Math will kill Trump’s infrastructure plan
In his address to Congress on Tuesday, President Donald Trump once again brought up his support for a large infrastructure package. And there's good reason for this: It's not nearly as polarizing as most other parts of his agenda and would stimulate economic growth in a way that would benefit blue-collar workers who were key to his election. But like much of his agenda, it's short on details, and the labor-market math doesn't add up.
Here's the napkin version. The trillion-dollar package being discussed is understood to be $100 billion of spending per year for 10 years. Leave aside the fact that infrastructure spending is notoriously messy and slow, as environmental delays and other project-specific concerns make it hard to spend the money as fast as a policymaker or economist would like. The labor question alone shows that this vision is impossible.
There are currently 6.8 million construction employees in the U.S. Annualized construction spending in the U.S. at the end of 2016 was $1.18 trillion. Dividing the two, we see that one construction worker supports around $175,000 in construction spending. (This doesn't mean that construction workers make $175,000 per year – that figure accounts for other labor-supporting projects and building materials.)
One more simple calculation shows the daunting labor needs. If one construction worker can support $175,000 worth of construction projects, then $100 billion in spending each year would require an additional 570,000 construction workers, which doesn't take into account truck drivers, project managers, environmental specialists, and all other support staff needed to complete projects. Perhaps infrastructure spending, which comprises 25 percent of all construction spending, is a little less labor- intensive than other types of construction spending. Maybe the shrewd administrative talent of this White House could generate some labor efficiencies. That still probably means 400,000 or 500,000 construction workers needed, not 50,000.
How realistic is construction employment growth of 570,000 workers? It hasn't happened since 1946. Even the peak of the housing bubble generated only one brief year-over-year increase of 500,000 construction workers.
A period of strong growth for construction employment is between 200,000 to 300,000 workers per year, like we saw in 2013 to 2015. As we get later in this economic cycle, construction employment growth is starting to slow – growth over the past 12 months was 170,000 jobs.
Also, infrastructure spending is only 25 percent of total construction spending. Residential construction continues to grow as housing recovers from its bust and millennials age into their family-forming years. Residential construction represents 40 percent of total construction spending and has added between 100,000 to 125,000 construction jobs over the past few years. So outside of residential construction, we're currently adding around 50,000 construction jobs per year. And residential construction growth, and hence its labor needs, should continue for sever- al more years, making it more difficult to find labor slack that could be redeployed into infrastructure projects.
The construction labor market at current wages is tight and has been tightening for the past several years. Last summer, when construction unemployment was at its seasonal low, there were only around 400,000 unemployed construction workers. This is around the lowest level we've seen for construction unemployment since the late 1990s. So if we're going to get an unprecedented amount of construction employment growth, they're going to have to come from other industries, outside the labor force, or abroad.
Immigrant labor, particularly undocumented workers, represent a significant proportion of the construction labor force. Bloomberg reported last week that up to 1.1 million construction workers in the U.S. are undocumented, so stepping up deportations would deplete an already-too-small construction labor pool.
Without any radical changes to immigration policy, we might have capacity for an additional 50,000 construction workers per year for infrastructure projects – far short of the 570,000 needed under Trump's infrastructure proposal. Significant growth beyond that would likely require much higher wages and poaching from other industries, creating labor shortages in those industries.
The infrastructure proposal is among Trump's most politically viable, but economics will kill it.