The study of the interrelation between human capital accumulation, training and productivity has a long-standing tradition in economics (Becker, 1964). Among workers, on-the-job training is the building block of the human capital accumulation process. In fact, almost half of the human capital individuals accumulate in their lifetime is associated with investments and activities related to work (Heckman, 1998). However, it was not until recently that the economic literature focused on the role of skills as key underlying drivers of the human capital accumulation process. During the last decade, a growing body of research has documented the essential role of socio-emotional skills – commonly referred to as “soft” skills – as determinants of labor market productivity, both at the individual and aggregate levels (Heckman and Kautz, 2012, Daruich, 2018). These include dimensions such as communication, teamwork, and planning and organizing, among others. This has triggered new interest in the effects of on-the-job training programs as, in principle, they impact labor market outcomes through skill development (Bassanini et al., 2005, Barrett and O’Connell, 2001, Dearden et al., 2006, Konings and Vanormelingen, 2015).
In this paper we use a field experiment to estimate the impact of a training intervention in- tended to modify critical socio-emotional skills – leadership and communication – among workers of one of the largest retailers in Latin America. Through an interactive five-day program, managers from a randomly selected sample of stores participated in activities specifically designed to improve their skills to succeed as team leaders. The program encourages the managers to think about their capabilities and provides techniques to develop new management skills. Likewise, through a two-day program, selected sales associates were trained to develop critical communication skills. The technical content of the sessions includes themes such as complaint handling, advising customers, and understanding the importance of customer service as core to success.
The empirical analysis is carried out using longitudinal information gathered by the firm and through two skills surveys. The Identification exploits the experimental design in the context of a difference-in-difference strategy. The results indicate large positive effects of the training program on store-level productivity. We further link these Findings to individual-level performance measures. In particular, we document positive effects on total sales and numbers of transactions for all workers. Regarding the mechanisms, we provide evidence suggesting that the intervention was more effective in boosting leadership skills than communication skills. Spillovers from trained managers to untrained sales representatives also contribute to the main effects. Our findings point towards the possibility of increasing productivity through training programs targeting critical skills.
The following is a list of our main findings:
• From the store-level analysis we conclude the intervention increased daily sales by 119 dollars in locations where only managers were treated -T1- and by 204 dollars (12,1%) when both sales associates and managers received training -T2- (relative to control stores). The results are robust to different sample definitions and, in general, statistically significant at conventional levels (the exception is T3, i.e., the treatment group where only sales associates received training).
• Positive effects also emerge using total daily transactions as the outcome of interest, suggesting an impact on the extensive margin of sales. In this case, the largest estimated impact comes from stores in which only managers were assigned to training (11% increase in total transaction per day).
• To a large extent, our worker-level estimates confirm the previous findings. For sales, the estimated impact of the training program is unambiguously positive (and statistically significant) across the different treatment groups. For the number of monthly transactions, we document positive effects as well. Statistically significant results are obtained for sales associates and in the regression that pools both managers and sales associates. The training for both the managers and the sales associated had an unambiguous effect on the intensive margin. However, our results suggest that only the training for sales associates impact the extensive margin.
• Importantly, the individual analysis allows us to investigate the direct impact of the intervention on skills. While for sales associates we do not find any indication of gains in communication skills, among store managers we do document a positive and significant impact on leadership/management skills equivalent to 0.4 standard deviation.
• We estimate positive spillover effects from trained store managers to untrained sales personnel, which are consistent with our store-level results from T1 stores (only managers received training). The managers training not only impacted their own sales but also had an indirect effect on the number of transactions and sales of the untrained workers in the stores. Since the identification comes from individuals switching stores after the training program was implemented, we mitigate potential threats to identification by controlling for individual- and store-level fixed effects.
• Our study highlights the important role of managers within the stores. Particularly, it shows that training them has both direct and indirect effects on productivity. Trained managers not only have large effects on untrained workers, but as shown by our store- level results, there might be complementarities with the enhanced skills of sales associates. These results are in line with the recent literature on the importance of managers to explain worker’s productivity (Lazear et al., 2015) and on the effectiveness of management practices (Gosnell et al., 2019). Moreover, they highlight the multiple effects of manager’s training on the productivity of the firm.
• Even though a large fraction of the impact on sales can be explained by the extra units of management skills, a significant proportion of the impact remains unexplained. We argue this can reflect both the limitations of our skill measures as well as the potential role of spillovers within stores. In this context, a future exploration of changes in the composition of the work force within each store during the post-treatment period might provide additional insights explaining the overall impact.
In sum, this paper shows not only how a tailored made training interventions can shape workers’ performance but also the effectiveness of investing in soft-skills as they can trigger productivity gains. More generally, the analysis sheds light on three critical aspects to be considered for public policies aimed at improving labor market outcomes: the potential impact of public-private partnerships, the importance of institutional and managerial capacities, and the fundamental role of quality assurance mechanisms. Although the external validity of this type of study is not guaranteed, our findings suggest it is possible for a well-targeted on-the-job-training intervention to positively impact workers’ productivity levels.
Chosen excerpts by Job Market Monitor. Read the whole story at Training, Soft Skills and Productivity: Evidence from a Field Experiment in Retail