II. Definitions
III. Source of Data and Survey Questionnaires
III.a. Units Authorized by Building
Permits
III.b. Housing Units Authorized,
but Not Started; Housing Starts; Housing Units Under Construction; and Housing
Completions
V. Sample Design
V.a. Units Authorized by Building
Permits
V.b. Housing Units Authorized,
but Not Started; Housing Starts; Housing Units Under Construction; and Housing
Completions
VI. Compilation of Data
VI.a. Units Authorized by Building
Permits
VI.b. Housing Units Authorized,
but Not Started; Housing Starts; Housing Units Under Construction; and Housing
Completions
VI.b.1. Methodology used for data
up through 1998
VI.b.2. Methodology used beginning
with 1999 data
VI.b.3. Adjustments for non-reporting
VII. Reliability of Data
VII.a. Units Authorized by
Building Permits
VII.a.1. Sampling and
non-sampling errors
VII.b. Housing Units Authorized,
but Not Started; Housing Starts; Housing Units Under Construction; and
Housing Completions
VII.b.1. Sampling errors
VII.b.2. Non-sampling
errors
VIII. Seasonal Adjustment
Northeast | Midwest | South | West |
Connecticut | Illinois | Alabama | Alaska |
Maine | Indiana | Arkansas | Arizona |
Massachusetts | Iowa | Delaware | California |
New Hampshire | Kansas | District of Columbia | Colorado |
New Jersey | Michigan | Florida | Hawaii |
New York | Minnesota | Georgia | Idaho |
Pennsylvania | Missouri | Kentucky | Montana |
Rhode Island | Nebraska | Louisiana | Nevada |
Vermont | North Dakota | Maryland | New Mexico |
Ohio | Mississippi | Oregon | |
South Dakota | North Carolina | Utah | |
Wisconsin | Oklahoma | Washington | |
South Carolina | Wyoming | ||
Tennessee | |||
Texas | |||
Virginia | |||
West Virginia |
The Census field representatives use interviewing software on laptop computers
to collect the data. Facsimiles of the computer-based questionnaires are
provided to respondents to familiarize them with the survey. These facsimiles
show the questions that are asked for housing units in single-family buildings
Form SOC-QI/SF.1 and in multiunit
buildings Form SOC-QI/MF.1.
For more geographic information, please refer to our definitions of Regions and Metropolitan Areas.
Within each of these 169 areas, the sample was selected from two different sample frames: permit-issuing places and land areas not covered by building permit systems.
Each permit-issuing place was assigned to one of six size classes based on a weighted average of 1978, 1981, and 1982 permit activity. The two most recent years available for the SOC redesign, 1981 and 1982, were unusual because of recession. Future activity was expected to return toward the pre-recession distribution. It was thought that averaging 1981 and 1982 with 1978 would include this tendency to return to the pre-recession distribution but also reflect a permanent change in the distribution of activity due to the changes in the economy. The activity levels were standardized by dividing by the US activity for each year because the activity levels were quite different in the three years. The permit places in each of the 169 areas were grouped into these six size classes and a systematic sample of places was selected from each size class. Places were selected at different sampling rates in each of the classes so that larger proportions of the places were selected from the larger size classes. For example, all places in the largest size classes fell into sample, but only an expected 1 in 40 of the places in the smallest size class fell into sample. Approximately 840 permit-issuing places were selected. Due to splits and mergers of local permit systems, there are now about 900 permit offices in the Survey of Construction.
Monthly, Census field representatives sample permits from these 900 permit-issuing places. They select permits for one-to-four-unit buildings with probability proportional to the number of units at an overall rate of 1 in 40. All permits for buildings with five units or more are selected.
Within each of the 169 areas, the land not covered by building permit systems, called nonpermit areas, was identified. Small land areas (1980 Census enumeration districts) in these nonpermit areas were grouped into two strata according to the 1980 population. Overall, 1 out of every 120 land areas was selected from the strata with the larger areas and 1 out of 600 was selected from the strata with the smaller areas. Monthly, census field representatives canvassed about 130 selected land areas looking for all housing units started.
In January 1995, the area covered by building permit systems was expanded
to 19,000 permit-issuing places. Canvassing was stopped in those selected
land areas now represented by permit-issuing places. Census field representatives
continue to canvass about 70 land areas still not covered by building permit
systems. Due to the small amount of new construction activity, many areas
are now only canvassed once every three months.
Monthly building permits data are available in four basic formats: State, Metropolitan Area (MA), County, and Place. Data are tabbed for the current month and for year-to-date. Year-to-date data include any late reports or corrections from prior months. (Summing the published monthly data will not generate the same estimate that we publish for year-to-date.) Monthly data are not revised except for the highest aggregates (US and region) after annual processing. State data includes division, region and US data. These are sample based estimates that represent the entire geographic area.
MA tables show all MAs, but most do not include complete counts on a monthly basis. All CMSAs and PMSAs and the top 25 MSAs (as of 1993) are completely covered monthly, the remaining are just the sum of monthly reporters with no estimate for annual reporters. Annual tables include estimates for all permit areas.
Monthly county data are the sum of the places requested to report monthly in a county, and for counties not fully covered by monthly reporters, county totals will be incomplete. Annual county totals include estimates for all permit offices.
Monthly place level data include municipalities requested to report monthly. Data for all permit-issuing municipalities are available annually.
There is a follow-up for nonreporters in which calls are made to delinquent offices to obtain data or correct address information. At the end of the year a second form is mailed to delinquent offices. If an office is 1-4 months delinquent, a form is sent for each missing month, and if 5 or more months delinquent, an annual form is sent.
Annual data are obtained by summing monthly data for monthly reporters
and using annual data for annual reporters. If both monthly and annual
data exist, the annual data are used. If no annual data are received, but
there were some months reported, the sum of the monthly reported and imputed
data is used rather than the imputed annual data. Building permits data
are not sample based on an annual basis, annual data are tabulated from
the entire universe of building permit offices.
The compilation of the housing starts series was a multistage process. First, an estimate was made monthly of the number of housing units for which building permits were issued in all 19,000 permit-issuing places as described above.
Second, for each permit selected in the 900 permit-issuing places, an inquiry was made of the owner or the builder to determine in which month and year the unit(s) covered by the permit was (were) started. In case the units authorized by a permit in a particular month were not started by the end of that month, follow-ups were made in successive months to find out when the units were actually started.
Ratios were calculated (by type of structure) of the number of units started to the number of units covered by permits based on estimates generated from the 900 SOC permit offices. Separate ratios were calculated for units started from permits of that month and of each preceding month for the past 7 years. These ratios, or starts rates, were then applied to the appropriate estimate of the number of units authorized by permits, based on the Building Permits Survey, in the corresponding months to provide estimates of the number of units started for each month of authorization.
Having produced estimates of the number of units started with permit authorization, two additional adjustments were made. An upward adjustment of 3.3 percent was made to the number of one-unit structures (single-family houses) started to account for those units started within permit-issuing areas but without permit authorization. (A study spanning a four year period during the mid 1960s indicated that permits were obtained for all buildings with two housing units or more.)
Adjustments were made to account for those units started prior to permit authorization and for late reports. These adjustments were based on historical patterns of pre-permit starts and late data.
The estimates for housing units started in the 19,000 permit-issuing places resulted from the procedures outlined above.
Third, units identified as started in the monthly canvass of nonpermit areas were weighted appropriately to provide an estimate of total housing starts in areas not covered by building permit systems.
Adding this estimate of starts in nonpermit areas to the estimate of starts in the 19,000 permit-issuing places resulted in an estimate of total private housing units started.
This same methodology was also used for the estimates of housing units authorized but not started, under construction, and completed.
The procedure described above was done by size of structure. A total of
14 different sets of starts rates that change from month-to-month were utilized
to calculate the number of housing units started by type of structure in
permit places. Eight sets of starts rates were used for one-unit structures:
separate sets of rates for metropolitan and nonmetropolitan areas within
each of the four regions. For structures with five units or more, separate
sets of starts rates were used for each of the four regions. Single sets of
starts rates were used for all regions for structures with two units and for
structures with three and four units. Starts by type of structure in nonpermit
areas were calculated directly in the estimating procedure described above.
The compilation of the housing starts series is a multistage process. First, an estimate is made monthly of the number of housing units for which building permits have been issued in all 19,000 permit-issuing places as described above.
Second, for each permit selected in the 900 permit-issuing places, an inquiry is made of the owner or the builder to determine in which month and year the unit(s) covered by the permit was (were) started. In case the units authorized by permits in a particular month are not started by the end of that month, follow-ups are made in successive months to find out when the units were actually started.
Ratios are calculated (by type of structure) of the number of units authorized by permits, based on the Building Permits Survey to the number of units authorized by permits based on estimates generated from the 900 SOC permit offices; separate ratios are calculated for that month and the prior 11 months. The 13th through 18th month back are summed and a ratio is calculated and all months from 19 months back through 60 months (5 years) are summed to get another ratio. These ratios are then applied to the appropriate estimate of the number of units started, based on the 900 SOC permit offices, in the corresponding months or groups of months to provide ratio adjusted estimates of the number of units started for each month or group of months.
Adjustments are made to account for those units started prior to permit authorization and for late reports. These adjustments are based on historical patterns of pre-permit starts and late data. No adjustment is made for units in permit areas built without a permit.
The estimates for housing units started in the 19,000 permit-issuing places result from the procedures outlined above.
Third, units identified as started in the monthly canvass of nonpermit areas are weighted appropriately to provide an estimate of total housing starts in areas not covered by building permit systems.
Adding this estimate of starts in nonpermit areas to the estimate of starts in the 19,000 permit-issuing places results in an estimate of total private housing units started.
This same methodology is also used for the estimates of housing units authorized but not started, under construction, and completed.
The procedure described above is computed by size of structure. A total
of 8 different sets of authorization ratios that change from month-to-month
are utilized to calculate the number of housing units started by type of
structure in permit places. The rates are calculated for one-unit structures
for each of the four regions and for all 2 or more unit structures for each
of the four regions. Starts by type of structure in nonpermit areas are
calculated directly in the estimating procedure described above.
The reported statistics are influenced by the following factors:
1. Some new residential construction work in building permit jurisdictions escapes recording. However, we suspect that the number of such unrecorded units is very small.
2. Detailed recent evidence is lacking as to how closely the valuation recorded for building permit purposes approximates the dollar amount of construction work involved.
3. Changes in boundaries of localities due to annexations, new incorporations, etc., result in some problems of comparability over time, even for statistics for the same places.
4. Some building permit jurisdictions close their books a few days before the end of the month, so that the time reference for permits is not in all cases strictly the calendar month.
To the extent that most of these limiting factors apply rather consistently
over an extended period, they may not seriously impair the usefulness of
building permit statistics as prompt indicators of trends in residential
construction activity. However, the geographic limitations of the data need
to be kept in mind. In addition, the dollar volume of residential construction
should be used with caution. Due to the nature of the building permit
application, we suspect that the valuations may frequently differ from the
true cost of construction. Any attempt to use these figures for inter-area
comparisons of construction volume must, at best, be made cautiously and
with broad reservations.
The particular sample selected for the Building Permits Survey is one of a large number of similar probability samples that, by chance, might have been selected using the same sample design. Each of the possible samples would probably yield somewhat different results. The standard error of a survey estimate is a measure of the variation of all possible survey estimates around the theoretical, complete coverage value. The relative standard error is defined as the standard error divided by the value being estimated.
Estimates of the relative standard errors have been computed from the sample data for selected statistics in these tables and are available in Table A.
Statistics on Counties and Metropolitan Areas are not based on samples.
Although not subject to sampling variability, they are subject to various
non-sampling errors. Explicit measures of their effects generally are not
available, but it is believed that most of the significant response and
operational errors were detected and corrected in the course of the Bureau's
review of the data for reasonableness and consistency.
Estimates of the standard errors have been computed from the sample data for selected statistics in this report. They are presented in the tables in the form of average relative standard errors. The relative standard error equals the standard error divided by the estimated value to which it refers.
The sample estimate and an estimate of its standard error allow us to construct interval estimates with prescribed confidence that the interval includes the average result of all possible samples with the same size and design. For example, suppose the Housing Starts table showed that an estimated 110,000 units in one-unit structures were started in a particular month. Further, suppose that the average relative standard error of this estimate is 3 percent. Multiplying 110,000 by 0.03, we obtain 3,300 as the standard error. This means that we are confident, with 2 chances out of 3 of being correct, that the average estimate from all possible samples of one-unit structures started during the particular month is between 106,700 and 113,300 units. To increase the probability to about 9 chances out of 10 that the interval contains the average value over all possible samples (this is called a 90-percent confidence interval), multiply 3,300 by 1.6, yielding limits of 104,720 and 115,280 (110,000 units plus or minus 5,280 units). The average estimate of one-unit structures started during the specified month may or may not be contained in any one of these computed intervals; but for a particular sample, one can say that the average estimate from all possible samples is included in the constructed interval with a specified confidence of 90 percent.
Ranges of 90-percent confidence intervals for estimated percent changes
are shown in the text. When the range of the confidence interval contains
zero, it is unclear whether there was an increase or decrease; that is,
the change is not statistically significant.
As described in the section, ``Methodology Used for Data up through 1998,'' a potential source of bias prior to 1999 was the upward adjustment of 3.3 percent made to account for one-unit structures started in permit-issuing areas without permit authorization.
Another source is the adjustment for units started prior to permit authorization
and for late reports. The final estimates of privately owned housing units
started are adjusted less than 2 percent for pre-permit starts and late
reports.
The mechanics of seasonal adjustment involve breaking down a time series into 'trend-cycle' and 'seasonal and irregular' components.
The trend-cycle is the long-term tendency of a series to grow or decline.
Seasonal effects are effects that are reasonably stable in terms of timing, direction and magnitude. Possible causes include natural factors (the weather), administrative measures and social/cultural/religious traditions.
The irregular component is anything not included in the trend-cycle or the seasonal effects (including trading-day or holiday effects). Its values are unpredictable with respect to timing, impact, and duration. It can arise from sampling error, non-sampling error, unseasonable weather, natural disasters, strikes, etc.
Monthly time series that are totals of daily activities can be influenced by each calendar month's weekday composition. This influence is revealed when monthly values consistently depend on which days of the week occur five times in the month. For example, building permit offices are usually closed on Saturday and Sunday. Thus, the number of building permits issued in a given month is likely to be higher if the month contains a surplus of weekdays and lower if the month contains a surplus of weekend days. Recurring effects associated with individual days of the week are called trading-day effects.
Trading-day effects can make it difficult to compare time series values or to compare movements in one series with movements in another. For this reason, when estimates of trading-day effects are statistically significant, we adjust them out of the series. The removal of such estimates is referred to as trading-day adjustment.
Most of the seasonally adjusted series are shown as seasonally adjusted annual rates (SAAR). The seasonally adjusted annual rate is the seasonally adjusted monthly value multiplied by 12. The benefit of the annual rate is that not only can we compare one monthly estimate with another, we can also compare monthly data to an annual total. The seasonally adjusted annual rate is neither a forecast nor a projection; rather it is a description of the rate of building permits, housing starts or housing completions in the particular month for which they are calculated.
The seasonal adjustment factors shown in this publication were developed using X-12-ARIMA. The X-12-ARIMA is a seasonal adjustment program developed at the U.S. Census Bureau. The program is based on the Bureau's earlier X-11 program and the X-11-ARIMA/88 program developed at Statistics Canada. For more information on X-12-ARIMA please see the Bureau's X-12 website.
Different procedures are used to develop seasonal adjustments for monthly New Residential Construction series. The differences are necessary due to publication requirements and the lack of identifiable and stable seasonality in certain series (such as two-to-four unit structures for all series except building permits). The procedures are described below.
Housing Units Authorized by Building Permits
Seasonally adjusted annual rates are developed each month for building permits by region and type of structure. Each month, 10 series are run through the X-12-ARIMA program. The seasonally adjusted U.S single-family total is the sum of the seasonally adjusted single-family structures in each of the four regions. The seasonally adjusted U.S. total is the sum of the seasonally adjusted U.S. total single-family, U.S. total for two-to-four unit structures, and U.S. total for structures with five units or more. The totals for each of the four regions are seasonally adjusted and modified so that the seasonally adjusted U.S total derived from the region totals equals the seasonally adjusted U.S. total derived from the structures.
Housing Units Started and Completed
Seasonally adjusted annual rates are developed each month for Starts and Completions by region and type of structure. Each month (separately for Starts and Completions), five series are run through the X-12-ARIMA program. These series are the four regional series for single-family structures and the U.S. total for structures with two units or more. The seasonally adjusted U.S. total is the sum of the five seasonally adjusted components.
The seasonally adjusted U.S. total for structures with five units or more is the product of the seasonally adjusted U.S. total for structures with two units or more and the ratio of the unadjusted U.S. total for structures with five units or more to the unadjusted U.S. total for structures with two units or more. Each seasonally adjusted regional total is the sum of the seasonally adjusted single-family regional total and the seasonally adjusted regional total for structures with two units or more (adjusted using the seasonal factor for the U.S. total for structures with two units or more). The seasonally adjusted regional totals add up to the seasonally adjusted U.S. total without any modifications.
Housing Units Under Construction
The seasonally adjusted numbers of units under construction are developed by region and type of structure. These estimates are computed using a procedure similar to that used for the monthly seasonally adjusted series for Starts and Completions.
The X-12-Arima program that we use to derive our seasonal adjustment calculates numerous diagnostic statistics. The table below shows some of these statistics. For a description of the diagnostics published in this table, please click here.
Average Percent Changes and Related Measures for Monthly New Residential Construction
Series | ||||||||
Average percentage change | Ratio of irregular to trend-cycle component | Moving seasonality present relative to stable seasonality | Overall quality assessment statistic | F-test statistic for stable seasonality | ||||
Original series | Seasonally adjusted series | Irregular component | Trend-cycle component | |||||
(O) | (CI) | (I) | (C) | (I/C) | (M7) | (Q) | (F) | |
HOUSING UNITS AUTHORIZED BY BUILDING PERMITS | ||||||||
U. S. total | 10.85 | 3.60 |
3.28 |
1.09 |
3.01 |
0.18 |
0.67 |
147.77 |
Northeast | 15.14 |
6.31 |
5.79 |
1.83 |
3.17 |
0.16 |
0.39 |
287.35 |
Midwest | 17.34 |
4.91 |
4.51 |
1.30 |
3.48 |
0.14 |
0.32 |
407.19 |
South | 9.86 |
4.69 |
4.40 |
1.15 |
3.82 |
0.26 |
0.50 |
72.71 |
West | 11.58 |
5.53 |
5.06 |
1.42 |
3.56 |
0.20 |
0.38 |
140.10 |
1 unit: | ||||||||
Northeast | 15.48 |
5.60 |
5.28 |
1.40 |
3.78 |
0.15 |
0.37 |
328.94 |
Midwest | 18.36 |
4.37 |
3.76 |
1.55 |
2.43 |
0.19 |
0.27 |
553.06 |
South | 10.01 |
3.23 |
2.87 |
1.12 |
2.56 |
0.16 |
0.27 |
278.44 |
West | 11.65 |
4.23 |
3.70 |
1.52 |
2.44 |
0.12 |
0.25 |
316.50 |
2 to 4 units | 15.21 |
8.57 |
8.52 |
0.89 |
9.54 |
0.25 |
0.86 |
73.86 |
5 units or more | 16.17 |
10.40 |
10.24 |
1.34 |
7.61 |
0.42 |
0.95 |
30.50 |
HOUSING UNITS STARTED | ||||||||
U. S. total | 10.00 |
4.88 |
4.55 |
1.14 |
3.99 |
0.24 |
0.61 |
149.25 |
1 unit: | ||||||||
Northeast | 18.20 |
10.08 |
9.97 |
1.30 |
7.69 |
0.27 |
0.93 |
72.50 |
Midwest | 19.85 |
8.96 |
8.78 |
1.03 |
8.51 |
0.21 |
0.65 |
168.62 |
South | 9.36 |
5.90 |
5.78 |
0.92 |
6.29 |
0.29 |
0.68 |
71.45 |
West | 11.52 |
8.34 |
8.18 |
1.04 |
7.89 |
0.30 |
0.74 |
54.53 |
2 units or more | 17.62 |
15.63 |
15.37 |
1.53 |
10.04 |
0.66 |
1.39 |
12.95 |
HOUSING UNITS UNDER CONSTRUCTION | ||||||||
U. S. total | 1.82 |
0.85 |
0.53 |
0.65 |
0.82 |
0.32 |
0.25 |
85.20 |
1 unit: | ||||||||
Northeast | 2.63 |
1.53 |
1.19 |
0.79 |
1.51 |
0.20 |
0.38 |
149.98 |
Midwest | 3.40 |
1.36 |
0.82 |
0.89 |
0.92 |
0.19 |
0.20 |
311.36 |
South | 2.20 |
1.12 |
0.74 |
0.79 |
0.93 |
0.11 |
0.09 |
391.65 |
West | 2.49 |
1.43 |
1.07 |
0.80 |
1.34 |
0.14 |
0.17 |
234.66 |
2 units or more | 1.87 |
1.70 |
1.03 |
1.27 |
0.82 |
0.82 |
0.44 |
11.17 |
HOUSING UNITS COMPLETED | ||||||||
U. S. total | 8.57 |
4.81 |
4.67 |
0.82 |
5.68 |
0.24 |
0.69 |
88.97 |
1 unit: | ||||||||
Northeast | 14.48 |
9.93 |
9.96 |
1.08 |
9.21 |
0.42 |
1.07 |
28.21 |
Midwest | 11.31 |
6.98 |
6.90 |
0.99 |
6.97 |
0.24 |
0.67 |
90.27 |
South | 10.15 |
5.69 |
5.70 |
0.82 |
6.92 |
0.32 |
0.73 |
45.57 |
West | 11.05 |
7.80 |
7.75 |
1.06 |
7.35 |
0.39 |
0.89 |
28.68 |
2 units or more | 15.53 |
11.93 |
11.82 |
1.42 |
8.30 |
0.47 |
1.10 |
22.32 |
For further information on time series and seasonal adjustment, please refer to the Manufacturing and Construction Division's frequently asked questions.
For additional information, call the Residential Construction Branch. Telephone: 301-457-1321.
Source: U.S. Census Bureau
Last revised: May 18, 2004