Research Note · Multifamily Absorption
How Fast New Apartments Lease: Census Survey of Market Absorption Data, 1970 to 2024
The Census Bureau's Survey of Market Absorption reports that 55 percent of new apartments completed in 2016 were rented within three months and 94 percent within twelve 1. Final quarterly estimates for 2016 to 2024 run from 41 to 75 percent 2; first estimates are revised down in 32 of 36 quarters 2; and the metropolitan tables ended with the 2018 releases 3.
Between 2016 and 2024, 41 to 75 percent of new apartments leased within three months.
The Survey of Market Absorption of New Multifamily Units (SOMA) is the Census Bureau's measure of how fast new apartments lease. The Department of Housing and Urban Development sponsors it, and it draws its sample from the Survey of Construction 4. For each completed building it records how many units were first rented or sold by 3, 6, 9 and 12 months after completion 2. The headline figure is the share of units absorbed within three months.
For apartments completed in 2016, the share absorbed was 55 percent at three months, 76 percent at six, 88 percent at nine and 94 percent at twelve 1. The annual three-month rate has run from 82 percent for 1978 and 1979 completions to 50 percent for 2008 completions, a range that follows the cycle and not a fixed standard 5. The final quarterly estimates for 2016 to 2024 completions run from 41 percent (fourth quarter of 2023) to 75 percent (third quarter of 2021), each with a margin of error of 2.2 to 4.1 points 2.
A first estimate is not the last word. In 32 of the 36 quarters from 2016 to 2024 the final three-month rate is below the initial one, by 2.2 points on average and 3.2 points for 2022 to 2024 2. The first-published rates for 2025 and 2026 completions (48, 47, 49 and 41 percent) are secondary: they are the first-published estimates as reported by the National Association of Home Builders, not final Census values 43444546.
SOMA is a market-rate benchmark. It covers only new buildings of five or more units and leaves subsidized and tax credit units out of the absorption rate 26. It counts first rentals, not stabilized occupancy, and it starts the clock when a building is half occupied or available 2. The method for turning this series into a lease-up forecast is in absorption and lease-up, part of the analytical framework. This note is the data reference behind that method.
SOMA counts first rentals in new buildings at fixed intervals after completion.
Universe and scope
SOMA covers new multifamily construction only, and the Census Bureau states the scope in one sentence: "All statistics from the SOMA refer to apartments in newly constructed buildings with five units or more." 7 Buildings completed in areas that do not issue building permits are outside the survey, and less than two percent of new construction takes place in those areas 82. The survey is designed to estimate the rate at which privately financed, nonsubsidized units are rented or sold 8.
The rate used in this note is the rate for unfurnished rental apartments. Furnished units, condominium and cooperative units, subsidized and tax credit units and a residual group (time-sharing, continuing-care retirement and turnkey units) are counted separately in the annual completions table 6. The glossary defines subsidized units as those where a government program pays part of the cost of construction, mortgage or operating expenses 9. The three-month rate is therefore a market-rate benchmark, and it does not describe a tax credit or other subsidized property.
When a building is complete
The clock does not start at the first lease. The Census Bureau classifies a building as complete under the Survey of Construction rule: "Construction of a multifamily building is classified as complete when 50 percent of the units are occupied or available for occupancy." 2 A building that began leasing earlier is interviewed three months after half of its units were available, and the published tables do not separate leasing before the completion date from leasing after it 10. A three-month rate measures the first three months after a building is half available, not three months after the first lease.
What absorbed means
The glossary counts absorption as a unit "initially rented or sold after construction and is no longer available on the market." 9 Absorbed means a first rental. A later turnover does not add to the count, and the count is not an occupancy rate. Rent is the monthly payment to the owner 9, and the annual tables note that data on asking rent are collected at the initial interview 11. The survey records an asking rent at month three, not an effective rent after concessions.
Interviews and respondents
The initial interview occurs three months after a building is completed 12. It "collects information on amenities, rent or sales price levels, number of units, type of building, and the number of units taken off the market (absorbed)." 4 The Bureau adds: "Field representatives conduct subsequent interviews, if necessary, at 6, 9, and 12 months after completion." 4 The preferred respondent is the building manager or the rental or sales agent 12, and participation is voluntary 13.
Weights, estimates and revisions
The statements for 2022, 2023 and 2024 completions each describe 48 self-representing and 772 non-self-representing primary sampling units 14152. In the 2024 statement, buildings with five or more units are selected with certainty and SOMA has not subsampled since 2005 2. The weights are then adjusted: "The basic weights are raked to two sets of SOC known totals: Regional and Metropolitan Statistical Area (MSA) status." 2 The weighting controls the national and regional totals and the split between metropolitan and other areas, and no finer geography.
The Bureau produces quarterly estimates, estimates for the prior four quarters and annual estimates, and the annual estimates use the final figure from each quarter 2. Each quarter it publishes the three-month rate for the latest completion quarter and the six, nine and twelve month rates for the three quarters before it, with revisions to the earlier figures. The fixing rule is short: "After the fourth quarter, the estimates are no longer revised." 2 A three-month rate is therefore published four times, as the initial, first revised, second revised and final estimate, and only the last is permanent.
Release calendar and where the numbers live
The static reports ended in 2018. On September 27, 2018 the absorption data moved to the Table Creator, and the Bureau states: "The historical format of the quarterly, annual, and characteristics reports, tables, and figures will no longer be produced." 4 The Table Creator carries completions from 2016 16, groups rent categories before and after 2019 17 and suppresses an estimate whose coefficient of variation is 30 percent or higher 18. The last static annual tables are the 2017 report for 2016 completions 19, and the last static quarterly tables are the 2018 releases 20.
The release page, last revised August 17, 2026, lists Third Quarter 2026 Absorptions (Second Quarter 2026 Completions) for November 2026, the 2025 Completions Annual Tables for February 2027 and First Quarter 2027 Absorptions (Fourth Quarter 2026 Completions) for May 2027 21. A release is named for the quarter in which the three-month absorption is measured, the quarter after completion, and appears in the quarter after that, so a reader matches the completion quarter, not the release name. The release before the one scheduled for November 2026 was the August 2026 release for first quarter 2026 completions, as reported by the National Association of Home Builders 2146. The 2025 annual tables are not yet published.
The annual three-month rate fell from 82 percent in 1978 to 50 percent in 2008.
The long series is the share of unfurnished rental apartments completed in a calendar year that were absorbed within three months of completion. Table 1 sets it beside the number of unfurnished rental apartments in scope, for 1970 to 2016. The values come from the 1995 report for 1971 to 1974, the long-series table of the 2015 report for 1970 and 1975 to 2014, and the annual tables for 2015 and 2016 completions 2252324. For the national rate and completions, the long-series table and the annual reports agree: the 1999 report gives 72 percent for 1998 completions and the 2010 report gives 51 percent for 2009 completions 2526. The one national revision is a 100-unit difference in 2005 completions, 113,700 in the 2015 report and 113,800 in the 2006 report 2728.
| Completion year | 3-month rate (%) | Units completed in scope | Completion year | 3-month rate (%) | Units completed in scope |
|---|---|---|---|---|---|
| 1970 | 73 | 328,400 | 1994 | 80 | 104,000 |
| 1971 | 68 | 334,400 | 1995 | 73 | 155,000 |
| 1972 | 68 | 497,900 | 1996 | 72 | 191,300 |
| 1973 | 70 | 531,700 | 1997 | 73 | 189,200 |
| 1974 | 68 | 405,500 | 1998 | 72 | 209,900 |
| 1975 | 70 | 223,100 | 1999 | 72 | 225,900 |
| 1976 | 80 | 157,000 | 2000 | 72 | 226,200 |
| 1977 | 80 | 195,600 | 2001 | 64 | 193,100 |
| 1978 | 82 | 228,700 | 2002 | 59 | 204,100 |
| 1979 | 82 | 241,200 | 2003 | 61 | 166,500 |
| 1980 | 75 | 196,100 | 2004 | 62 | 153,800 |
| 1981 | 80 | 135,400 | 2005 | 64 | 113,800 |
| 1982 | 72 | 117,000 | 2006 | 58 | 116,400 |
| 1983 | 69 | 191,500 | 2007 | 55 | 104,800 |
| 1984 | 67 | 313,200 | 2008 | 50 | 146,400 |
| 1985 | 65 | 364,500 | 2009 | 51 | 163,000 |
| 1986 | 66 | 407,600 | 2010 | 61 | 89,100 |
| 1987 | 63 | 345,600 | 2011 | 59 | 74,700 |
| 1988 | 66 | 284,500 | 2012 | 63 | 104,500 |
| 1989 | 70 | 246,400 | 2013 | 63 | 134,200 |
| 1990 | 67 | 214,300 | 2014 | 61 | 210,800 |
| 1991 | 70 | 165,300 | 2015 | 60 | 261,800 |
| 1992 | 74 | 110,200 | 2016 | 55 | 262,500 |
| 1993 | 75 | 77,200 |
Source: U.S. Census Bureau, Survey of Market Absorption annual absorption reports (1995, 2015, 2016 and 2017 reports; 2017 report Table 9), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 22523246.
Reading the cycle
Table 2 groups the years into eras. The boundaries are a reading choice, and the means are simple averages, not completion-weighted rates. The pattern is a series of level shifts: 68 to 73 percent in 1970 to 1975, 72 to 82 in 1976 to 1982, 63 to 70 in 1983 to 1991 and 74 to 80 in 1992 to 1994. The six years from 1995 to 2000 sit at 72 to 73 percent. After 2000 no year reached 65 percent; the highest values are 64 percent in 2001 and 2005 5.
| Completion years | Annual 3-month rate, range (%) | Simple mean of annual rates (%) | Units completed in scope per year, range |
|---|---|---|---|
| 1970 to 1975 | 68 to 73 | 69.5 | 223,100 to 531,700 |
| 1976 to 1982 | 72 to 82 | 78.7 | 117,000 to 241,200 |
| 1983 to 1991 | 63 to 70 | 67.0 | 165,300 to 407,600 |
| 1992 to 1994 | 74 to 80 | 76.3 | 77,200 to 110,200 |
| 1995 to 2000 | 72 to 73 | 72.3 | 155,000 to 226,200 |
| 2001 to 2007 | 55 to 64 | 60.4 | 104,800 to 204,100 |
| 2008 to 2009 | 50 to 51 | 50.5 | 146,400 to 163,000 |
| 2010 to 2016 | 55 to 63 | 60.3 | 74,700 to 262,500 |
| 1970 to 2016 | 50 to 82 | 67.7 | 74,700 to 531,700 |
Source: U.S. Census Bureau, Survey of Market Absorption annual absorption reports (1995, 2015, 2016 and 2017 reports; 2017 report Table 9), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 22523246.
The low point is 2008 and 2009: 50 percent for 2008 completions and 51 percent for 2009 completions, the latter on 163,000 unfurnished completions 26. The rate recovered to 61 percent for 2010 completions and held at 59 to 63 percent through 2015. The 55 percent for 2016 completions is the lowest annual rate since 2009 24.
The series does not show a simple relationship between volume and rate. Across the 47 years the correlation between the number of unfurnished completions and the three-month rate is 0.06 (0.26 for 1995 to 2016), on the published rounded values. The five years with the most completions (1973, 1972, 1986, 1974 and 1985) had rates of 70, 68, 66, 68 and 65 percent. The five with the fewest (2011, 1993, 2010, 1994 and 2012) had rates of 59, 75, 61, 80 and 63 percent 56. The recent run points the other way: 210,800, 261,800 and 262,500 completions in 2014, 2015 and 2016, with rates of 61, 60 and 55 percent. The series has too few turning points to support a causal estimate. Whether new supply is being absorbed is a question for the pipeline analysis in supply and competition, built from the permitted, financed and under-construction units around the subject, and not a reading of this national series.
The mix of buildings also changes the meaning of the rate. Unfurnished rental units in scope were 84 percent of apartments finished in structures of five or more units during 2015 and 2016, and 41 percent in 1981, 1982, 2006 and 2007. In 2006 condominium and cooperative units were 104,600 of 283,500 completions (37 percent); in 2016 they were 20,800 of 311,000 (7 percent), and subsidized and tax credit units were 20,200 6. The survey's count of all apartments finished in structures of five or more units during 2016, 311,000, equals the New Residential Construction count 296.
Two series breaks matter. The user note says weighting adjustments changed in 1990 and 1992 and advises caution in comparing completions in 1992 and later years with earlier years, and that interviewing software began in April 2014 and the asking rent ranges were revised at the same time 30. A comparison across 1992 or 2014 carries that caution.
Quarterly final estimates for 2016 to 2024 run from 41 to 75 percent, with margins of 2 to 4 points.
Appendix B1 of the 2024 completions statement on source and accuracy gives the three-month rate for every completion quarter from 2016 to 2024, from the initial estimate to the final one, with the coefficient of variation of each 2. Table 3 reproduces the initial and final rates and the final coefficient of variation. The margin of error is 1.645 times the standard error, which is the coefficient of variation times the final estimate; the Bureau states that "The 90 percent confidence level is the Census Bureau standard for measuring statistical significance." 2 The final coefficients of variation run from 2.1 to 4.2 percent, which gives margins of 2.2 to 4.1 points, 2.9 on average.
| Completion quarter | Initial estimate (%) | Final estimate (%) | Final minus initial (points) | Final coefficient of variation (%) | Final margin of error (points) |
|---|---|---|---|---|---|
| 2016Q1 | 58 | 55 | -3 | 3.0 | 2.7 |
| 2016Q2 | 58 | 57 | -1 | 3.2 | 3.0 |
| 2016Q3 | 61 | 59 | -2 | 2.5 | 2.4 |
| 2016Q4 | 48 | 47 | -1 | 3.1 | 2.4 |
| 2017Q1 | 55 | 54 | -1 | 3.1 | 2.8 |
| 2017Q2 | 62 | 58 | -4 | 2.8 | 2.7 |
| 2017Q3 | 58 | 56 | -2 | 3.5 | 3.2 |
| 2017Q4 | 50 | 48 | -2 | 3.6 | 2.8 |
| 2018Q1 | 55 | 53 | -2 | 3.8 | 3.3 |
| 2018Q2 | 61 | 57 | -4 | 4.0 | 3.8 |
| 2018Q3 | 60 | 56 | -4 | 3.5 | 3.2 |
| 2018Q4 | 52 | 50 | -2 | 3.0 | 2.5 |
| 2019Q1 | 55 | 53 | -2 | 3.7 | 3.2 |
| 2019Q2 | 58 | 55 | -3 | 3.7 | 3.3 |
| 2019Q3 | 58 | 56 | -2 | 2.6 | 2.4 |
| 2019Q4 | 54 | 55 | +1 | 2.6 | 2.4 |
| 2020Q1 | 45 | 43 | -2 | 3.8 | 2.7 |
| 2020Q2 | 60 | 57 | -3 | 3.7 | 3.5 |
| 2020Q3 | 51 | 49 | -2 | 4.2 | 3.4 |
| 2020Q4 | 50 | 51 | +1 | 3.3 | 2.8 |
| 2021Q1 | 58 | 57 | -1 | 3.5 | 3.3 |
| 2021Q2 | 72 | 71 | -1 | 3.5 | 4.1 |
| 2021Q3 | 76 | 75 | -1 | 2.1 | 2.6 |
| 2021Q4 | 67 | 68 | +1 | 3.1 | 3.5 |
| 2022Q1 | 62 | 63 | +1 | 2.4 | 2.5 |
| 2022Q2 | 77 | 72 | -5 | 2.8 | 3.3 |
| 2022Q3 | 61 | 59 | -2 | 3.3 | 3.2 |
| 2022Q4 | 59 | 56 | -3 | 2.6 | 2.4 |
| 2023Q1 | 61 | 57 | -4 | 4.2 | 3.9 |
| 2023Q2 | 65 | 61 | -4 | 3.0 | 3.0 |
| 2023Q3 | 57 | 51 | -6 | 2.6 | 2.2 |
| 2023Q4 | 42 | 41 | -1 | 4.0 | 2.7 |
| 2024Q1 | 53 | 49 | -4 | 2.9 | 2.3 |
| 2024Q2 | 55 | 50 | -5 | 3.8 | 3.1 |
| 2024Q3 | 49 | 47 | -2 | 3.3 | 2.6 |
| 2024Q4 | 45 | 42 | -3 | 3.8 | 2.6 |
Source: Census Bureau SOMA statement on 2024 completions, Appendix B1, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables; margin of error is 1.645 times the coefficient of variation times the final estimate 2.
The shape of the finals
The finals sit between 47 and 59 percent from 2016 through 2019, with a one-quarter fall to 43 percent for first quarter 2020 completions. Three quarters are above 70 percent: 71 for second quarter 2021, 75 for third quarter 2021 and 72 for second quarter 2022. The rate then falls to 41 percent for fourth quarter 2023 completions and stays between 42 and 50 percent for the four quarters of 2024. Five quarters read 47 or lower: fourth quarter 2016 (47), first quarter 2020 (43), fourth quarter 2023 (41) and third and fourth quarters 2024 (47 and 42) 2.
Table 4 averages the four final rates in each year. The appendix carries rates and coefficients of variation but no completion counts, so these are simple means and not the completion-weighted annual rates the Bureau publishes. The simple mean for 2016 is 54.5 against the published 55 224.
| Completion year | Mean of final quarterly rates (%) | Mean of initial quarterly rates (%) | Lowest final quarter (%) | Highest final quarter (%) |
|---|---|---|---|---|
| 2016 | 54.5 | 56.2 | 47 | 59 |
| 2017 | 54.0 | 56.2 | 48 | 58 |
| 2018 | 54.0 | 57.0 | 50 | 57 |
| 2019 | 54.8 | 56.2 | 53 | 56 |
| 2020 | 50.0 | 51.5 | 43 | 57 |
| 2021 | 67.8 | 68.2 | 57 | 75 |
| 2022 | 62.5 | 64.8 | 56 | 72 |
| 2023 | 52.5 | 56.2 | 41 | 61 |
| 2024 | 47.0 | 50.5 | 42 | 50 |
Source: Census Bureau SOMA statement on 2024 completions, Appendix B1, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables; the means are not Census annual rates; exact halves are rounded to the even digit 2.
A seasonal pattern the annual mean hides
Averaging the nine final rates for each quarter of the year from 2016 to 2024 gives 53.8 percent for the first quarter, 59.8 for the second, 56.4 for the third and 50.9 for the fourth. The quarterly tables published to 2018 show the same direction: from 2012 to 2017 the not seasonally adjusted rate for fourth-quarter completions averaged 4.3 points below the seasonally adjusted rate, and the rate for second-quarter completions averaged 4.7 points above it 10. A building that completes in the fourth quarter should be compared with fourth-quarter results, not with an annual mean that is several points higher.
The revision pattern
For the 36 quarters in Table 3 the final rate is below the initial rate in 32, above it in 4 (by one point each, for completion quarters 2019Q4, 2020Q4, 2021Q4 and 2022Q1) and equal in none. The mean change is minus 2.2 points: minus 1.75 for the 24 quarters of 2016 to 2021 and minus 3.2 for the 12 quarters of 2022 to 2024, 11 of which were revised down. The largest revisions are minus 6 points (57 to 51, third quarter 2023) and minus 5 points (77 to 72 for second quarter 2022 and 55 to 50 for second quarter 2024). In 12 of the 36 quarters the revision is larger than the margin of error of the final estimate 2. First quarter 2024 completions were 53 percent absorbed in the initial estimate and 49 in the final one, and fourth quarter 2024 completions were 45 and 42 2.
Most of the correction comes early. Table 5 splits the average change by stage: 62.5 percent of it arrives in the first revision, one quarter after the initial figure 2. The practical rule is to cite the final estimate wherever one exists and to write "initial" beside any first-published figure.
| Revision stage | Mean change (points) | Share of the initial-to-final change (%) |
|---|---|---|
| Initial to first revised | -1.39 | 62.5 |
| First revised to second revised | -0.47 | 21.3 |
| Second revised to final | -0.36 | 16.3 |
| Initial to final | -2.22 | 100 |
Source: Census Bureau SOMA statement on 2024 completions, Appendix B1, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables; exact halves in the share column are rounded up 2.
First-published rates for 2025 and 2026 completions run from 41 to 49 percent, and they are not final.
The final Census values in this note end with fourth quarter 2024 completions. For later completions this section reports the first-published estimates as reported by the National Association of Home Builders, which writes a short note on each release. These values are secondary: a trade association's reading of a Census release, and first estimates. Table 6 sets them beside the Census initial and final values wherever those are in this note, so that the reader can see how the first-published figure has related to the final one. For 2025 and 2026 completions the table gives no Census value.
| Completion quarter | Reported by NAHB, first published (%) | Census initial (%) | Census final (%) | NAHB minus Census final (points) | NAHB note |
|---|---|---|---|---|---|
| 2021Q3 | 75 | 76 | 75 | 0 | 47 |
| 2021Q4 | 66 | 67 | 68 | -2 | 48 |
| 2022Q1 | 62 | 62 | 63 | -1 | 49 |
| 2022Q2 | 77 | 77 | 72 | +5 | 50 |
| 2022Q3 | 62 | 61 | 59 | +3 | 51 |
| 2022Q4 | 59 | 59 | 56 | +3 | 52 |
| 2023Q1 | 61 | 61 | 57 | +4 | 53 |
| 2023Q2 | 65 | 65 | 61 | +4 | 54 |
| 2023Q3 | 57 | 57 | 51 | +6 | 55 |
| 2023Q4 | 42 | 42 | 41 | +1 | 56 |
| 2024Q1 | 53 | 53 | 49 | +4 | 57 |
| 2024Q2 | 55 | 55 | 50 | +5 | 58 |
| 2024Q3 | 50 | 49 | 47 | +3 | 59 |
| 2024Q4 | 45 | 45 | 42 | +3 | 60 |
| 2025Q1 | 48 | n/a | n/a | n/a | 43 |
| 2025Q2 | 47 | n/a | n/a | n/a | 44 |
| 2025Q4 | 49 | n/a | n/a | n/a | 45 |
| 2026Q1 | 41 | n/a | n/a | n/a | 46 |
Source: National Association of Home Builders, Eye on Housing posts on each SOMA release (secondary source), and Census Bureau SOMA statement on 2024 completions, Appendix B1; assembled by Feasibility Study Company out of the published documents, retrieved November 9, 2026 2.
What the comparison shows
For the 14 completion quarters from third quarter 2021 to fourth quarter 2024, the first-reported rate equals the Census initial value in 10 quarters and is within one point of it in all 14 24748495051525354555657585960. It is 3 to 6 points above the Census final value in each quarter from second quarter 2022 to fourth quarter 2024 except fourth quarter 2023, where it is 1 point above 2. The reported figure is the initial figure, and the initial figure is the one that is revised down.
The 2025 and 2026 values
The first-reported rates are 48 percent for first quarter 2025 completions 43, 47 percent for second quarter 2025 44, 49 percent for fourth quarter 2025 45 and 41 percent for first quarter 2026 46. Table 6 has no value for third quarter 2025. The association counts consecutive quarters below 50 percent at 4 in its March 2026 note, 6 in May 2026 and 7 in August 2026, which also calls 41 percent the lowest rate since the fourth quarter of 2023, and it states that the data had never before shown more than two consecutive quarters under 50 percent 444546. Those counts rest on the association's figures, which are first-published only for the newest quarter; third quarter 2024, which each count includes, was first reported at 50 percent and stands at 49 in the Census initial estimate and 47 in the final one 259. On the Census final values in Table 3, the quarters at or below 50 percent run from fourth quarter 2023 through fourth quarter 2024, five quarters, with second quarter 2024 at exactly 50 2.
Twelve-month rates, asking rent and completions
The August 2026 note reports that 89 percent of the 93,120 units completed in the first quarter of 2025 were absorbed after twelve months, the second lowest twelve-month rate in the SOMA data behind 86 percent for first quarter 2020 completions 46. The annual twelve-month rate in the long series was 87 percent for 2008 completions 5. The association reports a median asking rent of $2,034 for first quarter 2026 completions, up 6.5 percent from $1,909 a year earlier, and says fourth quarter 2025 was the first quarter in which the median topped $2,000 4645. The median was $1,666 for first quarter 2018 completions 31. Both are nominal asking rents at the three-month interview, for different cohorts. The first-reported count of units completed was 77,380 for fourth quarter 2025, the lowest since second quarter 2022 (76,630), and 73,510 for first quarter 2026 4546.
The wider supply context is the Census New Residential Construction series. Five-unit-and-larger buildings reached 591,700 completions in 2024, the highest annual count since 1974 (692,900), and 467,700 in 2025 29. The release of October 20, 2026 gives a September 2026 seasonally adjusted annual rate of 339,000 for such units 32. These counts include condominium, subsidized and other units outside the SOMA rate, so they overstate the units in scope. Table 7 gives the annual series.
| Year | Units completed |
|---|---|
| 2012 | 157,600 |
| 2013 | 186,200 |
| 2014 | 255,600 |
| 2015 | 310,300 |
| 2016 | 311,000 |
| 2017 | 346,900 |
| 2018 | 335,600 |
| 2019 | 342,900 |
| 2020 | 364,700 |
| 2021 | 363,700 |
| 2022 | 359,100 |
| 2023 | 438,300 |
| 2024 | 591,700 |
| 2025 | 467,700 |
Source: U.S. Census Bureau, New Residential Construction completions time series, file modified October 27, 2026, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 29.
Regional three-month rates differ by 12 to 18 points in recent years, and the ranking does not persist.
The survey reports four Census regions. The Northeast is the six New England states with the three Middle Atlantic states, and the South stretches from the District of Columbia to the Gulf Coast 7. Table 8 gives the three-month rate by region for 2007 to 2016 completions. The spread between the highest and lowest region averaged 16.8 points for 1995 to 2016 completions and was 12 to 18 points in each year from 2013 to 2016. The widest spread was 30 points, in 2009 (Midwest 74, West 44); in 2012 it was 28 points (Northeast 88, Midwest 60) 5232461. For 2012 completions the long-series table of the 2015 report prints 66, 61, 72 and 42 against the four regions, which are the 2013 report's figures for inside core-based statistical areas, inside and outside principal cities, and outside core-based statistical areas; the regional figures here are the 2013 report's 561. The South had the lowest rate in 11 of the 22 years from 1995 to 2016.
| Completion year | United States | Northeast | Midwest | South | West |
|---|---|---|---|---|---|
| 2007 | 55 | 66 | 58 | 52 | 58 |
| 2008 | 50 | 52 | 58 | 48 | 51 |
| 2009 | 51 | 56 | 74 | 49 | 44 |
| 2010 | 61 | 79 | 68 | 60 | 55 |
| 2011 | 59 | 67 | 53 | 60 | 54 |
| 2012 | 63 | 88 | 60 | 61 | 61 |
| 2013 | 63 | 63 | 78 | 60 | 64 |
| 2014 | 61 | 54 | 60 | 61 | 66 |
| 2015 | 60 | 62 | 67 | 53 | 66 |
| 2016 | 55 | 49 | 56 | 51 | 63 |
Source: U.S. Census Bureau, Survey of Market Absorption annual absorption reports (2013, 2015, 2016 and 2017 reports; 2012 from Table 1 of the 2013 report), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 5612324.
The ranking moves
Table 9 weights the annual regional rates by the unfurnished completions of each region and year. In 2007 to 2011 the West had the lowest weighted rate (51.1) and the Midwest and Northeast the highest (62.4 and 62.7). In 2012 to 2016 the West had the highest (64.5) and the Northeast fell to 58.7 52761622324. A region's rank in one period is not a basis for assuming the same rank in the next, and a study that uses a region's average for a state or a metropolitan area is using a number that moved by more than ten points between two five-year periods.
Small regions are noisy. The Northeast reported 4,700 unfurnished completions in 2005 and a rate of 75 percent, then 5,800 completions in 2006 and a rate of 41 percent. The South is the largest region by volume: 126,800 of the 262,400 regional completions in 2016 (48 percent), and 44 to 56 percent of each year's total from 2012 to 2016 5272324.
| Geography | Completion years | Units completed | 3-month rate, weighted (%) | 12-month rate, weighted (%) | Lowest annual 3-month (%) | Highest annual 3-month (%) |
|---|---|---|---|---|---|---|
| United States | 2007 to 2011 | 578,000 | 54.0 | 89.7 | 50 | 61 |
| United States | 2012 to 2016 | 973,800 | 59.6 | 95.2 | 55 | 63 |
| Northeast | 2007 to 2011 | 41,300 | 62.7 | 92.8 | 52 | 79 |
| Northeast | 2012 to 2016 | 111,300 | 58.7 | 95.6 | 49 | 88 |
| Midwest | 2007 to 2011 | 78,700 | 62.4 | 86.4 | 53 | 74 |
| Midwest | 2012 to 2016 | 154,800 | 62.7 | 96.2 | 56 | 78 |
| South | 2007 to 2011 | 324,900 | 52.1 | 90.2 | 48 | 60 |
| South | 2012 to 2016 | 465,400 | 56.2 | 94.2 | 51 | 61 |
| West | 2007 to 2011 | 133,000 | 51.1 | 88.7 | 44 | 58 |
| West | 2012 to 2016 | 242,100 | 64.5 | 96.7 | 61 | 66 |
Source: U.S. Census Bureau, Survey of Market Absorption annual absorption reports: rates for 2007 to 2011 and 2014 from the 2015 report's Table 15, for 2012 and 2013 from Table 1 of the 2013 and 2014 reports, and for 2015 and 2016 from Table 1 of the 2016 and 2017 reports; completions from the 2015 report's Table 14 and Table 1 of the 2016 and 2017 reports; retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables; weights are the published unfurnished completions of each region and year 52761622324.
Geography in the last static quarter
Table 10 is the geography table for first quarter 2018 completions, with margins of error. The national rate was 55 percent with a margin of 3.4 points. Inside principal cities the rate was 51 percent (margin 2.3) and outside them 63 percent (margin 7.2). By region the margins were 16.7 points for the Northeast, 6.8 for the Midwest, 2.3 for the South and 4.8 for the West 31. A Northeast rate of 73 percent with a margin of 16.7 points is consistent with anything from about 56 to about 90 percent. Later regional estimates are in the Table Creator for apartments as a whole, because its crosswalk says region cannot be combined with the furnished and unfurnished split, and a study should read the current release and its margin for the region it uses 63.
| Geography | Units completed | Margin of error | Median asking rent ($) | Margin of error ($) | Share of units (%) | Rented within 3 months (%) | Margin of error (points) |
|---|---|---|---|---|---|---|---|
| United States | 62,400 | 3,115 | 1,666 | 91 | 100 | 55 | 3.4 |
| Inside principal city of a CBSA | 38,500 | 4,371 | 1,807 | 103 | 62 | 51 | 2.3 |
| Outside principal city of a CBSA | 23,900 | 5,167 | 1,543 | 110 | 38 | 63 | 7.2 |
| Northeast | 9,500 | 2,579 | 2,261 | 386 | 15 | 73 | 16.7 |
| Midwest | 7,300 | 388 | 1,247 | 97 | 12 | 48 | 6.8 |
| South | 28,200 | 1,200 | 1,532 | 66 | 45 | 51 | 2.3 |
| West | 17,400 | 1,475 | 1,796 | 260 | 28 | 56 | 4.8 |
Source: U.S. Census Bureau, SOMA 2018 quarterly Table 8 (second quarter 2018 release), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 31.
Rent class explains little of the three-month rate, except at the top of the range.
The survey groups units by asking rent in nominal dollars at the three-month interview. For 2016 completions the median asking rent was $1,480 and the classes ran from "Less than $850" to "$2,450 or more" in $200 steps 1. Table 11 gives the share absorbed at 3, 6, 9 and 12 months by class. The three-month rate is 51 to 60 percent in every class up to $2,449 and 45 percent in the top class, which held 15 percent of the units. The top class is also the slowest at the other three horizons (67, 81 and 89 percent against 71 to 84, 85 to 93 and 92 to 97 percent elsewhere) 1.
| Asking rent class | Units completed | Share of units (%) | 3 months (%) | 6 months (%) | 9 months (%) | 12 months (%) |
|---|---|---|---|---|---|---|
| All classes | 262,500 | 100 | 55 | 76 | 88 | 94 |
| Less than $850 | 20,600 | 8 | 53 | 74 | 88 | 95 |
| $850 to $1,049 | 33,400 | 13 | 58 | 84 | 93 | 97 |
| $1,050 to $1,249 | 35,900 | 14 | 60 | 78 | 88 | 94 |
| $1,250 to $1,449 | 37,000 | 14 | 59 | 79 | 90 | 96 |
| $1,450 to $1,649 | 28,800 | 11 | 56 | 77 | 89 | 95 |
| $1,650 to $1,849 | 25,000 | 10 | 51 | 75 | 88 | 95 |
| $1,850 to $2,049 | 19,400 | 7 | 58 | 79 | 89 | 95 |
| $2,050 to $2,249 | 10,400 | 4 | 52 | 76 | 88 | 95 |
| $2,250 to $2,449 | 12,700 | 5 | 51 | 71 | 85 | 92 |
| $2,450 or more | 39,300 | 15 | 45 | 67 | 81 | 89 |
Source: Census Bureau, 2017 annual SOMA report (completions in 2016), Table 2, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 1.
Below the top class the gradient is weak: the three-month rates run 53, 58, 60, 59, 56, 51, 58, 52 and 51 percent from the lowest class upward, with no order. A study that assumes a luxury tier leases as fast as the market average is not supported by this table. The rent here is the asking rent, and what the tenant pays after concessions is a different number; that difference is treated in the note on effective rent versus asking rent 1.
The single-quarter version
Table 12 repeats the exercise for first quarter 2018 completions, the last static quarter. The median asking rent was $1,666 (margin 91) on 62,400 units. Three-month rates ran from 50 to 65 percent across the classes, with margins of 3.3 to 10.4 points. The class with the highest rate, $1,850 to $2,049 at 65 percent, has a margin of 10.4 points, and the top class of $2,450 or more, at 50 percent, held 19 percent of the units against 15 percent in 2016 33. Single-quarter rent classes are too noisy to rank. By bedroom count the rates were 56 percent (margin 3.6) for fewer than two bedrooms and 55 percent (margin 4.1) for two or more 33.
Rent classes are nominal. The Bureau revised them in 2014 30, and the Table Creator groups its rent categories before and after 2019 17. No real-dollar tiers exist, and the share of units in a fixed top class rises with inflation, so the 15 percent in 2016 and the 19 percent in 2018 reflect that as well as any change in the product.
| Asking rent class | Units completed | Share of units (%) | Margin of error (points) | Rented within 3 months (%) | Margin of error (points) |
|---|---|---|---|---|---|
| All classes | 62,400 | 100 | X | 55 | 3.4 |
| Less than $850 | 2,200 | 4 | 1.8 | 57 | 8.2 |
| $850 to $1,049 | 3,300 | 5 | 1 | 63 | 4.7 |
| $1,050 to $1,249 | 7,300 | 12 | 1.9 | 57 | 5.7 |
| $1,250 to $1,449 | 9,600 | 15 | 2.7 | 59 | 5.9 |
| $1,450 to $1,649 | 8,400 | 13 | 2.6 | 52 | 5.1 |
| $1,650 to $1,849 | 6,100 | 10 | 1.3 | 51 | 3.3 |
| $1,850 to $2,049 | 7,300 | 12 | 3 | 65 | 10.4 |
| $2,050 to $2,249 | 3,200 | 5 | 0.8 | 50 | 6.4 |
| $2,250 to $2,449 | 3,000 | 5 | 1.5 | 52 | 7.9 |
| $2,450 or more | 12,000 | 19 | 3 | 50 | 6.9 |
Source: U.S. Census Bureau, SOMA 2018 quarterly Table 5 (second quarter 2018 release), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 33.
Absorption is fastest in the first three months, and 2016 completions reached 90 percent in about 10 months.
The curve is the cumulative share of units absorbed at 3, 6, 9 and 12 months after completion. For 2016 completions it is 55, 76, 88 and 94 percent 1. The three-month rate alone can mislead, because two years with the same rate can have different curves: 2007 and 2016 completions both read 55 percent at three months, and the twelve-month rates are 92 and 94 percent. In 2008 the three-month rate was 50 percent and the twelve-month rate 87 percent, the lowest twelve-month rate in the annual series 5.
The implied pace in Table 13 comes from the published points. Cumulative absorption is taken as zero at completion and the published shares are joined by straight lines; the pace is the change in share per month, in percentage points of completed units. Because the survey does not report leasing before completion separately, the pace in the first three months is an average over the interval and not a measured rate. For 2016 completions the pace is 18.3, 7.0, 4.0 and 2.0 points per month in the four intervals, and 90 percent is reached at 10.0 months. For 1978 completions the pace is 27.3, 3.7, 1.3 and 0.7 points per month and 90 percent is reached at 5.2 months. Interpolated months to 50 percent run from 1.8 months (1978 and 1979) to 3.0 months (2008) 2252324.
| Completion year | 3-month rate (%) | 12-month rate (%) | Gap (points) | Pace, months 0 to 3 | Pace, months 3 to 6 | Pace, months 6 to 9 | Pace, months 9 to 12 | Months to 90 percent |
|---|---|---|---|---|---|---|---|---|
| 1978 | 82 | 99 | 17 | 27.3 | 3.7 | 1.3 | 0.7 | 5.2 |
| 1987 | 63 | 94 | 31 | 21.0 | 6.3 | 2.7 | 1.3 | 9.0 |
| 1994 | 80 | 98 | 18 | 26.7 | 4.3 | 1.3 | 0.3 | 5.3 |
| 1995 | 73 | 97 | 24 | 24.3 | 5.3 | 1.7 | 1.0 | 6.6 |
| 1997 | 73 | 97 | 24 | 24.3 | 5.7 | 1.7 | 0.7 | 6.0 |
| 2001 | 64 | 95 | 31 | 21.3 | 6.0 | 3.0 | 1.3 | 8.7 |
| 2005 | 64 | 96 | 32 | 21.3 | 6.7 | 2.7 | 1.3 | 8.2 |
| 2007 | 55 | 92 | 37 | 18.3 | 6.3 | 4.0 | 2.0 | 11.0 |
| 2008 | 50 | 87 | 37 | 16.7 | 6.0 | 3.7 | 2.7 | not reached by month 12 |
| 2009 | 51 | 90 | 39 | 17.0 | 6.7 | 3.7 | 2.7 | 12.0 |
| 2012 | 63 | 94 | 31 | 21.0 | 6.0 | 3.0 | 1.3 | 9.0 |
| 2016 | 55 | 94 | 39 | 18.3 | 7.0 | 4.0 | 2.0 | 10.0 |
Source: U.S. Census Bureau, Survey of Market Absorption annual absorption reports (1995, 2015, 2016 and 2017 reports), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables; the pace is a straight-line interpolation of the published 3, 6, 9 and 12 month shares, and exact halves are rounded to the even digit 2252324.
The gap between the twelve-month and three-month shares measures how much leasing is left after the first quarter. It was 17 points for 1978 and 1979 completions, 24 for 1995, 31 for 2001, 37 for 2007 and 2008, and 39 for 2009 and 2016 52224. A wider gap means a longer tail. In 2008 the 90 percent mark was not reached within twelve months, and in 2009 it was reached at 12.0 months. No completion year from 2001 to 2016 reached 90 percent in fewer than 8.2 months (2005), while 1995 to 2000 took 6.0 to 7.3 months and 1976 to 1979 took 5.1 to 5.3.
The quarterly tables show the same curve with margins. Table 14 gives the 3, 6, 9 and 12 month rates for the 2016 and 2017 cohorts, with margins of error in parentheses. The twelve-month margin is 1.0 to 1.7 points on these cohorts and the three-month margin is 2.4 to 2.9, so the tail of the curve is better measured than the head 10. The same release implies cohort leasing of 3,180 units a month in the first three months for first quarter 2012 completions and 16,720 for third quarter 2015 completions. These are totals for the national cohort and cannot be scaled down to one property without a comparable lease-up.
| Completion quarter | Units completed (margin) | 3 months, % (margin) | 6 months, % (margin) | 9 months, % (margin) | 12 months, % (margin) |
|---|---|---|---|---|---|
| 2016Q1 | 55,100 (1,916) | 55 (2.8) | 79 (2.2) | 90 (1.7) | 95 (1.0) |
| 2016Q2 | 62,000 (1,462) | 57 (2.9) | 78 (2.3) | 87 (1.8) | 94 (1.1) |
| 2016Q3 | 73,400 (1,861) | 59 (2.4) | 78 (1.8) | 89 (1.4) | 95 (1.0) |
| 2016Q4 | 72,100 (4,157) | 47 (2.4) | 71 (2.1) | 86 (1.4) | 93 (1.2) |
| 2017Q1 | 61,400 (2,002) | 54 (2.8) | 77 (2.5) | 87 (2.0) | 92 (1.7) |
| 2017Q2 (revised) | 76,400 (2,110) | 58 (2.7) | 78 (2.5) | 87 (2.2) | 94 (1.6) |
Source: U.S. Census Bureau, SOMA 2018 quarterly Table 1 (second quarter 2018 release), retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 10.
Metropolitan estimates ended with the 2018 releases, and many cells were withheld.
The Census Bureau published quarterly metropolitan absorption tables for the 2018 releases 3. Each of the two summary files lists 45 core-based statistical areas 3435. In the first release, covering completions as of the fourth quarter of 2017, 31 of the 45 areas had a published three-month estimate, 10 were withheld and 4 reported no construction. In the second, covering completions as of the first quarter of 2018, 20 had a published estimate, 17 were withheld and 8 reported no construction. The letter D marks an estimate "withheld to avoid disclosing data for individual builders." 34 Published margins of error had a median of 5.4 points in the first release and 4.2 in the second, and reached 15.2 and 17.9 points 3435. Table 15 summarizes the two releases.
| Release | Areas listed | Published | Withheld (D) | No construction | Median margin (points) | Largest margin (points) | Published with margin of 10 or more |
|---|---|---|---|---|---|---|---|
| Q1 2018 release (completions as of 4th Quarter 2017) | 45 | 31 | 10 | 4 | 5.4 | 15.2 | 6 |
| Q2 2018 release (completions as of 1st Quarter 2018) | 45 | 20 | 17 | 8 | 4.2 | 17.9 | 2 |
Source: U.S. Census Bureau, SOMA 2018 metropolitan summary files, first and second quarter 2018 releases, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 3435.
Table 16 gives the three-month rate in the first column of the first release for selected areas; the file heads that column "4th quarter 2017 through 3rd quarter 2017" and does not say whether it covers one completion quarter or several. The published rates run from 31 percent (Jacksonville) to 99 percent (Colorado Springs). The 99 percent figure has a margin of 0.1 points, which is the pattern of very few buildings and not a precise reading of a market. Chicago (38 percent, margin 12.1), Indianapolis (52, margin 13.7) and Baltimore (48, margin 15.2) show how wide the margins were in larger areas. Detroit, Kansas City and Minneapolis are withheld, and Milwaukee reported no construction 34.
For a study the conclusion is plain. Metropolitan detail ended in 2018, and what existed was a three-month estimate for fewer than half of the listed areas by the last release. The current Census descriptions divide absorption within metropolitan areas into inside and outside the principal city 7. A study of one market uses the national and regional curve as a benchmark and takes the market itself from comparable lease-ups, a method developed in absorption forensics.
| Core-based statistical area | Absorbed within 3 months (%) | Margin of error (points) |
|---|---|---|
| Atlanta-Sandy Springs-Roswell, GA | 46 | 5.6 |
| Baltimore-Columbia-Towson, MD | 48 | 15.2 |
| Charlotte-Concord-Gastonia, NC-SC | 33 | 3 |
| Chicago-Naperville-Elgin, IL-IN-WI | 38 | 12.1 |
| Colorado Springs, CO | 99 | 0.1 |
| Dallas-Fort Worth-Arlington, TX | 55 | 2.1 |
| Denver-Aurora-Lakewood, CO | 34 | 2.6 |
| Houston-The Woodlands-Sugar Land, TX | 39 | 2.5 |
| Indianapolis-Carmel-Anderson, IN | 52 | 13.7 |
| Jacksonville, FL | 31 | 1.7 |
| Las Vegas-Henderson-Paradise, NV | 75 | 1.7 |
| Memphis, TN-MS-AR | 70 | 1.2 |
| New York-Newark-Jersey City, NY-NJ-PA | 58 | 6 |
| Detroit-Warren-Dearborn, MI | withheld (D) | |
| Kansas City, MO-KS | withheld (D) | |
| Minneapolis-St. Paul-Bloomington, MN-WI | withheld (D) | |
| Milwaukee-Waukesha-West Allis, WI | no construction |
Source: U.S. Census Bureau, SOMA 2018 metropolitan summary file, first quarter 2018 release, retrieved November 9, 2026; figures calculated by Feasibility Study Company using the published tables 34.
The weighted response rate fell from 99.4 percent for 2014 completions to 90.7 percent for 2024.
The Census Bureau produces response rates as its measure of nonsampling error for SOMA 2. The rate is the weighted number of completed building interviews divided by the weighted number of eligible buildings, calculated at the initial interview and revised after the fourth interview 2. An eligible building is privately financed, residential and has five or more units, and only completed interviews enter the estimates 2. Table 17 gives the annual rates, each from that year's statement.
| Completion year | Sample size | Eligible interviews | Eligible noninterviews | Ineligible | Weighted response rate (%) | Statement |
|---|---|---|---|---|---|---|
| 2014 | 3,600 | 3,500 | 30 | 60 | 99.4 | 2 |
| 2015 | 4,100 | 3,900 | 100 | 60 | 97.9 | 2 |
| 2016 | 4,300 | 4,000 | 60 | 150 | 98.4 | 2 |
| 2017 | 4,400 | 4,200 | 100 | 50 | 97.4 | 236 |
| 2018 | 4,000 | 3,900 | 70 | 30 | 98.1 | 37 |
| 2019 | 4,100 | 3,900 | 150 | 60 | 97.1 | 38 |
| 2020 | 4,500 | 4,200 | 250 | 150 | 95.7 | 39 |
| 2021 | 4,000 | 3,800 | 100 | 100 | 96.6 | 40 |
| 2022 | 3,900 | 3,800 | 150 | 40 | 97.0 | 14 |
| 2023 | 4,900 | 4,500 | 300 | 70 | 92.9 | 15 |
| 2024 | 6,000 | 5,300 | 600 | 100 | 90.7 | 2 |
Source: Census Bureau SOMA statements: 2014 to 2017 from Appendix A of the 2024 statement and 2018 to 2024 from Table 1 of each statement (the 2017 statement prints unrounded counts of 4,394, 4,224, 119 and 51 and the same 97.4 percent); assembled by Feasibility Study Company out of the published documents, retrieved November 9, 2026 236.
The rate was 99.4 percent for 2014 completions and held at 95.7 percent or higher through 2022. It fell to 92.9 percent for 2023 and 90.7 percent for 2024. In 2024 the sample was 6,000 buildings, of which 5,300 were eligible interviews, 600 were eligible noninterviews and 100 were ineligible; in 2014 the sample was 3,600, with 3,500 interviews and 30 eligible noninterviews. The lowest quarterly rate is 88.7 percent for first quarter 2024, followed by 89.4 percent for fourth quarter 2024 2.
The Bureau states that no other direct measurements of nonsampling error have been obtained and that most response and operational errors are detected and corrected in review 2. The margins of error in Table 3 measure sampling error only. A response rate of 90.7 percent does not by itself show bias, but the managers who decline to be interviewed may differ from those who respond, and a reviewer should read the 2023 and 2024 estimates as possibly carrying more uncertainty than the published margin, and should not read a one-point difference as a finding.
Ten limits a reviewer should check before relying on a SOMA rate.
The clock starts when a building is half available
Completion is 50 percent of units occupied or available, so a project that pre-leased is measured from a later point than one that opened its first unit 2. A lease-up counted from the first move-in is not on the same footing as the three-month rate (Taxonomy F4a).
Absorbed is not stabilized
The survey counts first rentals at 3, 6, 9 and 12 months 2. A lender's occupancy test measures occupancy over a window, and a first rental does not guarantee a unit stays occupied (Taxonomy F4c).
Rent is an asking rent at month three
The tables record asking rent at the initial interview 11. Concessions and loss to lease are not in it, and a study that carries the survey's rent into revenue is on asking rent, the failure that the page on rent comparability analysis lists as Taxonomy F5a.
Initial estimates are revised, and usually down
32 of 36 quarters were revised down, by 2.2 points on average 2. A figure from the latest release without the word "initial" is not traceable to a final value (Taxonomy F9b).
Cell margins are wide
National quarterly margins are 2.2 to 4.1 points, and regional, rent and metropolitan cells run to 17.9 points 23135. For the static annual tables, which ended with the 2017 report, the Bureau published generalized standard-error tables, not a value per cell; since second quarter 2018 completions the Table Creator gives a margin for every estimate 41422.
The scope is market-rate and varies
Unfurnished rental units in scope were 41 to 84 percent of the completions in structures with five or more units, and subsidized and tax credit units are outside the rate 6.
Nonresponse has risen
The weighted response rate was 90.7 percent for 2024 completions against 99.4 percent for 2014 2.
A cohort is not a building
The rate is the share of a national cohort of units. A single property can lease far faster or slower, and the national rate is a benchmark, not a prediction for a given building (Taxonomy F4a).
A study should state which SOMA series it uses, then test its pace against local comparables.
A study that cites SOMA states the series, the completion period, the region and the stage (initial or final). It cites the final estimate where one exists and gives the margin of error. It uses the 3, 6, 9 and 12 month curve and not the three-month rate alone. It names the 50 percent completion rule 2 and says how the subject's own lease-up clock is counted. It treats a regional or rent-tier cell as a range. The monthly ramp, the reserves sized from it and the lender's test window are in the absorption and lease-up page; this note supplies the benchmark inputs to that method. The survey records units absorbed at 3, 6, 9 and 12 months after completion 4. Each state housing agency writes its own absorption and capture rules for market studies; the capture rate thresholds by state note compares them, while this one supplies only the observed pace.
A reviewer can test a schedule against the long series, selected years of which are in Table 13. Only 1976 to 1981, 1992, 1994 and 1997 reached 90 percent in six months or less, a span of 5.1 to 6.0 months. A schedule that reaches 90 percent leased in eight months is faster than every completion year from 2001 to 2016, the fastest of which took 8.2 months (2005). The study should then say why the subject beats the national benchmark, and the reason has to be local and specific: a pre-leasing campaign, a supply gap, a rent below the market 2252324.
The common errors are the ones in the Limits section. A study that takes the first-published rate for the latest quarter, a regional cell as a market rate, a tax credit property against a market-rate series, a metropolitan estimate that no longer exists, or the three-month rate as a stabilized occupancy has not met Taxonomy F4a or F9b. The same series, read with the final estimate, the curve, the margin and the scope, supports a defensible benchmark.
Survey of Market Absorption data questions.
What is the Survey of Market Absorption?
It is the Census Bureau's survey of how quickly new apartments in buildings of five or more units are rented or sold after completion, sponsored by the Department of Housing and Urban Development and drawn from the Survey of Construction 4. Interviews at 3 months, and if needed at 6, 9 and 12 months, give the share of units absorbed at each interval 42.
How fast do new apartments lease according to the Census?
For 2016 completions, 55 percent of unfurnished rental apartments were absorbed within three months, 76 percent within six, 88 percent within nine and 94 percent within twelve 1. The annual three-month rate has ranged from 50 percent (2008) to 82 percent (1978 and 1979) 5, and final quarterly rates for 2016 to 2024 ranged from 41 to 75 percent 2.
What is the Census absorption rate for new apartments?
It is the share of units in a completion cohort that have been initially rented or sold and are no longer available 9. The headline rate is for privately financed, nonsubsidized, unfurnished rental units, measured three months after a building is complete, meaning half its units are occupied or available 82.
Are the first published quarterly rates final?
No. Each quarter's rate is revised in the next three quarters and fixed after the fourth 2. Of 36 quarters from 2016 to 2024, the final three-month rate was below the initial rate in 32, by 2.2 points on average and 3.2 points for 2022 to 2024 2. A study should cite the final estimate and label any first-published figure as initial.
Does the survey publish metropolitan or city data?
Metropolitan tables for 45 core-based statistical areas were published through the 2018 releases, with many areas withheld to avoid disclosing individual builders 3435. The Census descriptions now divide metropolitan areas into inside and outside the principal city 7. A single market's pace has to come from comparable lease-ups.
Where can the latest SOMA data be found?
Since September 27, 2018 the data have been in the Census Bureau's Table Creator, which carries completions from 2016 416. The release page, last revised August 17, 2026, lists the 2025 annual tables for February 2027 and the second quarter 2026 completions for November 2026 21.
Does the survey cover affordable or tax credit apartments?
No. The absorption rate covers privately financed, nonsubsidized units, and subsidized and tax credit units are a separate category in the completions tables 86. The Census description of the data adds that privately financed units include privately owned housing subsidized by state and local governments 7. The glossary defines subsidized units as those where a government program pays part of the cost 9. The rate is a market-rate benchmark, not a measure for a tax credit property.
Where this research is applied.
Asset classes
Loan programs
Worked cases
- Multifamily Feasibility Study, Arizona: An Agency / Conventional Worked Case
- Multifamily Feasibility Study, Florida: An Agency / Conventional Worked Case
- Multifamily Feasibility Study, Georgia: An Agency / Conventional Worked Case
- Multifamily Feasibility Study, New York: An Agency / Conventional Worked Case
- Multifamily Feasibility Study, North Carolina: An Agency / Conventional Worked Case
- Multifamily Feasibility Study, Texas: An Agency / Conventional Worked Case
Related research notes
- Absorption Forensics: Testing a Lease-Up Assumption Before the Reviewer Does
- Capture Rate Thresholds by State: What 38 Housing Credit Agencies Require
- Office Repurposing: A Field Guide to the Conversions That Actually Pencil
- The Break-Even Occupancy Trap: Why a Deal Can Clear DSCR and Still Not Cash-Flow
- The Concession Gap: Why Underwriting to Asking Rent Breaks a Multifamily Pro Forma
- The Coverage-Constrained Takeout: When the Permanent Loan Sizes to Coverage, Not Value
Benchmarking a lease-up to Census data? Show the stage, the margin and the scope.
Feasibility Study Company prepares studies that cite the final estimate, the full 3, 6, 9 and 12 month curve and the survey's scope, and that test the pace against comparable lease-ups.
Request a methodology briefingNumbered references.
Census Bureau pages, statements, annual reports, releases and files are numbered 1 to 42 and 61 to 63, and the secondary sources 43 to 60; each entry gives its edition or date and the retrieval date.
- U.S. Census Bureau, SOMA 2017 annual absorption report, Table 2: asking rent and absorption rates for unfurnished apartments by geographic area (completions in 2016, file ann17t2.xlsx), 2017 annual report, file dated October 3, 2018; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2024 Completions (Appendices A1 and B1), statement issued 2026, clearance CBDRB-FY26-SEHSD003-062; retrieved November 9, 2026.
- U.S. Census Bureau, 2018 Quarterly Metropolitan Absorptions Tables, page last revised October 8, 2021; retrieved November 9, 2026.
- U.S. Census Bureau, About SOMA, page last revised May 12, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2015 annual absorption report, Table 15: absorption rates for unfurnished apartments by region, 1970 and 1975 to 2014 (file ann15t15.xls), 2015 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2017 annual absorption report, Table 9: apartments completed in buildings with five units or more, 1970 to 2016 (file ann17t9.xlsx), 2017 annual report, file dated April 3, 2018; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Characteristics of Data, page last revised October 8, 2021; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Source and Accuracy, page last revised May 19, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Glossary, page last revised May 12, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2018 quarterly absorption Table 1: absorption rates of privately financed, nonsubsidized, unfurnished rental apartments, 2012 to 2018 (second quarter 2018 release, file q218tab1.xlsx), second quarter 2018 release; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 1995 annual absorption report, Table 2: asking rent and absorption rates (completions in 1994, file ann95t2.txt), 1995 annual report; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Frequently Asked Questions, page last revised May 12, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA items booklet, April 2019; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2022 Completions (revised file), statement for 2022 completions, clearance CBDRB-FY24-POP001-0041; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2023 Completions, statement for 2023 completions, clearance CBDRB-FY25-POP001-0052; retrieved November 9, 2026.
- U.S. Census Bureau, About the SOMA Table Creator (completions 2016 and later), page last revised May 19, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Table Creator updates released May 2024, May 2024; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Table Creator users guide (cell suppression, margin of error), current edition; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Data Tables (static tables, 1995 to 2018), page last revised May 12, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, 2018 Quarterly Absorptions Tables, page last revised October 8, 2021; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA Upcoming Release Dates, page last revised August 17, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 1995 annual absorption report, Table 12: long series of absorption rates (completions in 1994, file ann95t12.txt), 1995 annual report; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2016 annual absorption report, Table 1 (completions in 2015, file ann16t1.xls), 2016 annual report, file dated May 2, 2017; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2017 annual absorption report, Table 1: absorption by geography (completions in 2016, file ann17t1.xlsx), 2017 annual report, file dated April 3, 2018; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 1999 annual absorption report, Table 1 (completions in 1998, file ann99t1.txt), 1999 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2010 annual absorption report, Table 1 (completions in 2009, file ann10t1.xls), 2010 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2015 annual absorption report, Table 14: apartments completed by region (completions in 2014, file ann15t14.xls), 2015 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2006 annual absorption report, Table 1 (completions in 2005, file ann06t1.txt), 2006 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, New Residential Construction: completions time series by structure size (comps_cust.xlsx), file modified October 27, 2026 (September 2026 data); retrieved November 9, 2026.
- U.S. Census Bureau, SOMA User Note (1990 weighting, 2014 interviewing software and asking rent ranges), page last revised November 7, 2024; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2018 quarterly absorption Table 8: number, asking rent and absorption rate of unfurnished apartments completed in the first quarter of 2018 by geographic area (second quarter 2018 release, file q218tab8.xlsx), second quarter 2018 release; retrieved November 9, 2026.
- U.S. Census Bureau, New Residential Construction press release, September 2026, release of October 20, 2026; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2018 quarterly absorption Table 5: asking rent and number of bedrooms of unfurnished apartments completed in the first quarter of 2018 and those rented within 3 months (second quarter 2018 release, file q218tab5.xlsx), second quarter 2018 release; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2018 metropolitan (CBSA) quarterly absorption estimates, summary file for the first quarter 2018 release (18q1hud.xls), first quarter 2018 release; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2018 metropolitan (CBSA) quarterly absorption estimates, summary file for the second quarter 2018 release (18q2hud.xls), second quarter 2018 release; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2017 Completions, statement for 2017 completions; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2018 Completions, statement for 2018 completions; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2019 Completions, statement for 2019 completions; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2020 Completions, statement for 2020 completions; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: Source and Accuracy Statement for 2021 Completions, statement for 2021 completions, clearance CBDRB-FY23-POP001-0057; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2015 annual absorption report, Table B: standard errors of estimated percentages (file ann15tb.xls), 2015 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2017 annual absorption report, appendix: standard errors for estimates of apartments (completions in 2016, file ann17appen.pdf), 2017 annual report, file dated April 3, 2018; retrieved November 9, 2026.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rises in the Second Quarter, August 29, 2025; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rate Remains Below 50%, March 3, 2026; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Rent Prices Continue to Rise, While Absorption Remains Low, May 29, 2026; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rate Remains Below 50%, August 28, 2026; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Absorption Continued to Soar while Condo Absorption Flat, February 28, 2022; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Absorption Increased while Completions Fell, June 2, 2022; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Rent Price Increased for Fifth Consecutive Quarter, August 25, 2022; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Absorption Rate Reaches 25-year High, December 1, 2022; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Absorption Rate Falls but Remains above 60%, February 27, 2023; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Absorptions Slightly Weaker in Fourth Quarter of 2022, May 25, 2023; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rates Move Higher, August 24, 2023; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Absorption of New Multifamily Units Rises as Completions Near 100K in Second Quarter of 2023, November 28, 2023; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rates Fall for Third Quarter Completions; Share of Subsidized Units Rose in 2022, February 28, 2024; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Rates Fall to Decade Low as Completions Remain Elevated, May 30, 2024; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Completions Rising, August 29, 2024; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Apartment Completions Rise Resulting in Slower Absorption Rates, November 27, 2024; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Completions Rise Again Pushing Absorption Rates Lower, February 28, 2025; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- National Association of Home Builders (Eye on Housing blog), Multifamily Absorption Moves Lower for New Apartments, May 30, 2025; retrieved November 9, 2026. Secondary source: trade-association commentary on a Census release, giving the first-published estimates for the completion quarter named; the estimates are later revised.
- U.S. Census Bureau, SOMA 2013 annual absorption report, Table 1: absorption rates for unfurnished apartments completed by geographic area (completions in 2012, file ann13t1.xls), 2013 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, SOMA 2014 annual absorption report, Table 1: absorption rates for unfurnished apartments completed by geographic area (completions in 2013, file ann14t1.xls), 2014 annual report, file dated December 16, 2016; retrieved November 9, 2026.
- U.S. Census Bureau, Survey of Market Absorption: crosswalk from the static tables to the Table Creator, current file; retrieved November 9, 2026.