What causes the real wage rate to increase?

7 Economic Growth in Countries with Net Emigration

The large real wage differentials that can still be observed between developed and developing countries are indicative of the economic gains that can be achieved in terms of global welfare when workers would be allowed to freely cross borders to where their human capital has its greatest return. These welfare gains and the impacts for the distribution of income across owners of capital, workers in migrant-sending countries, the native-born in host countries, and the migrants were already discussed earlier in this chapter by means of Figure 19.2 and Table 19.3. The overall gains are much larger than those that would result from a further reduction of trade barriers between countries. Clemens (2011) reviewed the available evidence and concluded that a removal of barriers to international movement could conservatively add 20–60% to global GDP. Similarly, Kennan (2013) estimated that the net gains from open borders, taking migration costs into account, could be more than double the income level in less-developed countries. Of course, the societal impacts of free global labor mobility in terms of, for example, population size and distribution, social development, cultural identities, and national sovereignty are potentially huge as well. Walmsley and Winters (2005) calculated that a more realistic additional emigration rate of 1.6% from developing countries might add another 1.2% to world GDP.

Such gains in global welfare are the result of a one-off improvement in spatial distribution, and therefore the allocative efficiency, of the world's labor force. However, an important question from a long-run perspective is whether such short-run gains also lead to higher “dynamic efficiency”—that is, higher rates of global economic growth. Growth models such as formulated by Klein and Ventura (2009) suggest that the dynamic gains can also be very large. The neoclassical and endogenous growth theories discussed in this chapter make it clear that long-run growth effects would depend on the strength of a range of potentially favorable flow-on effects that would be triggered by greater integration of the world's labor markets. Such flow-on effects would include greater agglomeration, FDI, trade, and innovation. Since the empirical evidence reviewed in this chapter suggests that such flow-on effects are indeed plausible, it can be expected that a more integrated global labor market may achieve an endogenously determined higher rate of long-run growth.

However, the extent to which such additional growth leads to convergence or divergence between countries and regions is not a priori clear. Several theoretical mechanisms were discussed earlier in the chapter that could lead to either outcome. In the remainder of this section we consider briefly the implications of greater cross-border mobility for migrant-sending countries that are likely to lose a significant proportion of their labor force, skilled or unskilled.

Columns 4 and 5 were suggestive of population redistribution from migrant-sending countries to migrant-receiving countries leaving the former countries worse off in the short to medium term (a statistically significant positive coefficient of the net migration rate in column 5), while leaving growth in income per capita in the latter largely unaffected (an insignificant coefficient in column 4). In the longer run, migrant-sending countries in the developing world could expect higher growth in income per capita (a negative coefficient in column 8 at a lag of three decades), while there are long-run growth benefits for the high-income countries as well (a positive coefficient in column 10 at a lag of two decades). Using a multi-sectoral model calibrated with data from 60 developed and developing countries, di Giovanni et al. (2012) also detected an asymmetry between the long run and short run. With their model, they found that migrant-receiving countries benefit from larger scale and variety (as in the models we reviewed in Section 4), but only in the longer run. The potentially negative impact of a smaller scale production with less variety in countries that send migrants is overcompensated by remittances, which raise the incomes of the population left behind both in the short run and in the long run.

As Figure 19.2 suggested, emigration is expected to lead initially to a higher wage in the migrant-sending country. Studies of, for example, migration from Mexico to the US (Mishra, 2007) or from Lithuania to the European Union (Elsner, 2013) and emigration from Moldova (Bouton et al., 2011) suggested that this is indeed the case. Any short-run decline in growth in income per capita in migrant-sending countries would then be due to lower aggregate demand or due to positive self-selection of emigrants in terms of skills and unmeasured ability—the so-called brain drain (e.g., Bhagwati, 1976)—lowering aggregate productivity. Clearly, endogenous models of growth would suggest that selective emigration of the high skilled would also lower long-run growth (see also Chen, 2006).

However, in recent years the literature is increasingly seeing net emigration of skilled people from developing countries in a more positive light (e.g., Duncan, 2008; Gibson and McKenzie, 2011). First of all, higher returns obtainable abroad to investments in education and training may encourage a greater proportion of the workforce to invest in human capital than otherwise. Not all of these higher educated workers will actually emigrate. The potential opportunities abroad therefore generate a positive spillover from human capital accumulation in the source country.

Another major benefit for the home country is the receipt of remittances, particularly when these trigger domestic investment rather than consumption of imported commodities. Remittances can also have positive effects on income distribution. Adams and Page (2005) found that international migration and remittances lead to a sharp reduction in poverty in the developing world. Although some researchers argue that highly skilled migrants are often less committed to their home country, Bollard et al. (2011) found that the more educated migrants in fact remit more than average. The impacts of remittances are discussed in detail in Chapter 20 of this Handbook.

Another benefit from emigration is that it appears to trigger foreign direct investment (FDI) in the home country by firms from the migrants’ host country (Foley and Kerr, 2011). Additionally, Nijkamp et al. (2011) found that immigration has a positive impact on FDI investment in both directions (inward and outward). They also concluded that these impacts are greater when migrants are relatively highly educated. Similarly, the ties between home and host countries forged by migration also increase bilateral trade, as is demonstrated by the meta-analysis of Genc et al. (2012). However, the meta-analysis showed a slight trade balance benefit in favor of a migrant host country. This would imply trade balance deterioration in the sending country, but such a negative impact might be offset by the growth in remittances.

An important aspect of emigration for economic growth in the home country is the impact of the resulting networks and ties between migrants and businesses in the home country. The impacts from diaspora for innovation in the home country have been recently highlighted by various case studies, particularly with respect to diaspora from China and India. Agrawal et al. (2011) argued that the emigration of highly skilled individuals has a detrimental effect on innovation and local knowledge networks in the home country but, once such emigrants contribute to innovation activity in the host country, personal networks with innovators back home can contribute to dissemination of new knowledge and practices.

Besides benefitting from network ties with high-skilled diaspora, it should be noted that developing countries can also benefit from increasing temporary migration, return migration and circulation, encouraged by lower real costs of air transportation and by cheaper information exchange through new information and communication technologies. It is estimated that up to about one-third of migrants may return to the home country in the long run, where the return of diaspora may raise human capital levels and entrepreneurship. This is another channel through which emigration can be a source of growth for the home country (Dos Santos and Postel-Vinay, 2003).

Fertility.

Over the period from 1830 to 1990 real wages increased by a factor of 9 – see Figure 1.1 This rise was propelled by a near 7-fold increase in market-sector total factor productivity (TFP) between 1800 and 1990. Such tremendous technological advance had a dramatic impact on everyday life. As an example, consider the effect that economic progress could have had on fertility. Raising children takes time. A secular increase in real wages implies that the opportunity cost of having a child, when measured in terms of market goods, will rise. The utility value of an extra unit of market consumption relative to an extra child should fall, however, as market goods become more abundant with economic development. So long as the marginal utility of market goods falls by less than the increase in real wages fertility should decline. And so fertility did decline, from 7 kids per woman in 1800 to 2 today.

In the short run, the ETS reduces employment relative to its base-case level; over time, the employment deviation remains fairly constant as the national real wage rate adjusts downwards

The explanation of macro effects begins with the impacts on the national labor market. Figure 9.13 shows percentage deviations in national employment, the national real wage rate and the national real cost of labor. The real wage is defined as the ratio of the nominal wage rate to the price of consumption. The real cost of labor is defined as the ratio of the nominal wage rate to the national price of output (measured by the factor-cost GDP deflator). Assuming competitive markets, the equilibrium nominal wage will be equal to the value of the marginal product of labor.

According to the labor market specification in MMRF (Section 9.5.3.1), the real wage rate is sticky in the short run (i.e. the nominal wage moves with the price of consumption), but adjusts with a lag downwards (upwards) in response to a fall (rise) in employment. When the ETS starts up, the emissions price increases the price of spending (e.g. household consumption) relative to the price of output and hence moves the nominal wage above the value of the marginal product of labor in the short run. In Figure 9.13 this shows as an increase in the real cost of labor relative to its base-case value and a fall in employment relative to base case.

If there were no further shocks, over time the real wage rate would progressively fall relative to base-case levels, reducing the real cost of labor and forcing employment back to its base-case level. In the ETS simulations, however, shocks continue with the permit price increasing under a progressively tighter regime of tradable permits. Hence, as shown in Figure 9.13, the employment deviation is never fully eliminated and the real wage rate declines steadily relative to its base-case value. In 2030, the employment deviation is –0.2%, while the real wage rate is down 2.6%.

Note that the deviations in employment and the real wage rate are not smooth, especially in the early years, despite the smoothness of the permit-price trajectory (Figure 9.7). This reflects a number of factors:

the changing coverage of the ETS scheme, with transport industries entering in 2012 and agricultural industries entering in 2015 (Table 9.8);

large changes in electricity generation and capacity by technology type projected by the detailed electricity modeling (Figure 9.9a and b);

swings in the national terms of trade projected by GTEM (Figure 9.11).

The swings in the terms of trade have a significant impact on the labor market in the short run. An increase in the terms of trade causes the price of final domestic demand (which includes import prices but excludes export prices) to fall relative to the price of GDP (which excludes import prices but includes export prices), leading to downward pressure on the real cost of labor. Hence, relative to base, changes in the terms of trade contribute positively to employment in the first few years of the projection when the terms of trade rise.

A final point to note is that even though the fall in national employment is fairly small, this does not mean that employment at the individual industry or regional level remains close to base-case values. In most industries and regions, there are significant permanent employment responses to the ETS, compounding or defusing existing (base-case) pressures for structural change.

5.3 The Cyclicality of Real Labor Compensation

We are now in a position to examine the cyclical behavior of real wages. Tables 2 and 3 report cyclicality estimates for six different measures of log real wages. For each measure of real wages, we regress the calculated wage series on an indicator of the business cycle (and a time trend and a constant). Table 2 examines the cyclicality of real wages with respect to the HP filtered unemployment rate. We use HP filtered unemployment rather than the unemployment rate in levels because the average unemployment rate changes substantially over the time period for the NLSY.ac Thus, the coefficients reported are semielasticities: the percent change in a real wage measure in response to a one percentage-point deviation of unemployment relative to its trend. The sample for columns 1–5 consists of 25 data points from 1979 to 2012, dropping the odd years between 1994 and 2012 (see Section 3.2). To construct the UCL, we need to impute values of wages for the odd years between 1994 and 2012. The final user cost series itself ends in 2007 because we require seven subsequent wage observations to calculate the value of the UCL in year t (again see Section 3.2 for details).

Table 2. Real wage cyclicality: Unemployment rate

Notes: OLS standard errors are in parentheses. Coefficients are multiplied by 100.

Table 3. Real wage cyclicality: GDP

Notes: OLS standard errors are in parentheses.

Columns 1–4 report results for average hourly earnings. For the BLS wage series, AHE-BLS, the coefficient on the unemployment rate is −0.507: real average hourly earnings fall by roughly 0.5% for each percentage point increase in the cyclical component of the unemployment rate. Columns 2–4 report results for our constructed measure of average hourly earnings from the NLSY data, AHE-NLSY. As noted in our discussion earlier, the dependent variables in Columns 2–4 are estimated time fixed effects from regressions of individual wages on the listed set of controls. The columns differ according to the number of controls included in the regression. Column 2 includes only experience and experience squared; column 3 adds industry fixed effects, job tenure and schooling; column 4 includes all of the aforementioned controls and adds individual fixed effects. The NLSY sample exhibits greater cyclicality for all of the measures of average hourly earnings, and the cyclicality rises with the number of controls for worker characteristics. We interpret this finding as being supportive of the basic composition-bias effect emphasized by Bils (1985) and Solon et al. (1994). Typically, as we add more controls for worker heterogeneity, the point estimate of the cyclicality rises (though note, the standard errors are high enough that we cannot say with any certainty that any one of these measures is clearly more or less cyclical than any other).

Column 5 reports results for the new-hire wage. The point estimate for the cyclicality coefficient is − 0.698, so a one percentage point increase in the cyclical component of unemployment corresponds to a 0.7% reduction in the real new-hire wage. By itself, the point estimate seems to be at odds with the findings in Haefke et al. (2013) who reported that in CPS data, the wages of newly hired workers appeared substantially more cyclical than average hourly earnings. We should note that while our point estimates do not indicate greater cyclicality of the new-hire wage, the estimates are quite noisy and admit a range of interpretations.

Column 6 reports results for the user cost of labor (UCL). Our measure of the UCL exhibits much greater cyclicality than either the composition-adjusted wage or the new-hire wage series. In Table 2, the cyclicality estimate is − 5.818 indicating that for every one percentage-point increase in the cyclical component of unemployment, the real user cost of labor falls by almost 6% (!).

The estimates in Table 2 are robust to alternate measures of the business cycle. Table 3 reports estimates for the same dependent variables as those in Table 2, but uses HP filtered GDP as the indicator of the business cycle instead of the unemployment rate. Again, average hourly earnings seem to be only moderately cyclical. When HP filtered GDP is above trend by 1%, AHE-BLS is above trend by only 0.311%. By contrast, holding the set of workers fixed in the NLSY and controlling for observed and unobserved heterogeneity increases this estimate to 1.165%. The point estimate of the cyclicality of the new-hire wage is more cyclical. The point estimate is a rise of roughly 1.3% for every 1% change in the cyclical component of GDP. Finally, as before, the UCL is the most cyclical wage measure. For each percent increase in GDP above trend, the UCL rises by approximately 3.1%.

What these results seem to suggest is that both composition bias and implicit contracting play important roles in shaping the wage payments made to workers over the business cycle. Quantitatively, controlling for composition (by including individual fixed effects and controls for observed worker differences in the wage regressions) increases wage cyclicality by perhaps as much as a factor of two relative to a group of workers without such controls. The effects of implicit contracting and wage-smoothing seem to be even greater than the effects of composition bias. According to our calculations, the user cost of labor has a cyclicality that is, in some cases, about six times greater than the log real wages of the base group. Since average payments are less cyclical than the user cost, workers hired in bad times are paid a wage greater than their user cost. In return, the workers expect to receive fewer and smaller wage increases over their employment spell.

Our findings (which are consistent with the results in Kudlyak, 2014) seem to corroborate the results in Beaudry and DiNardo (1991, 1995), who argued that current wage payments seem to be tied to past labor market conditions. In that paper, the authors showed that the maximum unemployment rate during a job spell and the unemployment rate that prevailed when the worker was hired both have a significant influence on current wage payments. The specification above, which we have adapted from Kudlyak's work, is a more general econometric specification than the one in Beaudry and DiNardo, but implicit contracts still appear to play an important role in shaping wage payments.

3.2.1 Example: Real Wage Estimation

The decision to work is usually assumed to be a function of the real wage (i.e., the purchasing power of the nominal wage). Suppose that a worker is interested in knowing his or her real wage. Although the worker knows his or her nominal wage, the price level is not evident at the time the worker makes the decision to work. Denote the logarithm of the nominal wage by W and the logarithm of the price level by p. Suppose that wage and prices are governed by the following equations:

W=z+uandp=z+v,

where z represents the movements in the price level that leave the real wage unchanged so that disturbance terms u and v determine the variations in real wage. Assume that the disturbances are serially uncorrelated processes with u∼N(0, σu2) and v∼N(0,σv2). Furthermore suppose that the disturbances are not correlated with each other or with z. The worker's best estimate of the logarithm of the real wage by observing the nominal wage and knowing the first and the second moments of all the variables is given by

·(W−p∩ )=w1W.

According to the Wiener-Hopf equation (Equations 3.11, 3.12), we have to find the following:

γyy =E(WW)=E(z+u)(z+u)=E(u2)+E(v2)γxy=E[(W(W−p)]=E(z+u)(u−v) =E(u2).

Substituting these expressions into the Wiener-Hopf equation (Equation 3.11), the optimum weight w1 is found to be

w1=E(u2)E(u2)+E(v2).

The weight of the nominal wage in making an estimate about the variation in the real wage lies between zero and one. The weight of nominal wage in real-wage estimation becomes zero if all of the variation in the real wage comes from the price level, E(u2) = 0. On the contrary, if E(u2)/E(v2) is large (i.e., change in real wage is caused mainly by the nominal wage movements), then the weight of the nominal wage approaches one.11

Imported primary inputs: partial equilibrium substitute but general equilibrium complement for skilled labor

One of the most interesting results is displayed in row 2. The real wage of skilled labor rises monotonically across the row. As barriers to foreign service providers fall, the X sector substitutes foreign services for domestic services and there is a substitution effect away from domestic skilled labor because foreign service providers use skilled labor less intensively than domestic service providers (V economizes on domestic skilled labor in producing ZM). However, the reduction in the quality adjusted cost of services lowers the cost of final output in the X sector and induces an output expansion there. In the simulation, the expansion of output in the X sector increases the X-sector’s direct demand for skilled labor. The output effect dominates the substitution effect resulting in an increase in the demand for skilled labor on balance. Thus, V and skilled labor are partial equilibrium substitutes but general equilibrium complements. These results are particularly dramatic if we want to think of V as largely consisting of imported skilled workers: they are clearly a general equilibrium complement to domestic skilled labor.

2.3. Real wage levels by skill group

A limitation of the college/high school wage premium as a measure of the market value of skill is that it necessarily omits information on real wage levels. Stated differently, a rising college wage premium is consistent with a rising real college wage, a falling real high school wage, or both. Movements in real as well as relative wages will prove crucial to our interpretation of the data. As shown formally in Section 3, canonical models used to analyze the college premium robustly predict that demand shifts favoring skilled workers will both raise the skill premium and boost the real earnings of all skill groups (e.g., college and high school workers). This prediction appears strikingly at odds with the data, as first reported by Katz and Murphy (1992), and shown in the two panels of Fig. 4. This figure plots the evolution of real log earnings by gender and education level for the same samples of full-time, full-year workers used above. Each series is normalized at zero in the starting year of 1963, with subsequent values corresponding to the log change in earnings for each group relative to its 1963 level. All values are deflated using the Personal Consumption Expenditure Deflator, produced by the US Bureau of Economic Analysis.

In the first decade of the sample period, years 1963 through 1973, real wages rose steeply and relatively uniformly for both genders and all education groups. Log wage growth in this ten year period averaged approximately 20 percent. Following the first oil shock in 1973, wage levels fell sharply initially, and then stagnated for the remainder of the decade. Notably, this stagnation was also relatively uniform among genders and education groups. In 1980, wage stagnation gave way to three decades of rising inequality between education groups, accompanied by low overall rates of earnings growth—particularly among males. Real wages rose for highly educated workers, particularly workers with a post-college education, and fell steeply for less educated workers, particularly less educated males. Tables 1a and 1b provide many additional details on the evolution of real wage levels by sex, education, and experience groups during this period.

Table 1a. Changes in real, composition-adjusted log weekly wages for full-time, full-year workers, 1963–2008: by educational category and sex (100 × change in mean log real weekly wages).

Source: March CPS data for earnings years 1963–2008. See note to Fig. 1.

Table 1b. Changes in real, composition-adjusted log weekly wages for full-time, full-year workers, 1963–2008: by experience, educational category, and sex (100 × change in mean log real weekly wages).

Source: March CPS data for earnings years 1963–2008. See note to Fig. 1.

Alongside these overall trends, Fig. 4 reveals three key facts about the evolution of earnings by education groups that are not evident from the earlier plots of the college/high school wage premium. First, a sizable share of the increase in college relative to non-college wages in 1980 forward is explained by the rising wages of post-college workers, i.e., those with post-baccalaureate degrees. Real earnings for this group increased steeply and nearly continuously from at least the early 1980s to present. By contrast, earnings growth among those with exactly a four-year degree was much more modest. For example, real wages of males with exactly a four-year degree rose 13 log points between 1979 and 2008, substantially less than they rose in only the first decade of the sample.

A second fact highlighted by Fig. 4 is that a major proximate cause of the growing college/high school earnings gap is not steeply rising college wages, but rapidly declining wages for the less educated–especially less educated males. Real earnings of males with less than a four year college degree fell steeply between 1979 and 1992, by 12 log points for high school and some-college males, and by 20 log points for high school dropouts. Low skill male wages modestly rebounded between 1993 and 2003, but never reached their 1980 levels. For females, the picture is qualitatively similar, but the slopes are more favorable. While wages for low skill males were falling in the 1980s, wages for low skill females were largely stagnant; when low skill males wages increased modestly in the 1990s, low skill female wages rose approximately twice as fast.

A potential concern with the interpretation of these results is that the measured real wage declines of less educated workers mask an increase in their total compensation after accounting for the rising value of employer provided non-wage benefits such as healthcare, vacation and sick time. Careful analysis of representative, wage and fringe benefits data by Pierce (2001, forthcoming) casts doubt on this notion, however. Monetizing the value of these benefits does not substantially alter the conclusion that real compensation for low skilled workers fell in the 1980s. Further, Pierce shows that total compensation—that is, the sum of wages and in-kind benefits—for high skilled workers rose by more than their wages, both in absolute terms and relative to compensation for low skilled workers.14 A complementary analysis of the distribution of non-wage benefits—including safe working conditions and daytime versus night and weekend hours—by Hamermesh (1999) also reaches similar conclusions. Hamermesh demonstrates that trends in the inequality of wages understate the growth in full earnings inequality (i.e., absent compensating differentials) and, moreover, that accounting for changes in the distribution of non-wage amenities augments rather than offsets changes in the inequality of wages. It is therefore unlikely that consideration of non-wage benefits changes the conclusion that low skill workers experienced significant declines in their real earnings levels during the 1980s and early 1990s.15

The third key fact evident from Fig. 4 is that while the earnings gaps between some-college, high school graduate, and high school dropout workers expanded sharply in the 1980s, these gaps stabilized thereafter. In particular, the wages of high school dropouts, high school graduates, and those with some college moved largely in parallel from the early 1990s forward.

The net effect of these three trends—rising college and post-college wages, stagnant and falling real wages for those without a four-year college degree, and the stabilization of the wage gaps among some-college, high school graduates, and high school dropout workers—is that the wage returns to schooling have become increasingly convex in years of education, particularly for males, as emphasized by Lemieux (2006b). Figure 5 shows this “convexification” by plotting the estimated gradient relating years of educational attainment to log hourly wages in three representative years of our sample: 1973, 1989, and 2009. To construct this figure, we regress log hourly earnings in each year on a quadratic in years of completed schooling and a quartic in potential experience. Models that pool males and females also include a female main effect and an interaction between the female dummy and a quartic in (potential) experience.16 In each figure, the predicted log earnings of a worker with seven years of completed schooling and 25 years of potential experience in 1973 is normalized to zero. The slope of the 1973 locus then traces out the implied log earnings gain for each additional year of schooling in 1973, up to 18 years. The loci for 1989 and 2009 are constructed similarly, and they are also normalized relative to the intercept in 1973. This implies that upward or downward shifts in the intercepts of these loci correspond to real changes in log hourly earnings, whereas rotations of the loci indicate changes in the education-wage gradient.17

The first panel of Fig. 5 shows that the education-wage gradient for males was roughly log linear in years of schooling in 1973, with a slope approximately equal to 0.07 (that is, 7 log points of hourly earnings per year of schooling). Between 1973 and 1989, the slope steepened while the intercept fell by a sizable 10 log points. The crossing point of the two series at 16 years of schooling implies that earnings for workers with less than a four-year college degree fell between 1973 and 1989, consistent with the real wage plots in Fig. 4. The third locus, corresponding to 2009, suggests two further changes in wage structure in the intervening two decades: earnings rose modestly for low education workers, seen in the higher 2009 intercept (though still below the 1973 level); and the locus relating education to earnings became strikingly convex. Whereas the 1989 and 2009 loci are roughly parallel for educational levels below 12, the 2009 locus is substantially steeper above this level. Indeed at 18 years of schooling, it lies 16 log points above the 1989 locus. Thus, the return to schooling first steepened and then “convexified” between 1973 and 2009.

Panel B of Fig. 5 repeats this estimation for females. The convexification of the return to education is equally apparent for females, but the downward shift in the intercept is minimal. These differences by gender are, of course, consistent with the differential evolution of wages by education group and gender shown in Fig. 4.

As a check to ensure that these patterns are not driven by the choice of functional form, Fig. 6 repeats the estimation, in this case replacing the education quartic with a full set of education dummies. While the fitted values from this model are naturally less smooth than in the quadratic specification, the qualitative story is quite similar: between 1973 and 1989, the education-wage locus intercept falls while the slope steepens. The 1989 curve crosses the 1973 curve at 18 years of schooling. Two decades later, the education-wage curve lies atop the 1989 curve at low years of schooling, while it is both steeper and more convex for completed schooling beyond the 12th year.

6.1 Real wage rigidities and wage indexation

As emphasized by Blanchard and Galí (2007, 2010) the presence of real wage rigidities may have implications for the optimal design of monetary policy that are likely to differ from the ones generated by a model with nominal wage rigidities only (like the one emphasized here). Among other things, in the presence of real wage rigidities, the policymaker cannot use price inflation to facilitate the adjustment of real wages. A simple way to introduce real wage rigidities would be to allow for (possibly partial) wage indexation to contemporaneous wage inflation between wage renegotiations. Formally, one can assume:

Wt+k||t=Wt+k-1|t(Pt+k/Pt+k-1)ς

for k = 1, 2, 3, … and Wt|t=Wt* , and where Wt+k||t is the nominal wage in period t + k for an employment relationship whose wage was last renegotiated in period t. Note that parameter ζ ∈ [0,1] measures the degree of indexation. An alternative specification, often used in the New Keynesian literature (e.g., Smets & Wouters, 2007) and adopted by Gertler et al. (2008), assumes instead indexation to past inflation. Formally,

Wt+k||t=Wt+k-1|t(Pt+k-1 /Pt+k-2)ς

for k = 1, 2, 3, … In the latter case, even with full indexation, price inflation can still be used to speed up the adjustment of real wage to shocks that warrant such an adjustment, due to the lags in indexation.

26.9 Conclusion

The last four centuries have seen alterations in the relationship between growth in output per worker and the real wage. In the run up to the Industrial Revolution (1620-1770), both increased, and wages tended to converge upwards. High wages relative to capital prices precipitated the invention of capital-intensive factory production. Competition between the new factories and the remaining handicraft producers during the Industrial Revolution (1770-1867) led to rising inequality in wages, flat average real wages overall, and rising inequality as profits per worker rose. Once the handicraft sector was competed out of existence by the factories, a long period of fairly steady growth in output per worker ensued from 1867 to 1973. The endless invention of new manufactured goods was essential in maintaining buoyant demand for so long. Real wages rose in pace with overall labor productivity. By the end of the nineteenth century, the real wages of skilled workers in the USA, in particular, were leaping ahead, and the chance to eliminate their expensive jobs led to the invention of mass production and the routinization of work–a process that continued through the 1960s. By the 1930s, the spread of collective bargaining throughout manufacturing created an industrial relations system through which the manufacturing worker force could secure increases that matched productivity growth. This favorable situation lasted until about 1973, when the growth in manufacturing output slacked markedly as consumers shifted spending towards services. Slowly growing manufacturing output in conjunction with the continual substitution of capital for labor via new production technology meant that the number of manufacturing workers went into long term contraction. More and more of the new work force was absorbed in the new ascendant sector–services. While it provided many high wage technical and professional jobs, it also recruited large numbers of workers (who would have become factory workers in previous generations) into low paid jobs, many of which were part time. Weakness in the labor market contributed to a rise in profits, which now originate mainly in services. Responding to this new economy is the challenge of the future.

Macroeconomic choices and rational behaviour

As with microeconomic behaviour, in the macroeconomic domain, conventional economics makes the case that decision-makers must be neoclassically rational. The evidence suggests this is not how individuals behave and this has had some major repercussions in the construction of macroeconomic theory and the same for finance theory (Akerlof, 2002; Akerlof & Shiller, 2009). But these reconstructions are rejected outright by those who remain strict adherence to the conventional assumptions of rationality combined with assumptions related to flexible factor prices and the capacity of micro decisions having direct and immediate impact in the macroeconomic domain. The latter ‘school of thought’, in their pre-Keynesian incarnation has been dubbed the classical school whereas their modern equivalents have been referred to as the new classical school of macroeconomics.

Many of the underlying revisions to macroeconomic theory were made decades ago by Keynes, in his articulation of business cycle theory, more specifically his theoretical narrative on the making of deep recessions and the mechanism involved in the economy transitioning from a deep recession or depression to recovery. Keynes’ narrative is largely based on the assumption of smart agents making decisions in a world of complex and asymmetric information with asymmetric power relationships across decision-makers. Keynes introduces the notion of ‘animal spirits’ as an important to the determination of the timing and depth of recessions and upturns. His narrative suggests that animal spirits as a determinant of decision-making are rational in the sense that decision makers are making due, doing their best, satisficing, given their decision-making environment. Hence, Keynes recognizes the importance that non-economic variables and heuristics can play in economic (macroeconomic) outcomes (Keynes, 1936).

In the conventional wisdom non-economic variables are assumed away. Keynes also recognizes the importance of sticky prices as being a possible and possibly important determinant of recession/depression, given negative demand shocks. But the relative importance of sticky prices in determining economic downturns, especially severe ones, is subject to heated debate amongst those writing in the Keynesian tradition. But there is no denying the empirical significance of sticky prices.

Most recently Akerlof (2002), in theory, and Bewley (1999), empirically, have made the case that sticky prices in the face of a negative demand shock are rationally determined. This is based on what is referred to efficiency wage theory, first modelled by Leibenstein (1957). Smart agents make local (within the firm) utility and profit maximizing decisions that have negative macroeconomic consequences, such as persistent unemployment. Firms don’t cut real wages for fear that workers will retaliate by cutting effort inputs thereby reducing productivity. Here effort is a variable in the production function. Employers are also concerned that their best workers will quit, given the opportunity, for what are perceived to be fairer firms, also damaging firm productivity. But workers maximize their utility by taking such action, which is common knowledge to employers. Akerlof considers sticky-price related unemployment to be involuntary. The employed don’t want to lose their jobs or keep others unemployed even though their locally rational decisions have this effect.

It is important to note that classical economists, old and new, pay no attention to this efficiency wage modelling of unemployment. But they interpret such behaviour as a reflection of labour’s preference for leisure at least on the margin. And, there would be the assumption that in the face of negative aggregate demand shocks, employment would be restored by cutting real wage below where it was prior to a particular negative demand shock. The assumption is also made that workers can determine their real wages as opposed to simply their nominal wages.

In terms of the narrative of this chapter, what’s critical for causal analysis and policy, is whether or not decisions-makers are rational and the implications of this for analysis and policy. The pre-Keynesian and new classical economics perspectives assume that rational agents would endeavour to clear all markets (prices are flexible) and behave as if prices are flexible. Hence, if unemployment exists or if it increases, this is related to the rational decision to keep real wages too high, for example. Here, unemployment or increases in unemployment are voluntary. There can no substantive demand side problem, especially in the longer run. The assumption here is that increases in the unemployment rate is a product of changing preferences of workers in favour of more leisure or non-labour market activities or government interventions that make labour markets less flexible and/or increase the real wage above what would be generated in a ‘pure’ market economy. The increased real wage is predicted to increase the rate of unemployment. Such institutional interventions (minimum wages and unions) increase the structural rate of unemployment. Here too the demand side is not of importance.

A popular rendition of this perspective was put forth in Friedman’s classic (1968) article, making a case for supply side determinants of macro outcomes. He focuses on what he refers to as the natural rate of unemployment, which is determined by structure of real wages. Not much attention is paid to severe negative demand shocks. Ultimately, if workers wanted more employment they should and would cut their real wages. It is assumed that this would not have a knock-on affect of reducing aggregate demand and therefore further increasing the rate of unemployment. Moreover, as unemployment increases, even dramatically so, it can be attributed to changing preferences of workers in favour of more leisure time or changing government policy that permanently increases real wages to higher levels—facilitating workers’ preferences for more leisure time.

Notice that amongst the old and new classical economists and, amongst many Keynesian economists, there is a prior assumption that decision-makers are rational, but the understanding of rationality differs across schools of thought, with significant implications for policy. Across the board, Keynesians regard spikes in unemployment yielding substantive increases in the unemployment rate to be involuntary. These increases in unemployment would be impossible for the market to deal with quickly and efficiently; that is in the real world of complex and asymmetric information, limited foresight, inflexible prices and the consequential reliance (to a lesser or greater extent) on decision decision-making heuristics, such as herding. There is no evidence that markets naturally clear swiftly after a severe demand-side shock. But the classicals assume that this reflects the preferences of decision-makers (there is very little modelling attention paid to different preferences and different power relationships across agents). This adds weight to the argument that one’s definition of rationality, what it means to be a smart decision-maker, and the realism of one’s modelling of the decision making process, is vitally important for causal analysis and, in the macro domain, for public policy.

A core Keynesian argument is that increasing demand either through monetary or fiscal policy will restore the economy to full employment in a relatively quick and efficient manner. Hence, the excessive demand-side related unemployment would be eliminated, and the economy restored to the prior and lower natural rate of unemployment. The higher unemployment rate that is realized during a depression or deep recession is not the natural rate of unemployment—which is the claim of the classical economists, old and new.

A critical assumption made by Keynesian economists is that for involuntary unemployment to be eliminated, workers must accept lower real wages, as increasing employment requires the formerly employed less productive workers (lower marginal product) to accept lower real wages. It is assumed here that a downward sloping marginal product of labour curve, over its relevant portion, characterizes the representative firm, which is a very big short run assumption indeed. The decreased real wage must coincide with adequate increases in aggregate demand. Classical economists argue that accepting lower real wages would not be the rational response of the typical worker. Hence, increasing aggregate demand can have no real effect on the economy, measured by increased employment. But the side-effect of such activist demand-side policy would be increased prices or increasing the rate of inflation.

Akerlof has attempted to provide a scientific quasi-rational basis for government policy to restore employment towards its pre-recession levels (Akerlof, 2002). He maintains that workers’ in some sense suffer from money illusion (quasi-rationality) and will therefore not pay attention to reductions in real wages that is a function of low rates of inflation. Basically, the transaction costs of computing the impact of low rates of inflation on real wages are not worth the benefits. Hence, increasing aggregate demand to increase employment should be effective so long as one buys into the realism of this transaction cost-based money illusion argument.

Decades earlier, Keynes rejected any presumption of money illusion on the part of workers, although he accepted the assumption that real wages need to be decreased for pre-recession or depression rates of unemployed to be restored. Workers would accept cuts to real wages that were generalized across sectors and occupations as these would be seen as fair especially when accompanied by increased employment. This could be achieved through aggregate demand-side induced inflation. Workers, themselves, could not orchestrate such a cut in real wages. This would have to be affected through macroeconomic government policy. There is no money illusion here at all. Moreover, Keynes theorizes that self-imposed cuts to money wages would simply reduce aggregate demand, further dampening animal spirits and thereby, further increasing unemployment.

Keynes (1936, pp. 14–15) argues:

…they [workers] do not resist reductions of real wages, which are associated with increases in aggregate employment and leave relative money-wages unchanged, unless the reduction proceeds so far as to threaten a reduction of the real wage below the marginal disutility of the existing volume of employment. Every trade union will put up some resistance to a cut in money-wages, however, small. But since no trade union would dream of striking on every occasion of a rise in the cost of living, they do not raise the obstacle to any increase in aggregate employment which is attributed to them by the classical school.

Simply because nominal wages are sticky in no way implies that real wages are not flexible enough in a world of rational (smart) agents, for employment to restored to pre-recession levels through monetary and fiscal policy. Increased longer term unemployment need not be a product of workers suddenly shifting their preferences towards more leisure, but rather of misconstrued macro policy that misreads sticky nominal prices (especially wages) with sticky real wages.

On a related note, given the empirics and theory underlying x-efficiency theory, even if real wages increase as aggregate demand increases, if this is accompanied by compensating increasing in labour productivity (a rational response by economic agents), increasing real wages would not impede the employment of more workers as aggregate demand increases. In this case, increasing real wages will not affect economic capacity of the firm to hire more workers on the margin. The marginal product of labour curve shifts to the right as real wages increase (Altman, 2006a, 2006b). Here too, by assuming rational individuals, one cannot logically deduce that increasing unemployment is a function workers’ preference for more leisure. Rather, a large reduction of aggregate demand requires a compensating increase in aggregate demand, given that rational or smart workers pose no fundamental obstacle to restoring employment to its pre-recession levels. This x-efficiency perspective strengthens the rational worker approach presented by Keynes in his narrative on workers accepting generalized, fair cuts to real wages, given the expectation that employment will increase as a consequence.

In this instance, rational inefficiency becomes a product of government not pursuing policy that restores aggregate demand, in the face of rational decision-making at the firm level. The latter is a product of the belief by government decision-makers in the capacity of markets to self-correct and that the ultimate source of the persistence in the increased level of unemployment following a severe economic downturn is the unwillingness of workers to reduce their real wages. This belief in the classical model might be rational given the information set of decision-makers. But they yield economic inefficiencies at the macroeconomic level, keeping unemployment rates unnecessarily high and output well below what it might otherwise be. Thus, government unnecessarily causes great economic and social harm to large segments of its population in the pursuit of a misconstrued economic model.