of the role and effects of private investment within a capitalist economy. Put simply, when extra investment is called for, because of above-normal capacity utilization, the investment expenditure adds further to demand at a time when overall demand is already strong, while adding to capacity only after a time lag. The effect is temporarily to push demand nearer to, and possibly beyond, practical capacity limits. As a result, there can be some pure inflation (in which prices inflate away demand that cannot physically be met with an expansion of production). Then, once the additional capacity has been installed, private investment, and therefore overall demand and the utilization rate, will suddenly drop off. In this way, the economy will typically oscillate around a trend, where the trend is determined by non-capacity-generating autonomous demand – most notably government expenditure – and the fluctuations are due to the demand and capacity effects of investment.
A very stylized representation of this process is shown in the diagram. Government expenditure G – taken for simplicity to be the only form of non-capacity-generating autonomous demand – is intentionally stepped up at a constant rate over time. Assuming, again for simplicity, that there is no change in the propensities to leak to taxes and saving, the government expenditure has a stable multiplier effect in which private consumption expenditure is induced. When output Y is pushed beyond normal, implying a capacity utilization rate greater than 1, private investment is induced as firms attempt to restore the normal degree of planned spare capacity. It is supposed that so long as actual output remains below the absolute maximum possible, the general price level can be taken as given, but that if demand pushes actual output beyond this point, prices will rise to inflate away the difference. Alternatively, there could be some non-price rationing. Once the additional capacity is installed, investment falls away for a time, and this has a negative impact on overall demand, causing actual output temporarily to fall below normal. However, if the government continues to step up its spending at a steady pace consistent with the intended rate of growth of real output, this along with the consumption expenditure it induces will once again push actual output toward and then beyond normal. And so on.
In theoretical terms, the instability is connected to Harrod’s ‘knife edge’. Actions to adjust capacity to demand temporarily push demand in the wrong direction, before the capacity effects of investment have had time to materialize. A stabilizing force is autonomous expenditure that does not directly add to private-sector productive capacity. For simplicity, Harrod abstracted from this form of expenditure. To the extent firms operate with planned margins of spare capacity, these autonomous expenditures can drive growth without necessarily causing pure inflation, and the system is able to exhibit a degree of stability. The knife edge is still at hand, though somewhat blunted. Franklin Serrano makes this point very clearly in his seminal analysis of the ‘supermultiplier’. Cycles were abstracted from in his work, so as to place the focus entirely on the long-period aspects of a demand-led growth process, but the resolution to Harrod’s kinfe-edge problem is essentially the same with or without cycles.
We have seen in the period leading up to the global financial crisis of 2007 that others can also attempt to be the source of non-capacity-generating autonomous demand when governments fall down on the job. But this other source – private autonomous consumption expenditure financed through private credit creation – is not a sustainable means of maintaining healthy growth in autonomous demand and, via the multiplier (and supermultiplier), economic growth. The process hits its limit when debt-servicing requirements go beyond the capacity of consumers to pay out of income.
For a currency-issuing government, it is different, because it faces no revenue constraint.
Even if the above argument is valid, there is, of course, an all-too-familiar political difficulty. Capitalists exhibit a clear preference for a growth strategy that centers on private debt-financed consumption rather than fiscal policy and worker-friendly regulations. The postwar class compromise only came about through massive mobilization of workers and the mounting of immense political and industrial pressure for progressive change. Ever since, capitalists have pushed back hard, putting us in our present malaise. Capitalists prefer a system based on private debt at least partly because:
- Finance capital can extract higher interest from private debtors than from currency-issuing governments. Any payment of interest by a currency-issuing government is purely at its own discretion, and can be zero (e.g. employing overt monetary financing or zero interest on reserves) if so desired.
- Capitalists prefer it when workers are docile and compliant. Private indebtedness helps to keep them so.
The political opposition to growth induced by government expenditure rather than private consumer debt underscores the need for grassroots efforts capable of pushing government policy in a more desirable direction.