Mario Seccareccia*







May 1999



*Full Professor, Department of Economics, University of Ottawa, Ottawa, Ontario, Canada (K1N 6N5); e-mail: <mseccare@uottawa.ca>. This paper was presented at the CEA-PEF meetings at the University of Toronto on May 30, 1999.




Introduction: Some Notes on Keynes's Original Policy Views to Achieve Full Employment

In commenting on the socio-economic realities of mass unemployment of the 1930s, John Maynard Keynes (1936: 373) wrote: "The outstanding faults of the economic society in which we live are its failure to provide for full employment and its arbitrary and inequitable distribution of wealth and income." To both Keynes and numerous post-Keynesian writers of the post-war period, full employment was the necessary prerequisite to the achievement of a more equitable distribution of income (and wealth). Indeed, as the point is also forcefully made by Arestis and Sawyer (1998: 181), the two goals could (and should) not be seriously dissociated. This is because, to achieve full employment, policies should be put in place whose impact would be inevitably to lead to income and wealth redistribution.

As it has been discussed in much greater detail elsewhere (Seccareccia 1993, 1994, 1995; and Seccareccia and Lavoie 1989), in the General Theory, Keynes had proposed a short ensemble of specific policies to achieve full employment which, in contrast with the textbook caricature of Keynesian short-term policies of "pump-priming", were of a long-term nature and whose impact would be on the long-term evolution of the functional distribution of income. These policies pertain to what Keynes had described as the "socialization of investment" and the "euthanasia of the rentier", that he believed would impact both upon the size of the economic surplus as well as on how that surplus would be divided up among social classes in a capitalist economy. For example, Keynes's notion of socializing investment is best understood in his support of the setting up of a National Investment Board (NIB) throughout most of the 1930s and early 1940s. The alleged purpose of such a Board was to achieve full employment by strategically regulating the aggregate flow of investment expenditures through the appropriate control of long-term financing. When such a NIB policy of directly intervening in the financial capital markets by redirecting investment were to be coupled with a monetary policy of pegging nominal interest rates at a very low level, the result would be that the economy would be hovering about the full employment level and, at the same time, massively redistributing income away from the rentier sector, thereby leading to the so-called "euthanasia of the rentier".

To our understanding, nowhere had Keynes shown interest in achieving full employment by means of short-term policy of "functional finance" of the type first put forth by Abba Lerner during the 1940s. As pointed out by Kregel (1985), Keynes regarded operating deficits in the "normal" or "ordinary" budget of the state as a visible sign of the government's lack of a strong public investment policy to achieve full employment by means of an activist NIB. For this reason, he was careful to distinguish between the current and the capital accounts of the state. Keynes's position was that, in an expenditure-led high growth economy propelled forward by the activity of the NIB, governments should aim at running surpluses in their "ordinary" (or recurring) budgets and that these surpluses should be "transferred to the capital Budget, thus gradually replacing dead-weight debt by productive or semi-productive debt." (Keynes 1980: 277). Hence, just as a private investment ultimately bears fruit for business enterprise, such a comprehensive public investment policy would pay for itself over time by engendering surpluses in the government's operating budget. Unlike private firms, however, governments would utilize this surplus not to extinguish their debts but to expand their capital expenditures strategically. The effect of this continuing transfer of operating surpluses to capital budgets would be gradually to socialize a growing proportion of total investment expenditures. Indeed, he argued that these public capital expenditures should be significant enough to achieve full employment by stabilizing aggregate investment spending over the business cycle and, in the long run, they should be "maintained at a level which will allow the growth of capital to the point where it ceases to be scarce." (Keynes 1936: 376) To this end, he eventually envisaged as much as two-thirds or three-quarters of total investment as being under the influence of the public authorities. The implication of all of this is that Keynes's original position (with regards to policies to achieve full employment) was concerned not only with the level of public spending but also with its composition over the long term. Moreover, by means of what may be dubbed a "reflationary" growth process, his public expenditure policy had clear long-term objectives in terms of income distribution: the gradual elimination of rentier income.

The Neoclassical Response: Income Supplementation Programmes without Full Employment

Keynes's radical policy system would neither be favourably received by the economic establishment of his times nor of ours. As Hayek (1966: 243) put it so succinctly, Keynes was reviving the most naive "inflationist fallacies" that no scarcity exists in the labour market (except at full employment). Despite the strong neoclassical opposition, governments both during W.W.II and the early post-war period did come to subscribe to some aspects of Keynes's thinking, largely because of conjunctural factors pertaining to the balance of political forces at the time that gave rise to what some may describe as a Fordist social structure of accumulation (cf. Paquette 1999: 967-71). As a byproduct of the early post-war political commitments to full employment in Western Europe and North America, the international policy drift in favour of high employment by means of public investment and low interest rate policy did follow a general policy pattern quite compatible with Keynes's vision. However, already in the late 1950s, growing fears of a "creeping inflation" associated with high employment brought many policy makers (as noted by Lerner 1963: 229) to "retreat from Keynesianism". These fears, especially as it came to be crystallized in the Phillips curve trade-off menu and subsequently in the Friedmanite natural rate hypothesis, ultimately came to spell the end of activist state policy to achieve full employment. As has been witnessed over the last three decades, policy thinking moved full circle back to the neoclassical supply-side doctrines of the 1930s. Once again, the only appropriate role of government has now become that of ensuring its long-run monetary and budgetary neutrality and actively to remove all institutional obstacles in the labour market that prevent wages from falling in order to reduce the equilibrium level of unemployment.

As understood within neoclassical theory, unemployment is the result of structural rigidities that do not allow the labour market to perform its functions in terms of allocation and clearance. Consequently, any macroeconomic policy that does not take heed of these underlying functions would generate inflationary pressures. Instead of seeking to boost progressively aggregate demand whose ultimate repercussion would supposedly be accelerating inflation, it is by dismantling various institutional obstacles in the labour market that would bring the equilibrium unemployment closer and closer to its full employment level (cf. Seccareccia 1991a, 1991b, and Bougrine and Seccareccia 1998, 1999). If problems of income distribution and poverty were to arise from the normal functioning of an unhindered and flexible labour market, the solution would be to set up a redistributive system that is presupposed to be least distortive in the sense of permitting wages to serve as the primary allocative and market-clearing device and, at the same time, not have the inherent inflationary features that were assumed to accompany Keynesian full employment policies. It is in this context that the various programmes of income supplementation such as Friedman's negative income tax and other notions of guaranteed income (GI) came to be seen as substitutes for an activist Keynesian policy of achieving full employment. As Friedman (1962: 191) pointed it out, unlike minimum wages and other similar income support systems such as unemployment insurance, a GI system would not much "distort the market or impede its functioning." This is because a GI policy would create incentives for individuals to offer their labour services while, at the same time, provide a basic income that would not hinder wage flexibility in the labour market.

The idea of a GI system is quite simple. In an unfettered labour market tending towards full employment, all individuals ought still to be guaranteed a minimum basic income (MBI). However, instead of most current transfer systems which allow individuals to receive either welfare payments or work (but not both), thereby creating work disincentives, a GI system would supplement a low-wage job by taxing back only a fraction of the GI transfer. The idea is appealing since it does concern itself with some income security objectives not presently met by numerous current welfare systems, especially with regards to the working poor. However, as it has been argued elsewhere (see Iacobacci and Seccareccia, 1989, and Seccareccia 1991a), such as system would only be desirable in an ideal neoclassical world in which full employment is presupposed to have been reached. On the assumption that the minimum basic income would be set close to what are present welfare transfers to the poor and in the absence of a comprehensive minimum wage legislation and Keynesian fiscal and monetary policies to achieve full employment, the effect would be significant wage deflation with the accompanying proliferation of low wage jobs. This is because the GI policy would significantly increase the number of employable welfare recipients seeking work in labour markets already characterized by chronically high unemployment. Due to the increased propensity to accept lower-paying jobs that would be generated by what in Iacobacci and Seccareccia (1989) was described as the "compensation effect" of the income support provided by the GI programme, such a policy would only intensify the downward pressures on wages and activate forces pushing towards a low-wage/part-time economy. Of critical importance to us, at the macroeconomic level, the adoption of such GI programmes would have an important effect on the responsiveness of wages to the business cycle. As cyclical unemployment rises, the number of workers willing to accept lower-paying jobs or a cut in wages (if already employed) would also rise, due to the compensation effect. In the process, the neoclassical flexible wage system would be rehabilitated; but, as Keynes (1936) had argued, flexible labour costs do not move the economy closer to full employment. Because of the negative feedback effects on aggregate demand, just the reverse would be true, especially if the gap between wages and profits might be widening in the long run owing to the proliferation of low wage jobs.

These wage deflating tendencies which would result from the above-mentioned compensation effect could exist regardless of the guaranteed minimum basic income (MBI) that would be set by the government. However, the higher the MBI, the lower would be the participation rate on the part of households, and thus the lower ought to be the underlying labour market pressures to bring down wage costs. As shown in Figure 1, where the MBI is measured on the vertical axis and the percentage employment (e) and participation (l) rates are measured on the horizontal axis, even if the employment rate remains insensitive to a small change in the MBI (because of the usual explicit assumption of "fiscal neutrality" of the GI programme [for a discussion, see Iacobacci and Seccareccia 1989: 162-63]), if the participation rate falls as the MBI rises, this would ultimately bring down the unemployment rate, reflected in the narrowing gap between e and l. In reality, however, labour demand (e) would probably not be inelastic to changes in the MBI. As the MBI moves upward, it would likely have a destabilizing effect on the wage structure as firms offering wages below the MBI would be under growing pressure to increase them, especially if labour becomes progressively scarce as the labour force participation rate (l) falls. To the extent that this positive effect on overall wages becomes significant (as the MBI rises), we could trace an upward-sloping adjusted e' curve reflecting the higher macroeconomic demand for labour arising from the increased consumption demand that ought to accompany the higher wages. It follows, therefore, that the choice of the minimum basic income is a critical factor in determining how significant would be the wage deflation bias in the operation of a GI programme. As the value of the MBI approaches MBI* (or MBI*') in Figure 1, it is likely that the wage deflating influence of the GI programme (due to the compensating effect) would be progressively outweighed by upward labour market pressures on wages as firms compete for qualified workers. However, as much as one may theorize about these latter effects, the dangers of inflation would be incredibly remote under current GI proposals. This is because all GI programmes are normally packaged so that the MBI is sufficiently low, usually below the poverty line and, indeed, even below existing social assistance levels (Iacobacci and Seccareccia 1989: 147-48). For this reason, not only do all proposed GI programmes meet the criterion of zero inflationary pressures that is of concern to neoclassical theorists but, as has been argued, the adoption of such programmes would have a clear deflationary bias in the labour market.

Figure 1

Relation between MBI and Employment/Participation Rates


















Government as Employer of Last Resort: An Alternative to the Deflationary Bias of Neoclassical Policies?

Over the last few years, a powerful alternative policy view (in which government is assumed to act as employer of last resort [ELR]) has emerged which completely stands on its head the neoclassical policy dilemma. Instead of the presumed incompatibility between inflation and unemployment that is at the core of the neoclassical policy perspective, advocates of ELR, such as Mosler (1997-98), Wray (1997, 1998a, 1998b), Forstater (1997) and Mitchell (1998), agree with Keynes in pointing out that true full employment is not only desirable but also achievable without the inflationary effects that are assumed to accompany it. Much like Keynes, full employment must be attained by means of demand-side policies and not through tinkering on the supply-side of the labour market. Furthermore, the whole notion of monetary/fiscal neutrality, so critical to the neoclassical policy system, is completely rejected since permanent deficits financed through money creation would be entirely compatible with price stability within the ELR framework (cf. Parguez 1998).

According to Wray (1998a: 125-26), an embryo of the ELR proposal can be found in Minsky (1986: 308), though one can probably trace ELR roots back to the early works of Abba Lerner in the 1940s. Not unlike the GI programmes, the basic structure of an ELR policy is also quite simple. As stated by Wray (1998a: 124), it ought to rest on two key features: (i) there would exist a perfectly elastic demand for labour by government at a given ELR wage and (ii) the basic ELR wage would be a discretionary variable set exogenously by the state. Given the vagaries of private sector employment, the ELR state would be prepared to hire all unemployed workers shed by the private sector at the basic ELR wage, thereby guaranteeing effective full employment.

To analyze the implications of ELR, it would be useful to address our questions within a particular Post-Keynesian or Kaleckian framework of analysis to be found in Eichner (1987), Nell (1988), and, among others, Seccareccia (1991b) and Bougrine and Seccareccia (1998, 1999). In a non-ELR world, let us assume that total income, Y, depends on aggregate expenditures which can be conveniently divided up into the discretionary (A) and the non-discretionary expenditures (C):

Y = C + A (1)

where A = I + D + Xn, and where I = private sector investment expenditures, D = government (non-ELR) budget deficit, and Xn = net exports. At the same time, let us suppose that total income, Y, can be classified as employment income, wL, and property income, B, so that:

Y = wL + B (2)

where w is the average wage rate of the combined private and public sectors, L is total employment and B is the flow of profit, interest and rent. Adopting a simple classical savings function whereby the propensity to consume out of wage income (Fw) is greater than the propensity to consume out of property income (FB), or in the extreme case that will be considered here FB = 0, we obtain that C = Fw(wL). Hence, introducing this assumption into equation (1) above, we have:

Y = Fw(wL) + A (1)

Finally, specifying a simple Leontief linear aggregate production function such that Y = aL, where a is average labour productivity, we arrive at the following fundamental equation:

aL = Fw(wL) + A (3)

and which, when re-organized, gives the aggregate labour demand curve for (non-ELR) labour at the macroeconomic level:



with the usual Keynes/Kalecki property that MLd/MW>0 and M2Ld/MW2>0; that is to say, that in contrast with neoclassical theory the wage-employment relation is necessarily positive and, in this case, also non-linear.

As can be found inter alia in Nell (1988), we can illustrate graphically this aggregate labour demand curve as in Figure 2. Any increase in A or Fw, or any fall in a, would have the effect of shifting the Ld curve outward to the right and/or modifying its slope. Introducing an inelastic labour supply curve (Ls), we can thus derive the usual Post-Keynesian results. At a given average wage, w0, associated with an excess supply of labour, there exists no self-correcting mechanism to bring the economy to its full employment level at the intersection point A. Indeed, if wages were flexible downward in the sense favoured by neoclassical theory, the effect of a fall in wages would be merely to make matters worse on the unemployment front. Hence, unemployment is not the result of wages being too high. Rather it is because of a lack of aggregate effective demand due to the fact that, for a given flow of autonomous expenditures, wages are too low. Consequently, only by an increase in discretionary spending, A, and/or an accompanying rise in real wages, w, can an economy reach full employment.

Figure 2

Macroeconomic Labour-Market Relation















Before analyzing ELR policy within this framework, let us discuss more carefully the nature of the aggregate labour supply curve. It is well know on the part of neoclassical economists that labour supply (i.e., labour force participation) is not completely inelastic to movements in the real wage (see Killingsworth 1983). If one were either to conceive of it historically or compare economies internationally, it is well known that in developing economies labour force participation rates would be relatively low (that is to say, there would exist vast labour reserves), and the desire to attain the high consumption levels of developed countries would be very high. In such economies, characterized by low real wages and high levels of disguised unemployment, labour supply would be sensitive to real wage movement and positively sloped, in conformity with traditional analysis (cf. Wisman 1989: 104-5). For instance, over the last two decades, the unemployment rate of a country such as Mexico has hovered around the 3.5 per cent value with forecasted rates of about 3.5 per cent for 1999 and 2000 (See OECD Economic Outlook, December 1998). Undoubtedly, such a low official unemployment rate masks a high degree of disguised unemployment revealing vast potential reserves of underutilized labour that would quickly decline as real wages rise. However, as the real wage rises, households would be attaining progressively their targeted consumption norms (à la Duesenberry) and, concomitantly, those substantial labour reserves would become depleted. At sufficiently high real income levels, the labour supply curve would become more inelastic and could eventually become backward bending. Indeed, this latter phenomenon is actually confirmed by household behaviour in North America. Real wages have generally fallen since the 1970s but labour force participation rates have risen dramatically throughout this period.

In Figure 3, we have traced a labour supply curve that depicts these specific features discussed above. Since the labour supply curve intersects the labour demand curve at two separate points, one may thus conclude that an economy can achieve full employment at two distinct levels: at a low wage level, wB, and employment level, LB, and at a high real wage, wA, and employment level LA. The intersection point B in the diagram represents a "Third World" type of solution to achieve full employment representing economies studied historically by development economists such as W.A. Lewis (1979), in which massive disguised unemployment (LA - LB) is the norm (cf. Cornwall 1977). On the other hand, the intersection point A represents a hypothetical state of veritable full employment for an advanced capitalist economy in which those potential labour reserves have essentially dissipated. As studied by Keynes during the 1930s, an advanced capitalist economy that must face a chronic shortage of effective demand would normally find itself within the shaded area of Figure 3 between the intersection points A and B, that is to say, in a state in which substantial involuntary unemployment is the norm.

Figure 3

Low-Wage/High Wage Full Employment and

Involuntary Unemployment

What would be the implication of an ELR programme in such an economy? Figure 4 shows how at w*ELR the programme would now set a floor so that the effective labour supply curve would resemble an inverse L-shaped relation. [The reader should take note that w*ELR represents the ELR-weighted average wage for the economy, inclusive, therefore, of the exogenously set ELR wage.] At this level (which was selected purposely for diagrammatical illustration), it coincides with the maximum stock of involuntary unemployed labour that would have to be absorbed by the ELR programme. The economy would thus be at its tangency point E in the diagram. At any other level of ELR-weighted wage between wA and wB, the ELR-adjusted labour demand schedule would cross the labour supply curve, thereby requiring less spending to absorb the lower volume of involuntary unemployed associated with it, say, at E'. There thus emerges an apparent trade-off from our diagram. Any wELR either greater or less than w*ELR would entail proportionally less expenditures on the part of the state to eliminate the involuntary unemployment. However, while the financial costs might be proportionally lower, the social costs would be very different as to whether one would choose the top or bottom side of the threshold level, w*ELR. The lower is the selected wELR that would move the economy closer to the low-wage full employment point at wB, the higher would be the level of disguised unemployment (or rate of underutilization of productive capacity) in an economy. This is because, barring forced labour (which is outrightly excluded [see Wray 1998a: 124]), the state presumably would not be able to constrain people to work for what some workers would deem to be an unacceptable wage. Naturally, on the top side of the threshold level w*ELR, the higher is the ELR wage and, once again, the lower is the volume of involuntary unemployed to be absorbed by the ELR programme. In contrast with the low-level ELR, however, the lower would be the number of disguised unemployed connected with this high average ELR-weighted real wage level. Finally, as the real wage continues to rise, the economy would reach the upper turning point of the labour supply curve in our diagram at intersection A. At this high end of our aggregate real wage scale, full employment is reached without the need for ELR jobs and in which no significant labour reserves exist.

Figure 4

ELR Full Employment










From this analysis alone, the choice of wELR becomes highly critical. One could imagine two types of effective full-employment systems under ELR: a low-wage ELR programme (presumably enacted by a right-wing government) with wELR gravitating around the bottom end of our relevant average real wage spectrum, and a high-wage ELR policy (possibly decreed by a social democratic regime) with less wastage of human resources due to the lower levels of both "involuntary" ELR workers and disguised unemployment associated with wELR>w*ELR. Given the critical role of the ELR wage, it is rather surprising that advocates of ELR have shown little concern with this thorny issue of what actual type of full employment is socially most desirable. Should it be a low-wage ELR system tending towards wB that would be accompanied by a large wastage of human resources or should it be a high-wage full employment system with a wELR approaching wA?

In going through this simple exercise, an obvious question that perhaps ought now to strike the reader is why go through this process of setting up an ELR programme in the first place. Surely, a Keynesian policy of high public investment, low interest rates, and high real wages (as described in our introduction) would take the economy close enough to the intersection point A in any case. Much like conventional wisdom, the answer that is provided by some defenders of ELR is that a Keynesian full employment policy would necessarily be inflationary as in a Phillips curve world (cf. Mitchell 1998). To his credit, Wray (1998a: 153, fn. 29; and 1998b) himself, is quite careful at times in stating that he does not necessarily endorse such an explanation of the inflationary process. However, while distancing themselves from the neoclassical explanation of inflation, it is quite clear that advocates of ELR are catering to those numerous mainstream economists who believe in a NAIRU whereby inflation is seen as the ineluctable consequence of Keynesian "tight" labour market policy (even though this mainstream view is highly disputed and, to a large extent, has been discredited [Galbraith 1997]).

If one believes in some Phillips curve process, then what is needed is a policy system that would recreate "loose" labour market conditions even at full employment. Does an ELR system provide such a "loose" labour market environment? It all depends, once again, on the choice of wELR. If one chooses the low-wage ELR regime, there would be some pressure to bear on wages to move downwards toward the low ELR anchor. For instance, in a period of fiscal austerity in the public sector, what would prevent a local government concerned with the size and cost of its own employed labour force from shedding some of its workers with relatively higher wages and trade union protection and replacing them with low wage ELR workers? "Workfare" policies may, to a certain extent, be envisaged in this light, i.e. as an attempt by certain governments in North America to cut public expenditures while trying to maintain some low-end public services. Wray (1997) does seem to recognize this danger; yet there appears to be nothing that is built in the ELR system to prevent such a deflationary wage effect to occur in the public sector. In much the same way, a low-wage ELR regime could destabilize downwards private sector wages if wELR is set below preexisting statutory minimum wages. Admittedly, the political prospects of this happening would be slight, but not impossible. However, if wELR is set low enough, it is not only the weight of the "buffer stock" of ELR workers that may pull other wages (situated further up the wage scale) downwards but, as many ELR analysts seem to forget, it is also the weight of the army of disguised unemployed as the chosen wELR is set closer to wB in Figure 3. Because the ELR wage sets a floor to wages, the deflationary effect would never be as strong as with the neoclassical GI systems (cum involuntary unemployment), but such an effect cannot be excluded.

Would there be a bias in favour of wage inflation under an ELR system? Once again this would depend to a large extent on the choice of wELR. The higher becomes the ELR wage (in Figure 4), and the more destabilizing would the implementation of such a programme be on the wage structure. As Keynes had understood it long ago, workers are perhaps more concerned with their relative wages than with the level of prices in the economy. Owing to these powerful normative forces in the labour market which make for a constancy of relative earnings (cf. Seccareccia 1991a: 48-49), if wELR is set relatively high, the effect would be not unlike what would happen if a statutory minimum wage is raised significantly in a fully employed economy. The effect would probably be to provoke what some may describe as a "wage-wage" spiral via the "leap-frogging" effect that it would generate. Hence, independently of the size of the ELR pool of workers associated with a given wELR, an inflation of the type which, for instance, Kaldor (1976: 708) had identified as best representing the "creeping inflation" of the 1950s and 1960s would be the most likely outcome of introducing an ELR system with a wELR that is "too high". Wray (1997, 1998a, 1998b) has insisted, however, that such a process would not be an inflation. It would produce merely a once-and-for-all jump in wages; it would not generate a "continuous pressure" on wages and prices. Regardless of one's terminology, to the extent that the setting of wELR generates an instability in the wage structure that may take a long time to wind down, economists have normally described this as an inflationary process. As shown in Figure 5, on the right-hand quadrant, we have depicted the possible relation between the exogenous setting of wELR and wage inflation, )w/w. The point of intersection on the wELR axis represents the ELR wage, w**ELR, that would preserve relative stability in the wage structure, and therefore in "wage-wage" inflation. Analogous to Galbraith's (1998) "ethical rate of unemployment" that stabilizes earnings relativities, we could conceive here of the "ethical" wELR that would least disturb relative wages.



Figure 5

ELR Anti-Inflation Policy








For the sake of the analysis, let us now assume that ELR policy makers (who have chosen a given w0ELR in Figure 5) are not satisfied with the presumed outcome in terms of generating some given level of wage inflation. The solution that would be available under ELR to deal with this undesirable inflation would be a very traditional one. In essence, government could either raise taxes and/or cut non-ELR expenditures or raise interest rates to slow down private spending so as to increase the "buffer stock" of ELR workers. Just to quote Mitchell (1998: 552):

"In the face of wage-price pressures, the BSE/ELR approach maintains inflation control in much the same way as monetarism)by chocking aggregate demand and introducing slack in the non-buffer stock sector."

In terms of Figure 4, the idea would be that it would shift the non-ELR aggregate demand curve to the left; or, in terms of Figure 5, this would entail an upward shift in the negative relation on the left-hand quadrant between wELR and rate of ELR employment (as a proportion of the working-age population) plus the number of disguised unemployed (DU) [also as a proportion of the working-age population], and simultaneously an upward shift in the wELR and )w/w relation on the right-hand quadrant, that is, up to the point where w0ELR is now compatible with a zero wage inflation.

Under ELR, the government would thus have two tools at its disposal to combat wage inflation: (i) to choose a sufficiently low wELR so that "wage-wage" inflation would not be a significant factor in a fully employed economy; or (ii) if it does not wish to tinker with the wELR, it could wring inflation out of the system by controlling aggregate demand of the non-ELR sector. The first instrument is nothing more than a peculiar type of incomes policy that focuses on regulating the income of the lowest end of the wage scale; while the second is, of course, merely a variant of the Phillips curve or NAIRU in disguise, with the "buffer stock" of ELR workers now replacing the involuntary unemployed as the preferred mechanism of controlling wage and price movements. In the final analysis, it may therefore be argued that the mechanisms in place are not that fundamentally different when comparing the current system of controlling inflation to the proposed ELR system. Within the present institutional arrangements, the state achieves its inflation targets merely by varying the reserve army of involuntary unemployed. In the ELR case, it would be through variations in the reserve army of ELR workers caught in a situation of "loose" full employment. While the latter ELR system can be termed a "Pareto improvement" when compared, say, either to the present wasteland of human resources or even to the deflationary GI alternatives that are sometimes offered by neoclassical economists, one could still legitimately ask the question once again: is this the type of full employment that one would find socially most desirable?


Concluding Remarks

After over a quarter century of being offered nothing but larger and larger doses of austerity in order to fight an anticipated inflation that does not exist, it is truly refreshing that there are economists out there who have had the courage to bring to the forefront once again the full-employment objective. What is at issue among those of us who are strongly commited to full employment is not the goal itself, which neoclassical economists adamantly reject in favour of some elusive NAIRU, but rather what type of full employment do we feel is most desirable for a free and caring society that would like to see involuntary unemployment eradicated. For those of us who believe in a socially responsible economics, this question needs to be seriously addressed.

As it was described in the introduction, in Keynes's original paradigm the full employment goal was not to be dissociated from the income distribution objective. By pegging interest rates at very low levels (consistent with his view on the "euthanasia of the rentier") and with a vigorous public investment policy (his so-called "socialization of investment") that would bring both employment and real wages up to point A in our Figure 3, Keynes was encapsulating these seemingly disparate goals into one key overall objective. Indeed, not only was he linking up full employment with his income distribution objective but, perhaps, even more fascinating, interest rate policy and even a preference for a specific composition of investment (and ouput) were all conceptually intertwined in Keynes's analytical framework. The ELR policy, on the other hand, would be able to achieve admirably the strict target of full employment but only by creating a "buffer stock" of ELR workers that would be needed to keep an effective lid on wage inflation. Instead of a system tending towards Keynes's full-employment point A, we have a system that can be anchored anywhere within the range of points A and B in Figure 3. As a result, this openess leaves room for competing views of ELR and perhaps, more importantly, it leaves entirely open the question of what type of income distribution. Furthermore, unlike Keynes's strong preference for public investment, the issue of the actual composition of ELR spending (and output) is also left completely open within the ELR analytics. Some may see virtue in a policy that allows so much open-endedness. However, without providing clearer objectives, especially on the income distribution side (such as offering strong safeguards against the low-wage ELR variant), one might simply be recreating an ELR system that mirrors the present system. That is to say, we would merely have changed the status of the involuntary unemployed to that of an ELR worker, but the individual's role would be substantially the same)that of preventing wage inflation and the labour market from inching upwards closer toward point A in Figure 3.

Finally, it is also for reasons pertaining to income distribution that an ELR system should not be brandished in the same way as GI systems are frequently presented, that is, as an all-purpose replacement of other existing social programmes. It is quite evident why statutory minimum wages would be redundant under ELR; however, it is not clear why other social programmes, such as unemployment insurance, would also disappear under ELR. With the political dangers that a low-wage ELR system be adopted, it would be important in this case that laid-off workers have a clear choice between ELR work and a multitude of existing institutional forms of social assistance in order to prevent any possible deflationary pressures from taking hold. In more recent publications, defenders of ELR have been more careful not to sell ELR in the same way that GI systems have been marketed historically; but, this was not always so in some earlier presentations of what ELR is intended to replace (cf. Wray 1997).


Arestis, P., and M. Sawyer (1998), "Keynesian Economic Policies for the New Millennium", Economic Journal, Vol. 108 (January), pp. 181-95.

Bougrine, H., and M. Seccareccia (1998), "Assurance-chômage et chômage : analyse des effets sur la demande globale de travail au Canada depuis la Guerre", Vers le plein emploi : pour un renouvellement des politiques publiques, P. Paquette and M. Seccareccia (eds), Montréal: Presses de l'Université de Montréal, pp. 151-75.

Bougrine, H., and M. Seccareccia (1999), "Unemployment Insurance and Unemployment: An Analysis of the Aggregate Demand-Side Effects for Post-War Canada", International Review of Applied Economics, Vol. 13, no. 1 (January), pp. 5-21.

Clark, C.M.A., and C. Kavanagh (1996), "Basic Income, Inequality, and Unemployment: Rethinking the Linkage between Work and Welfare", Journal of Economic Issues, Vol. 30, no. 2 (June), pp. 399-406.

Cornwall, J. (1977), Modern Capitalism, Its Growth and Transformation, London: Martin Robertson.

Eichner, A.S. (1987), The Macrodynamics of Advanced Market Economies, Armonk, N.Y.: M.E. Sharpe.

Forstater, M. (1997), "Selective Use of Discretionary Public Employment and Economic Flexibility", Jerome Levy Economics Institute, Working Paper No. 218, (December).

Forstater, M. (1998), "Flexible Full Employment: Structural Implications of Discretionary Public Sector Employment", Journal of Economic Issues, Vol. 32, no. 2 (June), pp. 557-63.

Friedman, M. (1962), Capitalism and Freedom, (in collaboration with R. Friedman) Chicago: University of Chicago Press.

Galbraith, James K. (1997), "Time to Ditch the NAIRU", Journal of Economic Perspectives, Vol. 11, no. 1 (March), pp. 93-108.

Galbraith, James K. (1998), "The Ethical Rate of Unemployment: A Technical Note", Journal of Economic Issues, Vol. 32, no 2 (June), pp. 531-37.

Hayek, F.A. (1966), "Personal Recollections of Keynes and the 'Keynesian Revolution'", in B. Caldwell (ed.), The Collected Works of F.A. Hayek, Vol. 9, Chicago: University of Chicago Press, 1995, pp. 240-6.

Iacobacci, M., and M. Seccareccia (1989), "Full Employment versus Income Maintenance: Some Reflections on the Macroeconomic and Structural Implications of a Guaranteed Income Program for Canada", Studies in Political Economy, No. 28 (Spring), pp. 137-73.

Kaldor, N. (1976), "Inflation and Recession in the World Economy", Economic Journal, Vol. 86 (December), pp. 703-14.

Keynes, J.M. (1936), The General Theory of Employment, Interest and Money, London: Macmillan.

Keynes, J.M. (1980), The Collected Writings of John Maynard Keynes, Vol. 27, ed. by D. Moggridge, London: Macmillan, pp. 264-419.

Killingsworth, M.R. (1983), Labour Supply, Cambridge: Cambridge University Press.

Kregel, J.A. (1985), "Budget Deficits, Stabilisation Policy and Liquidity Preference: Keynes's Post-War Proposals", in F. Vicarelli (ed.), Keynes's Relevance Today, London: Macmillan, pp. 28-50.

Lerner, A.P. (1963), "Keynesian Economics in the Sixties", in R. Lekachman (ed.), Keynes' General Theory: Reports of Three Decades, New York: St. Martin's Press, 1964, pp. 222-34.

Lewis, W.A. (1979), "The Dual Economy Revisited", Manchester School of Economic and Social Studies, Vol. 47, no. 3 (September), pp. 211-29.

Minsky, H.P. (1986), Stabilizing an Unstable Economy, New Haven, Ct.: Yale University Press.

Mitchell, W.F. (1998), "The Buffer Stock Employment Model and the NAIRU: The Path to Full Employment", Journal of Economic Issues, Vol. 32, no. 2 (June), pp. 547-55.

Mosler, W. (1997-8), "Full Employment and Price Stability", Journal of Post Keynesian Economics, Vol. 20, no. 2 (Winter), pp. 167-82.

Nell, E.J. (1988), Prosperity and Public Spending, Boston: Unwin Hymann.

Paquette, P. (1999), "Regulation Approach", in P.A. O'Hara (ed.), Encyclopedia of Political Economy, London: Routledge, pp. 967-71.

Parguez, A. (1998), "Full Employment, the Value of Money and Deficit Spending: The Theoretical Foundations of the Employer of Last Resort Approach within a Circuitist Framework", Paper presented at the Fifth International Post Keynesian Workshop, Knoxville, Tennessee, (June).

Seccareccia, M. (1991a), "An Alternative to Labour-Market Orthodoxy: The Post-Keynesian/Institutionalist Policy View", Review of Political Economy , Vol. 3, no. 1 (January), pp. 43-61.

Seccareccia, M. (1991b), "Salaire minimum, emploi et productivité dans une perspective post-keynésienne", l'Actualité économique, Vol. 67, no. 2 (juin), pp. 166-91.

Seccareccia, M. (1993), "On the Intellectual Origins of Keynes's Policy Radicalism in the General Theory", History of Economic Ideas , Vol. 1, no. 2, pp. 77-104.

Seccareccia, M. (1994), "Socialization of Investment", in P. Arestis and M.C. Sawyer (eds.), Elgar Companion to Radical Political Economy, Aldershot, England: Edward Elgar Publishing Ltd., pp. 375-80.

Seccareccia, M. (1995), "Keynesianism and Public Investment: A Left-Keynesian Perspective on the Role of Government Expenditures and Debt", Studies in Political Economy, No. 46 (Spring), pp. 43-78.

Seccareccia, M., and M. Lavoie (1989), "Les idées révolutionnaires de Keynes en politique économique et le déclin du capitalisme rentier", Économie appliquée, Vol. 42, no. 1, pp. 47-70.

Wisman, J.D. (1989), "Straightening Out the Backward-Bending Supply Curve of Labour", Review of Political Economy, Vol. 1, no. 1 (March), pp. 94-112.

Wray, L.R. (1997), "Government as Employer of Last Resort: Full Employment without Inflation", Jerome Levy Economics Institute, Working Paper No. 213, (November).

Wray, L.R. (1998a), Understanding Modern Money, The Key to Full Employment and Price Stability, Cheltenham, U.K.: Edward Elgar Publishing Ltd.

Wray, L.R. (1998b), "Zero Unemployment and Stable Prices", Journal of Economic Issues, Vol. 32, no. 2 (June), pp. 539-45.