Training is an investment in “embodied human capital” – assets stored in workers’ heads.
With indentured servitude now deservedly out of fashion, workers can choose to walk out the door without the boss separating those assets from their shoulders.
Why would employers invest in assets that can quit? In a simplified labour market model, employers have no reason to. Workers will own the resulting human capital. So (with some bold simplifying assumptions about competitive labour markets) they can capture the full benefit of any training investment by demanding higher wages for their now-more-productive labour.
In that model world, workers should bear the full cost of training (in cash or reduced wages and conditions while they train on the job). More sophisticated models introduce aspects of imperfect competition. Firms and workers then share the productivity benefits of training, but the level of training investment is less than socially optimal because workers and their current employers can’t expect to capture the full benefits over time.
In the real world, training costs are shared by workers, employers and taxpayers. Competing firms may work together to coordinate their training investments.
What leads employers and taxpayers to pay a major share of training costs? And how might cost shares and incentives to invest in training change as technology (emergent, feared or hoped for) reshapes the future of work?
- Expectations that an employment relationship will endure. If workers stay longer, employers are more likely to recoup their training investments. Positive incentives for workers to stick around can include pay and benefits that increase with length of service, and other things they value about the quality of their jobs. Barriers to finding another job, such as thin labour markets, high search/relocation costs, or reputational penalties for quitting, will also mean workers are more inclined to stay. Employment protections that restrict lay-offs or increase their cost can make retraining a more attractive option to employers. And if workers and employers build trust and team spirit, both may invest more in training, confident that the other won’t renege by seeking an unfair share of the benefits.
- Co-production of output and skill, where people can learn while doing their day job, mean it’s more efficient to train at work rather than elsewhere, off-the-clock. This creates room for deals between workers and employers about how they share the costs of training.
- Firm-specific knowledge and skills, which are worth more within the firm than if taken elsewhere. Workers will under-invest in firm-specific training if they bear all the cost themselves.
- Productivity spillovers within the firm, where each worker’s skill boosts the productivity of the people and capital they work with, or the firm’s ability to create or adopt new technologies.
Government’s reasons to invest in training increase with:
- Broader spillovers including:
- Social and fiscal benefits (better parenting, social engagement and tax revenues), and avoided costs (welfare costs, healthcare, etc) of a more skilled population.
- Broader productivity spillovers and coordination failures, where increasing workers’ skills boosts the productivity of workers and firms in clusters or across the economy. Government may intervene with regulation, levies and funding, and firms may try to coordinate to address “free rider” problems.
- Information and choice issues – people may systematically under-invest in skills, even for their own best interests, if they are myopic or don’t understand the payoffs of training.
Risk and uncertainty are also reasons to share the cost of training. Where the future value of a new skill is uncertain (e.g. mastering a new technology, or where the value of a skill outside the firm is unknown), it may pay firms and government to share the risk of training investments with workers.
How does technological change affect investment in workforce training and how training costs are best shared?
There are three main areas. The direction and pace of change in each area is not obvious. This depends on how specific technologies evolve, on business models and on market dynamics.
- Technology changing the skills firms need.
This can go three ways (all at once):
- creating demand for new skills to manage new technologies, tasks and products;
- devaluing the skills of firms’ existing workforces (skill obsolescence); and
- opportunities to leverage workers’ existing skills (e.g. new user interfaces that enable lower-skilled workers to perform more complex tasks).
- Technology changing how labour markets operate.
Digital labour market platforms (about which, more in a later post…) make it easier for both workers and employers to shop around for a “good match”. Platforms can also facilitate a shift from employment-based to more contract-based “outsourced” work relationships. These features of platforms could reduce firms’ incentives to invest in training, and drive workers to pursue more transferrable skills.
- Technology reshaping the learning process.
Digital tech has the capacity to drive major changes in the nature cost and quality of training, if other factors (including social perceptions and institutional settings) allow. Aspects up for change include programme design, modes of delivery and assessment, credentials, costs (in time and money), and how learning integrates with work and leisure. This may unpredictably alter the cost sharing calculus, and the optimal level of investment in training.1
Facing uncertainty, firms will need to make choices between strategies that treat their workforces as assets on which to build competitiveness, or as costs they should seek to minimise. The choices firms make are mediated through government policies and the values of communities and of firms themselves.2
Changes across the economy as a whole may differ significantly from the examples that get the most attention in the media and from those in the business of surfing each hype cycle’s crest.
Jobs aren’t all being stolen by robots or turning into “gig work”. Despite (or especially because of) ongoing technological change, there are some inherent advantages in “traditional” employment arrangements with their longer-term mutual commitments, greater job security and greater task flexibility. Employers, workers and government will all continue to invest in training and skills, on and off the job. But yesterday’s models need to evolve to capture the opportunities and manage the challenges that technological change poses for the future of work.
Some relevant reading:
There’s great stuff on platforms, intelligent tools, and the future of work in the working papers series from the Berkeley Roundtable on the International Economy (BRIE) at the University of California, Berkeley. 3
The first issue for this year was Beyond Hype and Despair: Developing Healthy Communities in the Era of Intelligent Tools. John Zysman, Martin Kenney and Laura Tyson look at the factors that may promote the development of “good jobs” business strategies as tech platforms and intelligent tools change economies and societies.
For a primer on returns to training, see:
Blundell, R., Dearden, L., Meghir, C., and Sianesi, B. (1999) Human capital investment: the returns from education and training to the individual, the firm and the economy, Fiscal Studies 20(1) 1-23.
- But I’ve haven’t met many educators who can imagine technology (or anything else, really) lowering the cost of education!
- Zysman, Kenney & Tyson (2019) p2, 15-16.
- Cal is my alma mater. Happy 50th jubilee, Goldman School of Public Policy. Go Bears!
This post was originally published on the Productivity Commission's website.