Buildings and Workers in Los Angeles Residential Construction

Place: Los Angeles • Date: August, 2026 • Partner: Office of the Chief Legislative Analyst, City of Los Angeles

The figures on this page are live: hover any chart for the underlying figures.

The City of Los Angeles is considering a Residential Construction Minimum Wage: a floor of $32.35 an hour plus a $7.65 healthcare credit, applied to residential developments of ten or more units and under eighty-five feet. A wage floor of that kind acts at the nexus of two things that are usually measured separately - the city's decisions about which buildings to authorize, and the economic forces affecting the people who build those buildings. This assessment puts the two side by side.

Part one tabulates thirty-one years of residential building permits. Part two evaluates twelve years of construction employment and wages among City workers. Both parts are descriptive: every quantity below is a measurement or an arithmetic consequence of one. Nothing here is a causal estimate, and the elasticities that would be needed for one are the subject of a separate technical assessment.

Part One: What Los Angeles Has Authorised

Los Angeles has permitted residential construction across two very dynamic growth cycles since 1995. The floor of the series is not the one usually cited: the City bottomed at 1,993 units in 1996, below the Great Recession trough of 2,607 in 2009. Any worst case anchored on the financial crisis is actually referencing the second-worst year of this generation.

Figure 1. Permitted units in the City of Los Angeles by structure size, 1995 to 2025, against California real GDP in billions of chained 2017 dollars on the right-hand axis. Permits are authorised units, not starts or completions. The GDP line begins in 1997, where the state accounts begin, rather than with the permits in 1995. The two scales are independent and are set to fit their own series, so the visual relationship between them is a matter of scaling and should not be read as a measured one. Hover for the year's figures.

The County peaked 22 years ago and has never matched its 27,429 units in 2004; 2022, its best recent year at 26,383, still fell about four per cent short despite eighteen years of sustained GDP growth.

Figure 2. The same measure for Los Angeles County, which includes the City.

Composition of the sector by building occupancy has shifted more dynamically than growth in units. In 2011 the average new building in the five-plus bracket jumped from 27.7 to 61.2 units in a single year - fewer buildings (117 to 84) delivering more units (3,237 to 5,139). At this time the City stopped permitting mid-sized apartment buildings and started permitting large podium projects and it has not gone back.

Figure 3. The same data as a share of each year's total. Buildings of five or more units rose from 61 per cent of City permits in 1995 to 87 per cent in 2011, then fell back to 59 per cent by 2025.

Five salient features of the period:

The real floor is 1996, not 2009. 1,993 units against 2,607.
2011: project scale steps up and stays up. 28.9 units per five-plus building before, 48.1 after.
Multifamily peaked in 2015 and has fallen for a decade. 13,671 units then, 6,599 in 2025 - a fall of 52 per cent. The total peaked later, in 2022, only because single-family and duplexes were rising.
2022 to 2024 is the sharpest non-recession fall on record. 16,707 units to 10,488, 37 per cent, and almost entirely the five-plus bracket.
The missing middle is gone; duplexes replaced it. Permits for three- and four-unit buildings have been exactly zero every year since 2017, while two-unit permits more than doubled at that moment - the timing of California's accessory-dwelling-unit reforms.

Part Two: Who Builds It, and What the Work Pays

There is no published City-of-Los-Angeles occupational wage series: the establishment survey stops at the metropolitan area and the payroll census at the county. This part of the assessment is therefore built from household records for residents of the City's own PUMAs. It is residence-based - a City resident framing houses in Pomona is counted, a Pomona resident working downtown is not - and it is a survey estimate rather than a payroll count.

Real pay has moved modestly in the middle and substantially at the bottom. Between 2013 and 2024, in 2024 dollars, the median rose 12 per cent while the tenth percentile rose 37 per cent and the twenty-fifth 27 per cent. The distribution compressed from below, and the statutory floor is what compressed it: by 2016 the minimum wage had passed the twenty-fifth percentile of measured construction pay outright.

Figure 4. Hourly wages in 2024 dollars. The dashed rule is the City minimum wage in force that year, also in 2024 dollars. 2020 has no American Community Survey release, so every line breaks there rather than interpolating.

Employment turned down before the buildings did. City residents working residential trades peaked at 82,088 in 2019 - three years before permits peaked - and stood at 70,947 in 2024. The structure and exterior trades, the group tied most closely to the multifamily pipeline, carry the whole of that decline.

Figure 5. Weighted counts of City residents working in each residential trade group.

Disaggregated by trade, the compression is concentrated where it would be expected - in the occupations that sit near the floor. The twenty-fifth percentile rose by roughly a quarter to a half for laborers, carpenters, painters and roofers, while the licensed trades barely moved at either percentile. The skilled premium narrowed from both ends.

Figure 6. Each pair of dots is one occupation's wage in the early period and the late period, in 2024 dollars. Occupations appear only where both periods clear thirty sample records.

One finding bears on the policy more directly than any other. A large minority of this workforce reports pay below the wage floor already in force, and the share has grown as the floor has risen: 21 per cent in 2013, 37 per cent in 2018, 30 per cent in 2024.

Figure 7. Share of workers whose implied hourly wage - annual wage income divided by weeks times usual hours - falls below the minimum wage in force that year. The American Community Survey is not a compliance instrument, so read the level as an order of magnitude and the trend as the finding.

Another handful of salient insights

Employment peaked in 2019, three years before permits did — 82,088 → 70,947 (−13.6%). Labour turned first.
The distribution compressed from below. Real p10 +37%, p25 +27%, median +12%, p75 +16%. The p75/p10 ratio fell 3.58 → 3.04. OEWS confirms the shape independently on establishment records: p10 +27.5% against a median that moved 2.8%.
The licensed trades stopped pulling away — electricians +2.1%, plumbers −2.3%, against laborers +13.9% and carpenters +21.1%.
Non-citizen labour supply and enforcement. 54% of these workers are non-citizens, and 68% in the structure and exterior trades. This project’s own estimation finds enforcement intensity contracts supply with a coefficient of −0.10 (SE 0.03) — the most precisely estimated parameter in the system.
A third report pay below the legal floor, and the share rose with the floor — 21% (2013) → 37% (2018) → 30% (2024). The 25th percentile went from 1.12× the minimum wage to 0.93×.

Three reasons labor supply needs to be better understood:

Labour leads buildings, not the other way round. Employment peaked in 2019 and permits in 2022.
A wage floor is already binding here, and already not fully reaching. The general City minimum reached $17.28 in 2024 and now sits above the twenty-fifth percentile of measured construction pay, with roughly a third of workers reporting pay below it. A second, higher floor inherits that compliance problem rather than starting clean of it.
The covered pool is the shrinking one. The five-plus bracket is both where the ten-unit threshold bites and where permitting has fallen by half since 2015.

Notes and Sources

Permits are authorised units: they lead construction activity, overstate it where projects lapse, and are not employment. Wages are deflated to 2024 dollars with the Los Angeles-Long Beach-Anaheim consumer price index. Occupations are harmonised across the 2018 Census occupation code revision by code pair rather than by code alone, without which painters and plumbers lose half their years. Employment counts are weighted survey estimates, and occupational detail is pooled over multi-year periods because annual cells run to a few dozen records.

Sources: All numbers presented here are BEAR estimates based on the following public information: US Census Bureau, Building Permits Survey and American Community Survey 1-year Public Use Microdata Sample; US Bureau of Labor Statistics, Occupational Employment and Wage Statistics and Consumer Price Index; City of Los Angeles Minimum Wage Ordinance; California Policy Lab on the ordinance's effects; Economic Roundtable on informal employment in California construction; and RAND and the UCLA Lewis Center on Measure ULA.

"Questions? Contact David Wells Roland Holst, PhD | admin@bearecon.com | www.bearecon.com