The human resources economy

What would it mean to take talent allocation more seriously?

Photo by K. Mitch Hodge on Unsplash

How much does talent allocation matter for the economy?

It’s not every day I read something that makes me think about economics in a very different way. But encountering this claim in the opening pages of Tyler Cowen and Daniel Gross’ book Talent was a real “whoa-if-true” moment:

If we look at the growth in US output since 1960, by the best available estimates at least 20 to 40 percent of that growth has stemmed from the better allocation of talent.

Perhaps this is my failing, but I hadn’t taken seriously the possibility that getting the right people in the right jobs could be quite so important. For example, last year, I edited a report exploring the different things a Labour government could do to boost growth. It looked at creating a more favourable investment environment, fiscal and monetary policy, energy transition, infrastructure, planning, devolution - all the usual things. It looked a little at human capital, but as I think is normal, mostly through the lens of how to get more of it i.e. investment in skills and education, and using immigration to grow the labour force. It said little directly about how to allocate that human capital.

I’m not sure how literally to take that 20-40% benchmark.¹ Even if we take it at face value, it relates to a specific historical change - the impact of more women and Black men entering the workforce, and high productivity professions in particular. The study that produced those numbers suggests that fully eliminating the remaining barriers to women and Black men would be expected to increase US GDP by a further 9.9%, which would be of huge benefit to the economy, but at the same time implies that the bulk of the gains have already been achieved.

On the other hand, maybe structural racial and gender barriers are just the tip of the iceberg. What about the persistent disadvantages of class and income? And beyond that more general questions of ‘fit’? On a micro level, this sort of thing clearly matters a lot. I put a lot of thought and effort into finding the job that suits me best, and work hard to find the right people to fill vacancies at my organisation.

How confident should we be that the labour market functions efficiently to address these issues? Sexism and racism can be seen as systemic ‘market failures’, but are there others? You might point to the existence of a huge recruitment sector, from public Jobcentres to private headhunters - the best part of a million people in the UK work in “employment activities” - and say that they exist to address this problem. But equally, economists are comfortable saying that the billions of pounds and thousands of highly paid people working in financial services don’t allocate financial capital perfectly efficiently. And allocating human capital seems like a much more complex problem. If you’re allocating financial capital, you broadly only need to think about one side of the equation: money is money, you just need to find the best place to invest it.² If you’re allocating human capital, both sides of the equation are in play - you need to understand the needs of the labour market, but also the humans that fill them, with their personalities, commitments and constraints. I would be shocked if we’re getting this perfectly right.

At the very least, the public sector raises issues. Consider common arguments that we need to recruit better civil servants (supposedly like Singapore) or better teachers (supposedly like Finland). From a talent allocation perspective, that raises the question of where we get them from, and which sectors should take mediocrities in exchange. Sceptics might say that it would be better for the best and brightest to be generating wealth in the private sector rather than pushing paper around, or actually doing things rather than training others to do things.

I’m not sure how to test the effectiveness of allocation in the job market. Job satisfaction and perceived performance as rated by workers and managers can give us some indication. In particular, it would be interesting to see dispersal of performance (including in wages) within particular roles for people of similar qualifications - the wider this is, the greater evidence of variation in fit. But I think it’s hard to be sure how big a problem this is.

Two questions I have about how to take talent allocation more seriously

1. To what extent are we talking about intelligence vs non-cognitive abilities?

A central argument of Cowen & Gross’ book is that intelligence, while important is overrated. We have a system that is already highly geared - though almost certainly not optimised - towards finding and developing cleverness. Between school exams and a selective higher education system, we put substantial effort into credentialising the intellectual elite, and serving them up to elite professions. Those processes might also select for some combination of conscientiousness and drive. Yet, in a theme picked up by David Brooks in a recent essay, all these attributes are less predictive of success than we might expect.

There are any number of characteristics that might make someone well suited to perform well in a particular role - leadership, ability to read people, organisation - any number of familiar ‘soft skills’, and dozens we probably don’t yet have a name for. To give one example, Cowen & Gross coin the term ‘sturdy’ to describe the personality type of someone (like me!) who tends to produce output regularly and consistently rather than in bursts e.g. writing a few hundred words each day rather than lurching between essay crises. A doctor who tends towards sturdiness might be better suited to routine operations than they are to working in an emergency department, where things are more chaotic and adrenaline fuelled.

How well are we a) identifying people’s non-cognitive abilities; b) developing them, in light of their ambitions and the needs of the economy; and c) matching them to jobs? My bet is less well than we are doing on cognitive ability.

2. Is this a strong link game or a weak link game?

I’m a little bit obsessed with a distinction Malcolm Gladwell makes between two alternative educational frameworks. On the one hand, there are those who see education as a strong link game: we need to make the most of our best players, put rocket boosters under them and they will pull the rest of us along. On the other hand, some see education as a weak link game: we only make progress by strengthening the least capable. Consider post-colonial India. Establishing elite Indian Institutes of Technology is a strong link strategy. Trying to achieve universal literacy is a weak link strategy.

Where are we in mature economies like the UK? A book like Talent starts from a relatively elitist premise, seeking brilliant people to receive venture capital millions for their start ups. That makes sense if you think we inevitably operate in a ‘superstar economy’ where the success of a relatively small number of outstanding firms is the key driver of economic success i.e. if you think we’re playing a strong link game. But equally, given the long tail of unproductivity we have in our economy, it could be that a lot of the potential is in the middle and rear.

In his essay, David Brooks seems to assume that shifting focus away from intelligence will lead to a broader based reckoning of talents. Perhaps. But it is possible that non-cognitive ability could also be unevenly distributed and we end up with similar inequality on different dimensions.

On the other hand, I would love Tyler Cowen to consider the weak link implications of his ideas, and write a companion to Talent called Mediocrity, asking whether better matching can mean better use of the average workers in our economy.

What might we do differently if we wanted to improve talent allocation?

So earlier I said that none of the chapters in my essay collection directly addressed talent allocation. However, a couple of the topics covered implicitly bear upon the issue. Broadly speaking, if we want to improve talent allocation, we should probably want thicker labour markets - if workers have more jobs to choose from, and employers more workers to choose from, we should increase the chances of getting better matches. Building housing in or near big cities should therefore improve talent allocation. So should liberalising migration.

But merely removing barriers to employers and workers finding each other may not be sufficient. I think an economics that worried more about talent allocation would be more concerned about careers information advice and guidance. Certainly, there is encouraging evidence that students at schools that follow good practice in careers support are less likely to drop out of the labour market. In addition we might hope that they also achieve more suitable jobs, at which they stay longer and happier and more productive. But careers guidance isn’t just for young people: if we care about talent allocation, we should be more attentive to the support that adults get - or don’t. Career path dependency would seem to me to be a significant risk to effectively deploying human capital.

There are implications for the education system, too. I’ve previously been sceptical of the signalling function of education - using the education system to identify and certify a person’s pre-existing level of ability. But signalling plays a vital role in effectively allocating talent. My worry is that our modes of signalling do not adequately account for non-cognitive ability. Our education system gives most people and their potential employers a pretty clear sense of where they stand in intelligence hierarchy. But can our schools, colleges and universities give people a clearer sense of their other strengths and weaknesses (which ought then to dovetail with careers advice)?

Finally, I’d be surprised if labour allocation wasn’t related to management quality. Certainly, in my own experience as a manager, one of my most significant responsibilities has been to understand the skills, inclinations and limitations of my staff, and allocate them appropriate responsibilities. It’s a difficult task, and one I largely had to figure out for myself. Again, it seems likely that this functions sub-optimally.

I’d love some firmer numbers against this stuff to give us a clearer sense of how much we should care about it all. But at the very least it seems plausible to me that economics is underrating the gains from better allocation of workers to jobs that fit them, especially on unobservable personality characteristics.

1. Here’s the study, which I have found tricky to parse. It tries to estimate the benefit of lower barriers to women and Black men entering various professional occupations in the US since the 1960s. It does not seem to show that these directly raised output, but models the implied impact of those falling barriers assuming that capable women and Black men were able to find more suitable jobs.

2. Broadly speaking. Obviously things like risk appetite and time horizon matter, but my point is that financial capital is less varied and constrained than human capital.