Labor Regulations Attached to Proposition HHH Lowered Los Angeles Housing Production Through Increased Costs, Developer Avoidance

A project labor agreement requiring that a primarily union workforce be used for housing projects for the homeless in Los Angeles appears to be one reason that the city's $1.2 billion Proposition HHH ballot initiative is falling short of its goal, according to a new RAND Corporation report.

If the rules—requiring that housing developments of 65 housing units or more be built using a primarily union workforce—had not been in place, the report estimates that about 800 additional units of housing could have been funded, in addition to the 7,305 units now in the pipeline. The project labor agreement was adopted by the city council after voters approved the historic ballot measure, which was intended to support construction of up to 10,000 units.

The report points toward two primary mechanisms driving this estimate. First, affordable housing developers responded to the project labor agreement by disproportionately proposing smaller housing projects that would not be covered by the agreement. Second, for larger projects affected by the agreement, construction costs increased by about $43,000 per unit, amounting to a 14.5% hike in construction costs for around 27% of the funded projects.

“Developer responses and higher construction costs related to the project labor agreement appear to be important contributors to Proposition HHH falling short of its announced housing goals,” said Jason M. Ward, author of the report and an associate economist at RAND, a nonprofit, nonpartisan research organization.

“Moving forward, considering the effects of these sort of labor requirements on the primary goals of public housing policy and making sure they are clearly communicated to the public would improve transparency and help ensure that realistic goals are set for such programs,” Ward said.

In 2016, voters in the city of Los Angeles passed Proposition HHH, approving the sale of $1.2 billion in bonds to support the construction of permanent supportive housing for people experiencing homelessness.

Campaign proponents suggested the funding would support the creation of up to 10,000 housing units. At present, virtually all funding has been committed and a total of 7,305 units of housing are in the pipeline.

The failure to meet the original target has been attributed in part to significantly higher-than-expected construction costs, which have averaged around $560,000 per unit, about 40% more than estimated during the campaign.

The RAND report analyzes the effects of a project labor agreement that was approved by city leaders about 18 months after the bond measure was approved.

The labor agreement includes a variety of rules concerning hiring authority, worker ratios (both union and nonunion workers), clauses guaranteeing no strikes or lockouts, and the hiring of certain levels of city residents and people from disadvantaged communities. The provisions apply only to projects that included 65 or more units.

The RAND report used two methods to assess the effects of the project labor agreement.

The first estimated how the labor agreement affected the size of proposed housing projects funded through HHH by comparing the difference in the shares of HHH-funded projects above and below the 65-unit threshold with the difference in these same shares among a sample of similar non-HHH-funded projects.

The second assessed the effect of the labor agreement on construction costs by comparing the size of cost differences at the 65-unit threshold between the two samples.

Ward found that the evidence strongly suggests that developers responded to the labor agreement by disproportionately building housing projects that fall below the 65-unit threshold. In total, 22 of the 98 construction projects analyzed are in the narrow range of 60 to 64 units. In contrast, between 65 and 69 units, there was just one proposed project.

While projects with 50–64 units make up more than 45% of the HHH sample, such projects make up less than 10% of the non-HHH sample. The overall share of housing projects with more than 65 units was much higher among projects not funded by HHH bond money.

Ward said it is unclear why developers responded so strongly to the presence of the labor agreement. The profit motive that might be relevant to market-rate housing developers is likely not a strong factor among the groups building HHH projects, which are primarily composed of nonprofit, mission-driven organizations that generally are allowed to charge a capped developer fee related directly to project costs.

“Developer concerns about the labor agreement adding uncertainty over costs and timelines may have been an important factor,” Ward said. “Deeply subsidized affordable housing projects already face considerable uncertainty related to community opposition, assembling the necessary funding, and uncertain timelines for regulatory approvals.”

Support for the report, “The Effects of Project Labor Agreements on the Production of Affordable Housing: Evidence from Proposition HHH,” was provided by the Lowy Family Group through its funding of the RAND Center on Housing and Homelessness in Los Angeles.

The RAND Social and Economic Well-Being division seeks to actively improve the health, and social and economic well-being of populations and communities throughout the world.