Knowledge-intensive sectors and the role of collective performance-related pay

ABSTRACT

The main contribution of this study is showing that the efficiency effects of collective performance-related pay (CPRP) are more pronounced in knowledge-intensive service sectors (KISs) than in other sectors. The hypothesis is that human resource practices such as CPRP are particularly useful for enhancing firm performance when innovation-supporting knowledge is distributed among multiple skill sets and employee creativity, knowledge creation and knowledge sharing are key success factors for the firm. Cross-sectional estimates obtained for a national sample of approximately 3,800 Italian firms confirm this prediction. These results are validated by adopting a treatment effect approach to solve the self-selection problem.

KEYWORDS:

• Collective bargaining
• performance-related pay
• firm performance

JEL:

• D23
• J33

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Wang, Zhao, and Chen (2017) also demonstrated that firms with greater specificity in knowledge structure need to properly design the compensation structure to improve firm economic performance. They showed that specificity in firm knowledge assets is positively associated with the use of restricted stocks in CEO compensation to discourage CEO dismissals.

² For aggregation of the manufacturing industry according to technological intensity and for aggregation of services by knowledge intensity, see Eurostat indicators on the high-tech industry and knowledge-intensive services, http://ec.europa.eu/eurostat/cache/metadata/Annexes/htec_esms_an2.pdf.

³ Gibbs (2016) collects the different views of a group of leading economists in this area of research, such as Kevin Hallock (Cornell University), Edward Lazear (Stanford University), Kevin Murphy (USC), and Canice Prendergast (University of Chicago).

⁴ One channel that removes these uncooperative actions is accessed when shared capitalism and collective performance-related pay schemes are adopted in combination with other complementary HRM practices. However, in the present paper, the availability of data does not allow to control for the combined role of HRM practices.

⁵ These activities have been grouped by Eurostat in the subset of knowledge-intensive business activities (KIBS) but are not considered separately from the other knowledge-intensive activities in our empirical analysis. From a thorough analysis of KISs, it clearly emerges that complex knowledge, teamwork and the need to share knowledge are widespread across all sectors included in KISs, as also documented by the case studies for different industries, such as healthcare and education, examined by the OECD report (2006).

⁶ Job titles (livelli di inquadramento) are defined by national industry-wide collective bargaining contracts, for which specific minimum wages apply. They are assigned according to different qualifications and skills of workers.

⁷ According to the SICA-LCS questionnaire firms can use more than one targets and criterion when they implement CPRP. In addition, data are collected for both cash variable compensation and employee share ownership plans, that involve employees participating in firm property rights. However, the second type of incentives are quite unusual in Italy and are not examined in this study. Finally, note that to avoid repetitions we use interchangeably terms such as collective bonuses and collective incentive pay schemes, although, from what we discussed above they are not exactly the same thing.

⁸ Unfortunately, the SICA-LCS survey does not provide similar information for individual incentives.

⁹ The return on sales is the ratio of gross operating margin on total sales.

¹⁰ The Eurostat Hi-tech classification of industries is available at

http://ec.europa.eu/eurostat/cache/metadata/Annexes/htec_esms_an3.pdf. To avoid excessive asymmetries across the classes size, we had to collapse the original four classes (high-tech, med-high-tech; med-low tech, low-tech) into two ones.

¹¹ It is worth noting that the values for TFP in levels are very close to those found in the related literature dealing with TFP estimation for Italian firms (Altomonte and Aquilante 2012; Aiello, Pupo, and Ricotta 2014).

¹² It must be remarked that we only have a binary variable for export (1 for an exporting firm, independent of the proportion of sales that come from selling abroad, and 0 otherwise). In addition, only firms with ten employees or more are in the sample.

¹³ Thus, the new variable is the past profitability, that we only use as instrument. However, the set of excluded instruments also includes covariates such as unions, tenure and training that proved to be highly correlated with CPRP and not influencing economic performances in OLS and IPWRA specifications (see Section 5).

¹⁴ According to Chatterjee and Hadi (2012), the presence of multicollinearity relies on two rules of thumb, which are used to interpret the VIF test. Multicollinearity is signaled by the largest VIF (among variables) being larger than 10 and the mean VIF being considerably larger than 1 (above 5); Table A2 shows that this is not true in our case.

¹⁵ These results are available upon request.

¹⁶ It means that the specification in IV_ATET is slightly different from that we use in the OLS and IPWRA estimations. In the former, unions, tenure, training and EGR only enter the set of excluded instruments. Since the purpose of this robustness check is only to test the potential strong endogeneity of CPRP that is not corrected by the IPWRA method, we do not care about differences in the size of the CPRP_ATET coefficients of Table A4 driven by differences in the model specification.

¹⁷ We obtain this result despite the relevance detected in the instruments set. Table A5, in the Appendix, reports findings for the first step probit, performed to calculate residuals ${\hat{\upsilon }}_i$. The five instruments (i.e. ROS_2007-2010, unions, EGR, training and tenure) are strong determinants of the treatment (CPRP) in the whole sample. Past profitability, unions and training remain highly significant drivers of CPRP adoption in KISs.