The job prospects for blue-collar workers is one of the hottest topics in U.S. politics. The current fascination with the potential of robotics has helped reveal the important truth that productivity increases have been the biggest and perhaps the inexorable factor reducing the number of manufacturing jobs, and that’s increasingly true in service sectors as well. Most observers expect to continue.

So far, the options for what to do about the loss of good jobs are pretty bad. Changes in trade policy have lots of drawbacks, but even if they could be enacted, they won’t address the massive productivity-driven job losses in industries like manufacturing. Retraining workers who lose jobs sounds like a good idea if there were lots of good jobs available for which they could be retrained. But that has not been the case, and it is unlikely to be so in the foreseeable future. What can we do to help increase the number of good jobs in the United States?

One limitation of the current debate is that it ignores the choices individual employers make about how to get work done and how to improve productivity, something employers have to do. But substituting machines and software for people is not the only way to do that. And it is not necessarily even the most effective way.

Why have businesses been so inclined to go in that direction? Because we’ve stacked the deck with policy decisions that favor buying equipment over investing in employees and management.

A strong case could be made that the most important factors driving productivity increases have had to do with management practices. Scientific management and assembly line techniques created 20th-century manufacturing. The biggest innovations in contemporary manufacturing have not come from technology but from management practices, first associated with quality improvement programs and more recently with the practices of lean production created by Toyota in the auto industry.

As my colleague John Paul MacDuffie and others have documented, while U.S. auto companies were trying to compete by replacing workers with robots, the Japanese companies beat the pants off them on productivity and quality with better systems for managing their employees.
Nor is this unique to manufacturing. My colleague Marshall Fisher demonstrated that the key source of competitive advantage in retail is better trained and managed front-line workers who are more able to make the last and most important step in the supply chain work by keeping the shelves stocked.

The important thing about these management practices is that they center on workers, giving them new skills and roles. Yes, machines are getting better all the time, but they still have limitations. They are expensive, they are less flexible than workers, and even at their best, they are only as good as the people programming them.

So why aren’t we seeing more of a focus on upgrading management and employee skills rather than replacing workers with machines?
For-profit businesses, especially public companies that have extensive reporting requirements for investors, hate fixed costs or those that can’t be adjusted down if business falls. That’s a reason we typically hear for why CFO’s in particular hate adding workers, because they represent fixed costs.

Robots and technology are massively bigger fixed costs than workers. We can’t layoff robots or reduce their hours if business goes down as we can with employees. You might think that would scare investors and businesses away from that approach. What we can do with robots and tech under contemporary accounting rules and can’t do with investments in employees or management is amortize investments in robots and depreciate them over time, spreading the costs out. There are also a range of special tax breaks for investments in capital that aren’t there for management and employee spending. Tax laws and accounting principles do not recognize training and management interventions as investments even though they surely are by any other definition.

Another way in which accounting stacks the deck against investments in human capital and management systems has to do with how those investments are reported, or not reported. An investor looking at any company can tell pretty quickly how much it has invested in capital improvements because the figure is right there on the balance sheet along with related investments like leased capital and even fuzzy assets like “good will.” Those investments are seen as a good thing for the business and are counted as “assets” held against “liabilities.”
But if I wanted to see investments in training and in management, where would I find those? You can’t.

They are lumped in with “general and administrative expenses,” typically described as “overhead” costs. These are liabilities on balance sheets, and investors like to see them as low as possible. Businesses that spend a lot on training and management improvements can look to investors like they are blowing money on office furniture. So we have the irony that business is rewarded for investments in capital that raise productivity by eliminating jobs but punished for investments in people and management that raise productivity and save jobs.

The Federal Government has been also spent billions directly and through agencies like the Department of Defense to develop robots and other manufacturing technology that displaces workers. The Centers for Advanced Manufacturing spread the knowledge of how to use them to businesses. There is little attention given to how workers might fit into these new systems.

Given all this, it’s not surprising that businesses favor spending to replace workers with capital equipment.

These ideas aren’t new, and the beauty about them from a political perspective is that there have not been groups organized to oppose them. Changing the tax code and accounting principles to unstack the deck against investments in employees is far easier and more likely to succeed than any of the other policies under debate.

It does require more effort for companies to invest in employees and change their management practices to make use of those employees than to buy equipment that vendors will set up and get going.

But the management challenges of investing in people and systems are also its saving grace. Any competitor can call up a vendor and get the latest robot, so it can’t be a source of competitive advantage for long, if at all. Interventions that are challenging to pull off like better management of the workforce can be. That makes them work in the long term for business and their employees.

Cappelli is a management professor at the University of Pennsylvania’s Wharton School.

Read more: