The National Council of State Housing Agencies (NCSHA) has released a comprehensive nationwide report on factors affecting the cost of developing low-income housing tax credit (LIHTC) properties.

The independent study was prepared by Abt Associates for the NCHSA with robust data. Fourteen syndicators provided data on 2,547 LIHTC developments containing 162,447 units placed in service between 2011 and 2016. Approximately 97% of the units are affordable while the rest of the units have market-rate rents in mixed-income communities. The sample includes approximately 47% of the units developed with 9% housing tax credits and 20% of the units developed with 4% housing tax credits placed in service between 2011 and 2016. It also factors in geographic diversity, including at least two developments in each state, the District of Columbia, Puerto Rico, and the Virgin Islands and over 25 developments in 35 states.

Stockton Williams
Stockton Williams

According to Abt Associates, the median total development cost (TDC) per unit, inclusive of soft costs and land costs, for the over 160,000 units was $164,757, adjusted for construction cost inflation. In the findings, 75% of the units had per-unit TDC at or less than $224,903 and 25% had per-unit TDC at or less than $121,254.

NCSHA looked to other research to compare the overall LIHTC development costs with the nation’s overall apartment development costs, finding that an average TDC for newly constructed multifamily apartments was roughly $196,000 to $204,000 between 2011 and 2016. Abt’s findings showed that the average LIHTC cost per unit of new construction, including land and soft costs, was slightly higher at about $209,000 during this timeframe. The higher cost per unit could be attributed to financing requirements not applying to market-rate developments.

According to the NCSHA, many of the factors that drive the development cost for multifamily projects, including those financed with LIHTCs, are beyond the control of state housing finance agencies (HFAs). This includes the costs of land, labor, materials, and local regulations. The Council, which represents the state agencies that allocate the LIHTC program, says the study suggests that state HFAs’ efforts to implement policies to require and urge developers to create LIHTC housing in an cost-effective manner has been effective.

“Since 1986, the housing credit has been our most successful tool for producing and preserving affordable housing—creating more than 3 million homes. This study is the clearest evidence yet that oversight of development costs by state housing agencies is not only working, but has been essential to making the most of this vital resource,” says Stockton Williams, NCSHA executive director.

Some of the key findings from Abt Associates include:

  • Median per-unit TDCs were found to be higher for developments in principal cities of metro areas, difficult development areas, and qualified census tracts.
  • Geographically, costs also were found to be higher in the Mid-Atlantic, New England, and Pacific regions compared with the rest of the country. The higher cost of land or construction wages, which were controlled for state-level differences in the study, were not the sole factors behind the findings. Abt says one explanation may be that developers adjust to higher land costs by employing different construction methods that carry a higher cost. Median per-unit TDCs were found to be the lowest in the Southern regions.
  • New construction developments were substantially more expensive than acquisition-rehab developments.
  • Developments with multiple financing sources were more expensive per unit, which, according to Abt, could be associated with the challenges of assembling multiple sources for the capital stack or the need to find multiple sources to pay for higher costs.
  • Projects developed with 9% LIHTCs tended to have higher costs than those with 4% LIHTCs, although this was not true for all models examined.
  • Costs also varied by average bedroom size, the study found. Projects where the average unit size was more than 2.5 bedrooms were more expensive.
  • Per-unit TDCs also decreased as the number of the units in the development increased, suggesting the benefit of economies of scale. Units located within developments with 100 units or less had median costs above the median for the entire database, while units located in developments with 100 units or more had median costs below the median.
  • Projects developed by nonprofits had higher costs than for-profits, although this was not the case in all of Abt’s models. Abt says this finding has been cited in other studies, saying potential reasons may be that projects developed by nonprofits may provide more supportive services and/or nonprofits may be more willing to take on developments with higher land costs, significant neighborhood opposition, or the need for substantial zoning changes.

It’s important to note that because the projects in the data were provided by participating syndicators, instead of selected at random, the sample is not statistically representative of all LIHTC developments. In addition, some factors likely to be associated with differences in costs could not be examined because of insufficient data, including a long development timeline related to local approval or NIMBYism, a tight labor market leading to higher wages, certain construction types, donated or below-market land acquisition, or a location in a master-planned development where approvals have been already granted.