Snowdon & Vane Modern Macroeconomics

Many economists share Akerlof’s concerns and are critical of models where the labour market is modelled in much the same way as a commodity or financial market. The flexible price–auction model employed by new classical economists does not seem to resemble observed labour market behaviour. There are fundamental differences between labour inputs and other nonhuman inputs into the production process:

1.Workers have preferences and feelings; machines and raw materials do not.

2.Workers need to be motivated; machines do not.

3.The productivity of a machine is reasonably well known before purchase, so that problems of asymmetric information relating to quality are much less significant.

4.Workers can strike and ‘break down’ because of ill health (stress and so on); machines can break down but never strike for higher pay or more holidays.

5.The human capital assets of a firm are more illiquid and risky than its capital assets.

6.Workers normally require training; machines do not.

7.Human capital cannot be separated from its owner; non-human capital can.

8.Workers’ utility functions are interdependent, not independent.

Because of these crucial differences, worker productivity is a discretionary variable; the effort or output of a worker is not given in advance and fixed for the future, irrespective of changes which take place in working conditions (see also Leibenstein, 1979). A machine does not get angry when its price fluctuates, nor does it feel upset if it is switched off. In contrast, workers are not indifferent to their price, nor are they unmoved by becoming unemployed against their will. For these and other reasons, the notion of fairness would seem to be an important factor in determining outcomes in the labour market. As Solow (1990) has argued, ‘The most elementary reason for thinking that the concept of fairness, and beliefs about what is fair and what is not, play an important part in labour market behaviour is that we talk about them all the time.’ The words ‘fair’ and ‘unfair’ have even been used by neoclassical economists at university departmental meetings!

The first formal model to bring in sociological elements as an explanation of efficiency wages was the seminal paper by Akerlof (1982), where issues relating to fairness lie at the centre of the argument. According to Akerlof, the willing cooperation of workers is something that must usually be obtained by the firm because labour contracts are incomplete and teamwork is frequently the norm. The essence of Akerlof’s gift exchange model is neatly summed up

in the phrase ‘A fair day’s work for a fair day’s pay’. Everyday observation suggests that people have an innate psychological need to feel fairly treated, otherwise their morale is adversely affected. In Akerlof’s model, workers’ effort is a positive function of their morale and a major influence on their morale is the remuneration they receive for a given work standard which is regarded as the norm. If a firm pays its workers a wage above the going market rate, workers will respond by raising their group work norms, providing the firm with a gift of higher productivity in exchange for the higher wage.

In subsequent work Akerlof and Yellen (1990) have developed what they call the ‘fair wage–effort hypothesis’, which is derived from equity theory. In the workplace personal contact and potentially conflicting relationships within a team of workers are unavoidable. As a result issues relating to fairness are never far away. Since there is no absolute measure of fairness, people measure their treatment by reference to other individuals within their own group. Fairness is measured by making comparisons with workers similarly situated (inside and outside the firm). Thus an individual worker’s utility function can be summarized as equation (7.13):

The utility of this worker (U) is dependent on the real wage (w) relative to the perceived ‘fair’ wage (ω ), the worker’s effort (e) and the unemployment rate (u). Assuming the worker wishes to maximize this function, the effort expended will depend on the relationship between w and ω for a given level of unemployment. Workers who feel unfairly treated (w < ω ) will adjust their effort accordingly. ‘The ability of workers to exercise control over their effort, and their willingness to do so in response to grievances, underlies the fair wage–effort hypothesis’ (Akerlof and Yellen, 1990, p. 262). Just as firms face a no-shirking constraint in the Shapiro–Stiglitz model, they face a ‘fair wage constraint’ in the fairness version of the efficiency wage model. Since the fair wage exceeds the market-clearing wage, this framework generates an equilibrium with involuntary unemployment.

The essence of this innovative approach to explaining real wage rigidity is that the morale of a firm’s human capital can easily be damaged if workers perceive that they are being unfairly treated. Firms that attach importance to their reputation as an employer and that wish to generate high morale and loyalty from their workforce will tend to pay efficiency wages which are perceived as fair.

It appears that American entrepreneur Henry Ford shared Marshall’s insight that ‘highly paid labour is generally efficient and therefore not dear labour’. In the autumn of 1908, Henry Ford launched the production of the

famous Model T Ford. During the period 1908–14, he pioneered the introduction of mass production techniques that characterized the ‘American System of Manufactures’ (Rosenberg, 1994). The assembly line production methods introduced by Ford required relatively unskilled workers rather than the skilled craftsmen he had previously needed to assemble automobiles one by one. The first moving assembly lines began operation in April 1913 but unfortunately for Ford, the introduction of these mass production techniques drastically changed the working environment and led to a massive and costly increase in absenteeism and the turnover of workers. In 1913 the annual turnover of workers at Ford was 370 per cent and daily absenteeism was 10 per cent. In January 1914 Ford responded to this problem by introducing a payment system of $5 for an eight-hour day for male workers over the age of 22 who had been with the company for at least six months. Previously these same workers had been working a nine-hour day for $2.34. For a given level of worker productivity an increase in the wage paid was certain to increase unit labour costs and, to contemporary observers, Ford’s policy seemed to imply a certain reduction in the firm’s profits. However, the result of Ford’s new wage policy was a dramatic reduction in absenteeism (down 75 per cent), reduced turnover (down 87 per cent), a massive improvement in productivity (30 per cent), a reduction in the price of the Model T Ford, and an increase in profits. It appears that Ford was one of the first entrepreneurs to apply efficiency wage theory. Later, Henry Ford described the decision to pay his workers $5 per day as ‘one of the finest cost cutting moves we ever made’ (see Meyer, 1981; Raff and Summers, 1987). There is no evidence that Ford was experiencing trouble recruiting workers before 1914 or that the new wage policy was introduced to attract more highly skilled workers. The most plausible rationale for the policy is the favourable impact that it was expected to have on workers’ effort, turnover and absenteeism rates, and worker morale. Raff and Summers (1987) conclude that the introduction by Ford of ‘supracompetitive’ wages did yield ‘substantial productivity benefits and profits’ and that this case study ‘strongly supports’ the relevance of several efficiency wage theories.

Insider–outsider models Why don’t unemployed workers offer to work for lower wages than those currently paid to employed workers? If they did so, wages would be bid down and employment would increase. There appears to be an unwritten eleventh commandment: ‘Thou shalt not permit job theft by underbidding and stealing the jobs of thy comrades.’ The insider–outsider theory also attempts to explain why wage rigidity persists in the face of involuntary unemployment (see Ball, 1990, and Sanfey, 1995 for reviews).

The insider–outsider approach to real wage rigidity was developed during the 1980s in a series of contributions by Lindbeck and Snower (1985, 1986,

1988a, 1988b). In this model the insiders are the incumbent employees and the outsiders are the unemployed workers. Whereas in efficiency wage models it is firms that decide to pay a wage higher than the market-clearing wage, in the insider–outsider approach the focus shifts to the power of the insiders who at least partially determine wage and employment decisions. No direct effects of wages on productivity are assumed.

Where does the insider power come from? According to Lindbeck and Snower, insider power arises as a result of turnover costs (Vetter and Andersen, 1994). These include hiring and firing costs such as those associated with costs of searching the labour market, advertising and screening, negotiating conditions of employment, mandatory severance pay and litigation costs. Other important costs are production-related and arise from the need to train new employees. In addition to these well-known turnover costs, Lindbeck and Snower (1988a) also emphasize a more novel form of cost – the insider’s ability and incentive to cooperate with or harass new workers coming from the ranks of the outsiders. If insiders feel that their position is threatened by outsiders, they can refuse to cooperate with and train new workers, as well as make life at work thoroughly unpleasant. By raising the disutility of work, this causes the outsiders’ reservation wage to rise, making it less attractive for the firm to employ them. To the extent that cooperation and harassment activities lie within the control of workers, they can have a significant influence on turnover costs by their own behaviour.

Because firms with high rates of turnover offer both a lack of job security and few opportunities for advancement, workers have little or no incentive to build reputations with their employers. Low motivation damages productivity and this represents yet another cost of high labour turnover.

Because it is costly to exchange a firm’s current employees for unemployed outsiders, the insiders have leverage which they can use to extract a share of the economic rent generated by turnover costs (the firm has an incentive to pay something to avoid costly turnover). Lindbeck and Snower assume that workers have sufficient bargaining power to extract some of this rent during wage negotiations. Although unions are not necessary for insider power, they enhance it with their ability to threaten strikes and work-to-rule forms of non-cooperation (For a discussion of union bargaining models and unemployment, see McDonald and Solow, 1981; Nickell, 1990; Layard et al., 1991.)

Although the insider–outsider theory was originally put forward as an explanation of involuntary unemployment, it also generates some other important predictions (see Lindbeck and Snower, 1988b). First, insider–outsider theory implies that pronounced aggregate shocks which shift the demand for labour may have persistent effects on wages, employment and unemployment. In countries with large labour turnover costs and powerful unions, this

‘effect persistence’ will be significant. Second, in cases where the shocks are mild, firms with high turnover costs have an incentive to hoard labour, and this reduces employment variability. Third, the insider–outsider model can provide a rationale for many features associated with ‘dual labour markets’. Fourth, this model has implications for the composition of unemployment. Lindbeck and Snower (1988b) argue that ‘unemployment rates will be comparatively high for people with comparatively little stability in their work records’. This offers an explanation for the relatively high unemployment rates which are frequently typical among the young, the female population and various minority groups.

While the insider–outsider theory and efficiency wage theories provide different explanations of involuntary unemployment, they are not incompatible but complementary models, since the amount of involuntary unemployment ‘may depend on what firms are willing to give and what workers are able to get’ (Lindbeck and Snower, 1985).

7.8New Keynesian Business Cycle Theory

New Keynesian economists accept that the source of shocks which generate aggregate disturbances can arise from the supply side or the demand side. However, new Keynesians argue that there are frictions and imperfections within the economy which will amplify these shocks so that large fluctuations in real output and employment result. The important issue for new Keynesians is not so much the source of the shocks but how the economy responds to them.

Within new Keynesian economics there have been two strands of research relating to the issue of aggregate fluctuations. The predominant approach has emphasized the importance of nominal rigidities. The second approach follows Keynes (1936) and Tobin (1975), and explores the potentially destabilizing impact of wage and price flexibility. We will examine each in turn. Consider Figure 7.8. In panel (a) we illustrate the impact of a decline in the money supply which shifts aggregate demand from AD0 to AD1. If a combination of menu costs and real rigidities makes the price level rigid at P0, the decline in aggregate demand will move the economy from point E0 to point E1 in panel (a). The decline in output reduces the effective demand for labour. In panel (c) the effective labour demand curve (DLe) shows how much labour is necessary to produce different levels of output. As the diagram shows, L1 amount of labour is required to produce Y1 amount of output. With prices and the real wage fixed at P0 and w0, respectively, firms move off the notional demand curve for labour, DL, operating instead along their effective labour demand curve indicated by NKL1 in panel (d). At the rigid real wage of w0, firms would like to hire L0 workers,

but they have no market for the extra output which would be produced by hiring the extra workers. The aggregate demand shock has produced an increase in involuntary unemployment of L0 – L1. The new Keynesian shortrun aggregate supply curve SRAS (P0) is perfectly elastic at the fixed price level. Eventually downward pressure on prices and wages would move the economy from point E1 to E2 in panel (a), but this process may take an unacceptably long period of time. Therefore new Keynesian economists, like Keynes, advocate measures which will push the aggregate demand curve back towards E0. In the new Keynesian model, monetary shocks clearly have non-neutral effects in the short run, although money remains neutral in the long run, as indicated by the vertical long-run aggregate supply curve (LRAS).

The failure of firms to cut prices even though this would in the end benefit all firms is an example of a ‘coordination failure’. A coordination failure occurs when economic agents reach an outcome that is inferior to all of them because there are no private incentives for agents to jointly choose strategies that would produce a much better (and preferred) result (see Mankiw, 2003). The inability of agents to coordinate their activities successfully in a decentralized system arises because there is no incentive for a single firm to cut price and increase production, given the assumed inaction of other agents. Because the optimal strategy of one firm depends on the strategies adopted by other firms, a strategic complementary is present, since all firms would gain if prices were reduced and output increased (Alvi, 1993). To many Keynesian economists the fundamental causes of macroeconomic instability relate to problems associated with coordination failure (see Ball and Romer, 1991; Leijonhufvud, 1992).

The second brand of new Keynesian business cycle theorizing suggests that wage and price rigidities are not the main problem. Even if wages and prices were fully flexible, output and employment would still be very unstable. Indeed, price rigidities may well reduce the magnitude of aggregate fluctuations, a point made by Keynes in Chapter 19 of the General Theory, but often neglected (see Chapter 2 above, and General Theory, p. 269). A reconsideration of this issue followed Tobin’s (1975) paper (see Sheffrin, 1989, for a discussion of this debate). Tobin himself remains highly critical of new Keynesian theorists who continue to stress the importance of nominal rigidities (Tobin, 1993), and Greenwald and Stiglitz have been influential in developing new Keynesian models of the business cycle which do not rely on nominal price and wage inertia, although real rigidities play an important role.

In the Greenwald and Stiglitz model (1993a, 1993b) firms are assumed to be risk-averse. Financial market imperfections generated by asymmetric information constrain many firms from access to equity finance. Equity-rationed

firms can only partially diversify out of the risks they face. Their resultant dependence on debt rather than new equity issues makes firms more vulnerable to bankruptcy, especially during a recession when the demand curve facing most firms shifts to the left. Faced with such a situation, a risk-averse equity-constrained firm prefers to reduce its output because the uncertainties associated with price flexibility are much greater than those from quantity adjustment. As Stiglitz (1999b) argues, ‘the problem of price-wage setting should be approached within a standard dynamic portfolio model, one that takes into account the risks associated with each decision, the nonreversibilities, as well as the adjustment costs associated with both prices and quantities’.

Greenwald and Stiglitz argue that, as a firm produces more, the probability of bankruptcy increases, and since bankruptcy imposes costs these will be taken into account in firms’ production decisions. The marginal bankruptcy cost measures the expected extra costs which result from bankruptcy. During a recession the marginal bankruptcy risk increases and risk-averse firms react to this by reducing the amount of output they are prepared to produce at each price (given wages). Any change in a firm’s net worth position or in their perception of the risk they face will have a negative impact on their willingness to produce and shifts the resultant risk-based aggregate supply curve to the left. As a result, demand-induced recessions are likely to induce leftward shifts of the aggregate supply curve. Such a combination of events could leave the price level unchanged, even though in this model there are no frictions preventing adjustment. Indeed, price flexibility, by creating more uncertainty, would in all likelihood make the situation worse. In the Greenwald–Stiglitz model aggregate supply and aggregate demand are interdependent and ‘the dichotomy between “demand” and “supply” side shocks may be, at best, misleading’ (Greenwald and Stiglitz, 1993b, p. 103; Stiglitz, 1999b).

In Figure 7.9 we illustrate the impact of an aggregate demand shock which induces the aggregate supply curve to shift to the left. The price level remains at P0, even though output falls from Y0 to Y1. A shift of the aggregate supply curve to the left as the result of an increase in perceived risk will also shift the demand curve of labour to the left. If real wages are influenced by efficiency wage considerations, involuntary unemployment increases without any significant change in the real wage.

In addition to the above influences, new Keynesians have also examined the consequences of credit market imperfections which lead risk-averse lenders to respond to recessions by shifting their portfolio towards safer activities. This behaviour can magnify an economic shock by raising the real costs of intermediation. The resulting credit squeeze can convert a recession into a depression as many equity-constrained borrowers find credit expensive or difficult to

Figure 7.9 The risk-based aggregate supply curve

obtain, and bankruptcy results. Because high interest rates can increase the probability of default, risk-averse financial institutions frequently resort to credit rationing. Whereas the traditional approach to analysing the monetary transmission mechanism focuses on the interest rate and exchange rate channels, the new paradigm emphasizes the various factors that influence the ability of financial institutions to evaluate the ‘creditworthiness’ of potential borrowers in a world of imperfect information. Indeed, in the new paradigm, banks are viewed as risk-averse firms that are constantly engaged in a process of screening and monitoring customers. In a well-known paper, Bernanke (1983) argues that the severity of the Great Depression was in large part due to the breakdown of the economy’s credit facilities, rather than a decline in the money supply (see Jaffe and Stiglitz, 1990, and Bernanke and Gertler, 1995, for surveys of the literature on credit rationing; see also Stiglitz and Greenwald, 2003, who champion what they call ‘the new paradigm’ in monetary economics).

Some new Keynesians have also incorporated the impact of technology shocks into their models. For example, Ireland (2004) explores the link between the ‘current generation of new Keynesian models and the previous generation of real business cycle models’. To identify what is driving aggregate instability, Ireland’s model combines technology shocks with shocks to household preferences, firm’s desired mark-ups, and the central bank’s monetary policy rule. Ireland finds that monetary shocks are a major source of

real GDP instability, particularly before 1980. Technology shocks play only a ‘modest role’, accounting for less than half of the observed instability of output in the post-1980 period.

7.9Hysteresis and the NAIRU

Since the early 1970s the natural rate of unemployment ‘seems to have taken a wild ride’ in OECD countries. For OECD countries in general, unemployment in the 1980s and 90s was higher than during the ‘Golden Age’ of the 1950–73 period. The steadily rising unemployment rates appear to have their origins in the two OPEC oil price shocks in 1973 and 1979 respectively (Phelps and Zoega, 1998) and in the case of the European OECD countries, unemployment that averaged 1.7 per cent in the early 1960s rose to 11 per cent by the mid-1990s. This high average also hides the large dispersion of unemployment rates across the European countries (Blanchard and Wolfers, 2000). Gordon’s (1997, 1998) estimates show that the US natural rate of unemployment has also varied during this same period although the long-run unemployment repercussions of the 1980s recessions appear to have been much more persistent in Europe than in the USA. Figure 7.10 shows the standardized unemployment rates for the USA and OECD Europe for the period 1972–98. While unemployment in OECD Europe was less than US unemployment until the early 1980s, since then European unemployment has remained stubbornly high while it has fallen in the USA.

While the problem of inflation was a major policy concern during the 1970s and early 1980s, by the mid-1980s economists were once again turning their attention to the problem of unemployment, in particular the rise in the estimated NAIRU (see Bean et al., 1986; Fitoussi and Phelps, 1988; Summers, 1990; Layard et al., 1991, 1994; Bean, 1994; Cross, 1995; Nickell, 1997, 1998; Siebert, 1997; Katz and Krueger, 1999; Blanchard and Wolfers, 2000; Fitoussi et al., 2000; Hall, 2003).

While estimates of the NAIRU are obviously subject to uncertainty given the broad range of determinants, recent OECD estimates shown in Table 7.1 indicate the much superior performance of the US economy compared to the euro area and G7 countries.

During the late 1960s, Friedman and Phelps independently put forward expectations-augmented models of the Phillips curve. In Friedman’s model the market-clearing rate of unemployment is called the natural rate of unemployment and is associated with a stable rate of inflation. As we noted in Chapter 4, many economists (especially those sympathetic to Keynesianism) prefer to use the ‘NAIRU’ concept (non-accelerating inflation rate of unemployment), rather than ‘natural rate’ when discussing long-run unemployment. The NAIRU terminology was first introduced by Modigliani and Papademos