The dynamics of profits and wages: technology, offshoring and demand

Tuesday, 15 November, 2016

Francesco Bogliacino, Dario Guarascio and Valeria Cirillo

1.     Introduction*

Over the last decades, economic inequalities have dramatically increased across both advanced and developing economies (Atkinson, 2015; Piketty, 2014). This evidence raises interest (again) in the dynamics of income distribution and its drivers (Franzini and Pianta, 2016). Looking back in time, classical economists such as Marx, Ricardo, or heterodox scholars such as Kalecki, regarded income distribution as the fundamental feature of capitalist economies, because distributive arrangements tend to reflect the interaction between the economy’s main driving forces.

On the other hand, most of the explanations for the recent increase in inequality focus on what happened in the labour market, looking, in particular, at the role of trade and technology. The former is important because of globalization, which forces advanced economies to face competition from abundant unskilled labour countries and threatens workers with the risk of offshoring of their jobs. The latter matters because the nature of innovations may favour some skills over others, a phenomenon popularized as the Skill Bias Technical Change (SBTC hereafter).

However, trade per se is not a good explanation because, contrary to textbook predictions, inequality in the labour market increased both in the EU and the US and in unskilled labour abundant economies such as China. In addition, most of the increase in inequality occurred within industries, and cannot be associated with labour reallocation across sectors (Acemoglu, 2002). Similarly, the SBTC hypothesis is unsatisfactory because the main driver behind the increase in wage inequality  turned out to be the set of labour market reforms that weakened workers’ bargaining power (Fana et al., 2016) rather than a technology-related displacement of medium skilled workers (Bogliacino and Maestri, 2014; OECD, 2011).[1]

From an empirical perspective, most of the recent studies on inequalities have focused on personal rather than on functional distribution of income. This occurred despite a growing importance of functional distribution in explaining the change in inequality (OECD, 2008; 2011). The personal distribution concerns how income is distributed across households, while functional distribution addresses how production factors – i.e. capital and labour - are remunerated.

The aim of this work is to discuss the existing consensus, starting from the premise that functional distribution is perhaps the more relevant part of the story of inequalities. In our framework, technology and trade are two key forces shaping the dynamics of income distribution. However, their impact depends on the relative power of the parties (i.e. capital and labour) involved in the bargaining process. Conceptually, we identify a set of key structural drivers shaping income distribution: i) the balance of power between capital and labour, linked to market structure and workers’ unionization ii) the dynamics of technological change affecting distribution through new products - leading to monopolistic rents shared between workers and capitalists according to their relative bargaining power - or new processes - making production more efficient but with the risk of destroying jobs and demand outpacing efficiency gains iii) the degree of openness of the economy, ensuring external demand flows (exports) and allowing firms to offshore parts of production as well as to choose foreign rather than domestic suppliers iv) quality of institutions, size of the welfare state and workers’ human capital endowment, all elements potentially smoothing the distributive set-up (Howell, 1999; Atkinson, 1999).       

We develop a structural model founded on two primary building blocks and characterized by a sequential timing. Wages are  negotiated before entering into production and take into account the constraints dictated by total employment, output decisions and available or expected rents (related innovation and organization of production). Profits are realized afterwards, depending, of course, on the surplus residual (in a Ricardian sense, i.e. after paying wages) and on demand level.

On the wage determination side, we are guided by Van Reenen’s hypothesis (1996) regarding ‘innovation rents’ captured by workers (in a similar vein, see Dunne and Schmitz, 1995).[2] Innovative rents are defined in Schumpeterian terms, and they should be derived from the temporary monopoly associated with a new product (Schumpeter, 1942). Van Reenen (1996) identifies three fundamental reasons why workers have legitimate access to portions of innovation rents: i) the time lag between input, R&D activities and output of innovation; ii) the difference in time horizon between workers and shareholders, which is shorter for the former due to the diffusion of temporary contracts; iii) the elements of randomness in the nature of innovation.

The second element, which affects the dynamics of wages, is offshoring. The latter can impact the dynamics of labour remuneration through three different channels. The first is a negative effect linked to the ‘threat’ faced by workers as employers have the opportunity of offshoring parts of production. Such a ‘threat effect’ is likely to reduce workers’ bargaining power, exerting a negative impact on wages - particularly of low-skilled ones that are relatively more substitutable. The second refers to a positive relationship that can emerge, again, as a consequence of the offshoring of more labour-intensive (and low-skill intensive) parts of production. In this case, a change in the skill composition that favours high-skilled workers is likely to positively affect the dynamics of average wage – i.e. the so-called skill composition channel (Fosse and Maitra, 2012). Finally, offshoring can work as an organizational innovation – that is, the inflow of foreign intermediate inputs incorporating new technologies may encourage the adoption of more efficient work practices – pushing upward the wages of workers adequately endowed to benefit from such innovations (most likely medium and high skilled workers). Overall, offshoring can exert both a negative and a positive effect on wages. [3] The prevalence of one or another effect depends on the relative bargaining power of capital and labour, workers’ skill endowment and technological characteristics of the firms and sectors involved (Pianta and Tancioni, 2008).

Concerning profits, we followed a standard Post-Keynesian approach. As a result, profits are driven by demand, which realizes outstanding surplus left after wage bargaining. Of course, there exists a heterogeneous impact exerted by domestic and foreign components of demand (Bogliacino and Pianta, 2013; Guarascio et al., 2015). Given the sequential structure put forth, profits are also determined by social conflict with labour, as well as by lagged internal investments capturing embodied technical change (Dosi, 1988).

We apply our framework to industry-level data for five European countries (Germany, France, Italy, Spain, and United Kingdom) over the period 1995-2010. Our database merges data from the Community Innovation Survey, OECD STAN and WIOD, thus allowing for the measurement of different sources of demand, technology and offshoring (more details on the adopted database are provided in the next section). The model is run relying on novel econometric techniques ensuring consistency in the estimation of our complex set of relationships. Moreover, the proposed approach allows feedbacks to be identified among the main variables – i.e. capital and labour remuneration – capturing both conflictual elements in the bargaining process as well as direct and indirect effects of technology and offshoring.

The article is organized as follows. In the next Section, we briefly illustrate the data and provide some descriptive evidence. In Section 3 the econometric strategy is presented and the results summarized. Finally, Section 4 concludes with some policy implications stemming from our findings. 

2.     Data and methodology

The data used in this work are drawn from the Sectoral Innovation Database (SID) developed at the University of Urbino (Pianta et al., 2011). The SID combines different data sources using the sector as the unit of analysis - two-digit NACE classification for 20 manufacturing and 17 service sectors. For innovation variables, such as R&D expenditure, average firm size and expenditure on new machinery and equipment, data stem from four European Community Innovation Surveys—CIS 2 (1994-1996), CIS 3 (1998-2000), CIS 4 (2002-2004) and CIS 6 (2008-2010)—and subsequently matched to industry-level data from the WIOD Nace Rev. 1 database.[4]

For production and demand variables - that is, wages, profits and demand - we use data from the World Input Output Database (WIOD). All data have been converted into euros and constant prices. The country coverage of the database includes five major European countries (Germany, France, Italy, Spain, and United Kingdom) covering 71% of the entire EU’s GDP. The selection of countries and sectors has been made to avoid limitations in data access (low number of firms in a given sector for a given country or because of the policies on data released by various National Statistical Institutes).

In the following we provide a descriptive picture of the main relationships investigated here. In Figure 1, we report two panels: on the left hand side, the dynamics of wages by industries’ R&D and offshoring intensity. On the right hand side, instead, the change in profits is analyzed, distinguishing industries according to their export intensity.[5]

As expected, Figure 1 (LHS) shows a positive and statistically significant association between R&D and wage growth, and a negative (and statistically significant) one between the latter and offshoring intensity.[6] Moreover, a strong association between profits and export growth emerges.

 

The relation between wage dynamics and offshoring is displayed, in greater detail, in Figure 2. We compare the average wage growth - for above and below the median offshoring level - of low, medium and high skilled workers. High, medium and low skilled workers are defined according to educational attainment (ISCED categories). The evidence shown in Figure 2 depicts the negative impact of offshoring on low skilled wages.

 

As can be seen from Figure 2, a more complicated picture emerges when workers are distinguished according to their skills. High skilled wages, in fact, display a positive correlation with offshoring intensity. Contrarily, offshoring is negatively associated with the change in low skilled wages, suggesting the presence of different channels at work, as we discussed in the Introduction. In the paper, we move beyond simple statistical analysis and we consider a number of frontier results in the econometric literature to identify the impact of technology and offshoring on wage growth and to quantify the role of demand in the growth of profits.[7] The next Section provides a synthetic description of the adopted methodology and a summary of the results.   

3.     Econometric strategy and results

We estimate a two-step structural model where wages are obtained first, using as explanatories innovation, offshoring and variables capturing industry-specific economic and demand dynamics.[8] Afterwards, we estimate profits including wages stemming from the previous step – incorporating the effect exerted by technological and offshoring factors - beside demand components (domestic and exports) and internal investments.

Our empirical strategy relies on standard instrumental variables and the recently proposed heteroskedasticity-based instrumental variables approach (Lewbel, 2012). Identification is achieved with the use of regressors not correlated with the product of heteroskedastic errors. With this approach, atheoretical instruments can be generated, and proper statistical tests can be provided for both the heteroskedasticity requirement and the over-identifying restrictions.

Table 1 reports a summary of the results of the structural model estimation. The latter are collapsed reporting key relations and main findings. The signs in brackets correspond to the direction of the relationships among variables as they emerged in the econometric model. The asterisks signal the significance of such relations.  

 

Table 1. Summary of the results – structural model estimations

 

Equations

R&D intensity

Offshoring

Interaction (R&D * offshoring)

Domestic demand

Exports

Wages per worked hour (%)

Wages (%) – High skilled

Wages (%) – Med skilled

Wages (%) – Low skilled

(1) Wages per worked hour (%)

(+)***

(-)*

(-)

 

 

 

 

 

 

(2) Wages (%) – High skilled

(+)***

(-)*

(-)

 

 

 

 

 

 

(3) Wages (%)– Medium skilled

(+)***

(-)*

(-)

 

 

 

 

 

 

(4) Wages (%)– Low skilled

(+)***

(-)*

(-)

 

 

 

 

 

 

(5) Profits (%)

 

 

 

(+)*

(+)***

(-)***

(+)***

(+)

(-)*

 

Note: full results are in Bogliacino et al. (2016). Row (1) refers to the 3SLS estimation of wages per worked hours and total profits; Rows 2-5 refer to the 3SLS estimation of wages per worked hour - distinguished in High, Medium and Low skilled - and total profits; the negative and significant impact of wages per worked hours on profits (Row 5) refers to the 3SLS estimation of wages not distinguished by skills and total profits. Asterisks report significance levels:  *** p< 0.01, ** p<0.05; * p<0.1. 

As shown by Table 1 we can put forth some key evidence: (1) there is a contrasting impact of offshoring—pushing wages downward—and innovation—pushing wages upward; (2) the presence of a non-linear effect in the R&D-wages relation; (3) social conflict matters, as captured by the negative effect of wage growth on profits; (4) the fundamental role of demand, particularly exports, as a driver of profits.

The most noteworthy element is that, to the best of our knowledge, the analysis undertaken here is the first attempt to measure the simultaneous impact of innovation and offshoring on wages by skill, while still accounting for the wage-profits bargaining conflict.

The most important findings are: a) consistent with the rent-sharing hypothesis formulated above, innovation spurs high- and medium-skilled wages, yet it is not correlated with low-skilled ones; b) high-skilled wages are found in relatively higher “offshoring intensive” industries, for they seem to benefit from the improved efficiency likely to be associated with production offshoring, while low-skilled wages tend to decrease in the same sectors, which points to the prevalence of a “threat effect” that hinders low-skilled workers’ bargaining power (though this does not speak to the situation of medium-skill wages—see SBTC literature for more information); c) the interaction between R&D efforts and offshoring, which is not significant in all other specifications, has a negative and significant impact on low-skilled wages, thus confirming the downward pressure exerted by offshoring on these wages; d) the wages-profits relationship undergoes far-reaching changes when skills are taken into account, in the sense that once these differences are accounted for, the heterogeneity in bargaining power softens the clear cut class clash between capital and labour tout court.

4.     Conclusions and policy implications

In this paper, we use a novel dataset at industry level, and we exploit some frontier econometric results to investigate structural determinants of distribution between wages and profits. Social conflicts, innovation, offshoring and demand emerge as key determinants of the theoretical relationship underpinning distribution.  

According to the causal claims discussed above, a more strictly regulated labour market is meant to be beneficial to the economy. In particular, the need emerges for stronger protection of workers from dismissal, and to encourage firms’ investment in human capital, diffusion of innovation and technological competitiveness strategies. This would result in a fairer distribution of innovation-related rents, which can help in reducing inequalities. Moreover, the strengthening of trade unions can mitigate the pressures for offshoring of production and avoid negative effects on employment, wages and demand. Finally, a public intervention aimed at ensuring a sustained dynamic of demand turns out to be fundamental to stimulate profits expectations and, as a consequence, new investments.

Summing up, there is increasing theoretical and empirical evidence in favour of the hypothesis that centralized bargaining mechanisms and strong unions can be good for innovation and growth. A complementary policy instrument could be related to the use of industrial policies. They can display their positive effects through two channels: i) a more equal income distribution - and higher wages - sustain domestic demand; ii) a cooperative environment within firms that encourages on-the-job skills upgrading, workers’ cooperation, and the adoption of new technologies and innovations. The empirical studies stress how this positive association is particularly evident in Europe. In this light, labour market policies supporting less adversarial social relations and strengthening unions’ bargaining positions may represent a key element within an overall policy for long-term sustainable growth. 

References

Acemoglu, D. (2002) Technical Change, Inequality, and the Labor Market, Journal of Economic Literature, 40(1), 7-72.

Akerlof, G. A. and Yellen, J. (1990). The fair wage-effort hypothesis and unemployment. Quarterly Journal of Economics, 105(2), 255-283.

Amiti, M., and Wei, S. (2004). Fear of Outsourcing: Is it Justified? (No. w.186). Washington: International Monetary Fund.

Antràs, P., Garicano, L. and Rossi-Hansberg, E. (2006). Offshoring in a knowledge economy. The Quarterly Journal of Economics, 121(1), pp. 31-77. MIT Press.

Atkinson, A. B. (1999) Is Rising Income Inequality Inevitable? A Critique of the Transatlantic Consensus. Wider Annual Lecture 3.

Bogliacino, F., and Maestri, V. (2014). Increasing economic inequalities? In W. Salverda, B. Nolan, D. Checchi, I. Marx, A. McKnight, and I. Tóth (Eds.). Changing Inequalities in Rich Countries: Analytical and Comparative Perspectives, pp. 15-48. Oxford University Press.

Bogliacino, F., Guarascio, D., Cirillo, V. (2016). The dynamics of profits and wages: technology, offshoring and demand (No. 2016/04). Laboratory of Economics and Management (LEM), Sant'Anna School of Advanced Studies, Pisa, Italy.

Bogliacino F. and Pianta M. (2013). Innovation and Demand In Industry Dynamics. R&D, New products and Profits. In A. Pyka and E.S. Andersen (Eds.). Long Term Economic Development. Berlin: Springer.

Burstein, A. and Vogel, J. (2012) International trade, technology, and the skill premium. 2012 Meeting Papers 664, Society for Economic Dynamics.

Dosi, G. (1988). Sources, procedures and microeconomic effects of innovation. Journal of Economic Literature, vol. 26, pp.1120-1171.

Dunne, T. and Schmitz, J. A, (1995) Wages, Employment Structure and Employer Size-Wage Premia: Their Relationship to Advanced-Technology Usage at US Manufacturing Establishments, Economica, London School of Economics and Political Science, vol. 62(245), pages 89-107, February.

Falk, M. and Koebel, B. M. (2002) Outsourcing, Imports and Labour Demand, Scandinavian Journal of Economics, 104(4): 567-586

Fana, M., Guarascio, D. and Cirillo, V. (2016). Did Italy Need More Labour Flexibility? Intereconomics, 51(2), 79-86.

Fosse, H. and Maitra, M. (2012). Import, offshoring and wages: rent sharing or composition? Mimeo, 2012. Available at: http://openarchive.cbs.dk/bitstream/handle/10398/8540/Fosse_2012_2.pdf?sequence=1

Geishecker, I. and  Görg, H. (2008). Winners and losers: A micro-level analysis of international outsourcing and wages. Canadian Journal of Economics, 41 (1), pp. 243–270.

Guarascio, D., Pianta, M., Bogliacino, F. (2015). Export, R&D and new products. a model and a test on European industries. Journal of Evolutionary Economics, 1-37.

Howell, D. R. (1999). Theory-Driven Facts and the Growth in Earnings Inequality. Review of Radical Political Economics, 31, pp. 54-86.

Lewbel, A. (2012). Using heteroscedasticity to identify and estimate mismeasured and endogenous regressor models. Journal of Business & Economic Statistics.

Mion, G and L Zhu (2013). Import competition from and offshoring to China: A curse or blessing for firms? Journal of International Economics, 89(1), pp. 202-215.

Munch, J. R. (2010) Whose Job Goes Abroad? International Outsourcing and Individual Job Separations. Scandinavian Journal of Economics, 112(2): 339-360.

OECD (2011). Divided We Stand: Why Inequality Keeps Rising. OECD Publishing.

OECD, (2008). Growing Unequal? Income Distribution and Poverty in OECD Countries. OECD Publishing.

Perani, C. and Cirillo, V. (2015) Matching industry classifications. A method for converting Nace Rev.2 to Nace Rev.1 (No. 1502). University of Urbino Carlo Bo, Department of Economics, Society & Politics - revised 2015.

Pianta, M. and Franzini, M. (2016). Disuguaglianze: Quante sono, come combatterle. Gius. Laterza & Figli Eds.

Pianta M. and Lucchese M. (2011). The Sectoral Innovation Database 2011. Methodological Notes, University of Urbino, Faculty of Economics. Discussion Paper.

Pianta, M. and Tancioni, M. (2008). Innovations, profits and wages. Journal of Post Keynesian Economics, 31(1), 103-123.

Schumpeter, J.A. (1975). Capitalism, Socialism and Democracy. New York: Harper (1st edn 1942).

Shapiro, C. and Stiglitz, J. (1984). Equilibrium unemployment as a worker discipline device. The American Economic Review, 74(3), pp. 433-444.

Sheng, L. and Yang, D. T. (2012). The Ownership Structure of Offshoring and Wage Inequality: Theory and Evidence from China, mimeo. Available at: https://www.econ.cuhk.edu.hk/dept/seminar/12-13/1st-term/shengyang2012.pdf

Slaughter, M. J. (2000) Production transfer within multinational enterprises and American wages. Journal of International Economics, 50(2), pp. 449–72.

Van Reenen, J. (1996). The creation and capture of rents: wages and innovation in a panel of UK companies. The Quarterly Journal of Economics, 111(1), pp. 195-226, February. MIT Press.

Figure 1. Mean annual rate of change of wages by intensity of offshoring and R&D (1996-2000; 2000-2003; 2003-2007; 2007-2010)

Source: Sectoral Innovation Database (Pianta et al., 2011).

 

Figure 2. Mean annual rate of change of wages by intensity of offshoring and skill group (1996-2000; 2000-2003; 2003-2007; 2007-2010)

Source: Sectoral Innovation Database (Pianta et al., 2011).  Low-Off. and High-Off. for high and low level of sectoral offshoring. 




* This note synthesizes the article ‘The dynamics of profits and wages: technology, offshoring and demand’ produced as part of ISIGrowth project on Innovation-fuelled, Sustainable, Inclusive Growth that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 649186 - ISI- Growth. The full paper is available at: http://www.isigrowth.eu/2016/02/11/the-dynamics-of-profits-and-wages-tec...  

[1] According to the SBTC theory, trade and technology could interact in affecting employment and wages: different machines may be more likely to replace certain occupations, negatively affecting those skills that have a comparative advantage in those tasks. In the most famous version of this theory, routine jobs tend to be offshored because of new technologies, and medium skills are those that suffer, generating a polarization in the labour market.

[2] Van Reenen’s hypothesis is predicated on the efficiency wage theory (Akerloff and Yellen, 1990). According to this theory, a causal relationship can be traced between wage level and workers’ on-the-job productivity. Employers are willing to pay wages above the market-clearing level in order to spur productivity growth; basically, worker productivity depends on wages received, implying higher wages provide stronger incentives for the worker to be productive. Furthermore, according to Shapiro and Stiglitz’s model (1984), a wage increase is shown to decrease a worker’s incentive to shirk. In other words, a wage increase boosts worker productivity and lowers direct monitoring expenses.

[3] The available empirical evidence on the relationship between technological change, trade and inequalities is mixed. Some contributions show consistency of the prevailing theoretical hypothesis with data, such as Slaughter (2000), Geishecker and Görg (2008), Mion and Zhu (2013). Some articles suggest large employment losses among low-skilled workers (Amiti and Wei, 2004; Munch, 2010; Sheng and Yang, 2012); others claim that the effect on wages is negligible (Antràs et al. 2006). Finally, some papers identify a positive effect on high skill wages (Falk and Koebel, 2002: Burstein and Vogel, 2012).

[4] In order to establish the requisite condition for comparability, innovation variables taken from CIS6 have been converted into Nace Rev.1 using the conversion matrix found in Perani and Cirillo (2015).

[5] The average annual rate of change of wages is analyzed, separating those sectors with an above the median R&D intensity from those with a below the median intensity, and the same for offshoring. To define offshoring we use a standard indicator: the share of only imported intermediates in a given industry from the same industry (corresponding to diagonal terms of the import-use matrix).

[6] The significance is tested using the Wilcoxon rank sum test. The latter assumes as null hypothesis the equality of distribution of the variables under comparison (i.e. wage growth in high and low offshoring or high and low R&D sectors. Results and further details are in Bogliacino et al. (2016).

[7] For a detailed methodological description, the reader is invited to check section 3 of the original paper (Bogliacino et al., 2016).

[8] We estimate the structural model using a Three Stage Least Squares (3SLS). This technique allows estimation of a simultaneous system of equations, addressing at the same time all the endogeneity issues. A number of econometric tests carried out adopting the state-of-the-art techniques to account for allmethodological issues are reported in Bogliacino et al. (2016).