Best Time of Year to Build: Monthly Workload Index for US Commercial Construction and State Seasonal Swing Ranking

How the index is built

Download the calculations (CSV)

How the monthly workload index is built

The index measures how far each calendar month sits above or below an average month, using raw (not seasonally adjusted) federal data. 100 means a normal month.

Seasonal index (spending): For each full year, each month's private nonresidential spending put in place is divided by that year's total and multiplied by 12 x 100. The twelve monthly values are then averaged across years. A value of 100 equals one twelfth of the year.

Seasonal index (employment): Employment is a level, not a flow, so each month's payroll count is divided by that year's average monthly level and multiplied by 100, then averaged across years.

Seasonal amplitude: (peak month index / trough month index – 1) x 100. It reads as the percentage by which the busiest month exceeds the slowest month in a typical year.

Spending comes from the Census Bureau not seasonally adjusted private construction table (release of 2026-10-01, data through August 2026). Census describes the survey as measuring construction work done each month, with sampled projects reporting progress monthly (methodology). The primary window is 2014 to 2025 excluding 2020; 2020 is tested separately because its pattern broke from every other year (lowest month December, highest June).

Employment comes from BLS flat files: national construction payrolls (CEU2000000001), nonresidential building (CEU2023620001) and nonresidential specialty trades (CEU2023800201), and statewide construction payrolls from the State and Area program. The employment window is the latest ten complete years, 2016 to 2025, with 2020 excluded.

Best time of year to build: what the national monthly index shows

Private nonresidential work put in place climbs from a January low (88.8) through spring, stays above average from May through December, and peaks in October (106.3). Payrolls follow the same arc with a smaller swing.

Across 2014 to 2025 excluding 2020, January's spending index is 88.8 and February's 89.6. March (94.7) and April (98.0) are still below average. The index crosses 100 in May (100.9) and stays above it through December (101.6).

The first quarter averages 91.0, the second 100.7, the third 104.3 and the fourth 104.0. On this measure, the last three months of the year carry the heaviest dollar volume of private nonresidential work in place.

Construction employment peaks earlier than spending. Payrolls reach index 103.0 in July and 102.9 in August, while spending is highest in October. Average weekly hours follow employment: 38.1 hours in January against 39.7 in August, 2016 to 2025 excluding 2020.

For an owner reading this as workload data: the January and February trough means fewer workers on payrolls, shorter weeks and less work in place. The summer and early autumn plateau means more crews already committed. Bid and start timing still depend on project-specific factors that these national data do not capture; see how to compare construction bids for bid-side checks.

State ranking: how strong the seasonal swing is in each state

The seasonal swing in construction employment ranges from 39.7% in Alaska to 1.7% in Florida. Cold-winter states sit at the top of the ranking and the Sun Belt at the bottom.

All ten widest swings are in states with a January average at or below 31.4 F: Alaska, North Dakota, Minnesota, South Dakota (27.7%), Iowa (26.1%), Vermont (23.0%), Montana (22.9%), West Virginia (22.0%), Wyoming (21.8%) and Illinois (20.4%).

The largest construction markets sit lower in the ranking. Texas swings 2.9%, California 5.9%, Florida 1.7% and New York 11.8%. In job terms, Illinois added 44.7 thousand construction jobs between its 2025 low and high, Texas 42.8 thousand and Minnesota 37.9 thousand.

Peak and trough month by state

Every state's construction payroll bottoms in January or February. Most peak in July; in warm states the curve is so flat that the peak month carries little meaning.

Of 51 states and DC, 27 have their lowest month in January and 24 in February. February troughs are concentrated in the Northern Plains, Upper Midwest and Northeast, where winter work stays depressed for longer.

Peaks fall in July in 28 states, August in 8, June in 7, October in 5, September in 2 and November in 1. The June and October peaks belong to Southern states such as Georgia, South Carolina and Arizona (June) and Texas, Louisiana and Alabama (October).

In those states the gap between the peak and the neighbouring months is small. Florida's index runs from 98.7 in January to 100.5 in November, and Texas's from 97.9 in January to 100.8 in October. A Minnesota contractor, by contrast, carries 111.5 in July against 84.7 in February.

Winter temperature explains most of the state differences

A state's January average temperature predicts its seasonal swing closely (r = -0.86).

Using NOAA Climate at a Glance statewide January temperatures averaged over 1991 to 2020, the Pearson correlation with state seasonal amplitude is -0.86 and the rank correlation -0.9 across 49 states (no statewide series exists for Hawaii or DC).

The 13 states with a January average below 25 F average a 23.7% swing. The 7 states at 45 F or warmer average 3.1%. Alaska (4.2 F) has the widest swing; Florida (58.3 F) is the warmest and ranks near the bottom.

Temperature does not explain every case. West Virginia (31.4 F) swings 22.0%, more than colder Maine (15.1% at 15.2 F). These data do not show what drives such exceptions.

Sensitivity test: windows, 2020 and method

January stays the low month and October the spending high in every window and method tested. The size of the swing depends on the method and has narrowed since 2020.

Window: the spending swing was 21.5% in 2014 to 2019 and 17.6% in 2021 to 2025. The employment swing fell from 10.0% to 7.1%. Including 2020 changes the spending swing to 18.2% and the employment swing to 7.7%. The calendar-year employment method, which keeps within-year growth, gives 9.9%.

Method: the share-of-year method mixes trend into the seasonal pattern. Private nonresidential spending grew 116.7% from 2014 to 2025, so in a growing year the later months look larger. A ratio to a centered 12-month moving average removes most of that trend and gives a 12.9% swing. Census's implied factors for 2023 to 2025 give 10.4%. The share-method 19.7% should be read as an upper bound.

States: 48 of 51 swings narrowed between the two windows, and the median fell from 11.9% to 9.3%. Rankings barely moved (rank correlation 0.98), so the states that were most seasonal before 2020 still are.

Single years vary. January was the lowest spending month in 8 of 12 years. In 2025 the spending curve was unusually flat (swing 8.6%), with December the low month, while 2022 had a 32.3% swing.

Components: building types, hiring and labor slack

Every major private nonresidential building type bottoms in January or February, but the swing ranges from 9.0% for power projects to over 30% for communication, transportation and educational work. Hiring and worker availability move ahead of work in place.

Educational construction is the one category that peaks in June (30.8% swing), matching the school summer break. Manufacturing, the largest category at $213,820 million in 2025, swings 23.1% and peaks in October. Power, the second largest, has the smallest swing at 9.0%.

Labor indicators lead work in place. Construction hires (JOLTS, not seasonally adjusted) average 472 thousand in April and 207 thousand in December, 2.3 times as many. The unemployment rate of experienced construction workers averages 7.6% in January and February against 4.0% in August and September. AGC noted in February 2026 that construction unemployment normally rises in winter months (AGC). For the structural side of labor supply, see the construction labor shortage tracker.

Do construction prices change by season?

Not in a warm-versus-cold pattern for finished buildings. Material input prices have tended to rise more in the first half of the year.

The PPI for new nonresidential building construction (WPU801) changes mostly in the first month of each quarter: a mean of 1.13% in January, April, July and October against 0.09% in the other months, 2016 to 2025 excluding 2020. That is a calendar step, not a weather effect: January to June averages 0.4% a month and July to December 0.47%. The BLS series catalog carries no seasonally adjusted version of WPU801 (wp.series).

Input goods (WPUIP2312001) rose a mean 0.93% a month from January to June and -0.08% from July to December. Excluding the 2021 to 2022 price surge, the gap shrinks to 0.44% against -0.04%. The tendency is real but small next to year-to-year price moves. Price levels and trends are covered in the commercial construction price index.

What the index cannot tell you

The index describes a typical year at national and state level. It does not measure local backlog, bid prices or availability of a specific contractor.

• The Census monthly spending table used here is national; it has no state detail.
• State figures cover all construction (residential, nonresidential and civil), not commercial building alone.
• Employment refers to the pay period including the 12th of each month (BLS); it does not measure crew idle time or backlog.
• The share method overstates the swing in growing years; the moving-average and Census factor checks give smaller values.
• Recent months (2026) are preliminary and are excluded from the averages, except as inputs to the moving average at the end of 2025.
• This report describes data patterns and is not legal, financial or project-planning advice.

Key findings

1. January is the low month for US private nonresidential construction spending put in place, at 11.2 index points below an average month (index 88.8), 2014 to 2025 excluding 2020.^[1]
2. October is the high month for US private nonresidential spending (index 106.3), with August to October all above 104.7, 2014 to 2025 excluding 2020.^[1]
3. The fourth quarter averages an index of 104.0 and the first quarter 91.0 for US private nonresidential spending, 2014 to 2025 excluding 2020.^[1]
4. Measured per calendar day, February's US spending index rises by 7.2 points to 96.8, so part of the February dip is the short month, 2014 to 2025 excluding 2020.^[1]
5. Census's own seasonal factors for US private nonresidential spending, 2023 to 2025, put January at 93.9 and October at 103.7, a 10.4% swing, smaller than the share-method 19.7%.^[1][2]
6. US construction employment (not seasonally adjusted) averages index 95.2 in January and 103.0 in July, a 8.2% swing, 2016 to 2025 excluding 2020.^[5]
7. US nonresidential specialty trade employment swings 7.3% and nonresidential building employment 5.5% between January and the summer peak, 2016 to 2025 excluding 2020.^[5]
8. In 2025, US construction payrolls (not seasonally adjusted) rose from 7,938 thousand in January to 8,476 thousand in July, a gain of 538,000 jobs (6.8%).^[5]
9. Alaska (39.7%), North Dakota (37.8%) and Minnesota (31.7%) have the widest seasonal swings in state construction employment, 2016 to 2025 excluding 2020.^[7]
10. Florida (1.7%), Arizona (2.3%) and Georgia (2.8%) have the narrowest seasonal swings in state construction employment, 2016 to 2025 excluding 2020.^[7]
11. Construction employment troughs in January in 27 of 51 states and in February in 24; it peaks in July in 28, in August in 8 and in June in 7, 2016 to 2025 excluding 2020.^[7][8]
12. States with a January average below 25 F (1991 to 2020) average a 23.7% seasonal swing, 7.6 times the 3.1% average of states at 45 F or warmer.^[7][18]
13. The median state swing fell from 11.9% in 2014 to 2019 to 9.3% in 2021 to 2025, while state rankings stayed stable (rank correlation 0.98).^[7]
14. US construction hires (not seasonally adjusted) average 472 thousand in April and 207 thousand in December, 2016 to 2025 excluding 2020.^[11]
15. The unemployment rate of experienced US construction workers averages 7.6% in January and 4.0% in August, 2016 to 2025 excluding 2020 (not seasonally adjusted).^[15]

Summary for decision makers

• US private nonresidential construction spending put in place is lowest in January (index 88.8) and highest in October (index 106.3), using Census not seasonally adjusted data for 2014 to 2025 excluding 2020.
• January was the lowest spending month in 8 of the 12 years from 2014 to 2025; the high point moved between June and November, so the busy season is a plateau rather than a single month.
• US construction employment (not seasonally adjusted) troughs in January (index 95.2) and peaks in July (index 103.0), a swing of 8.2% over 2016 to 2025 excluding 2020; in 2025 the swing was 538,000 jobs.
• State swings range from 39.7% in Alaska and 37.8% in North Dakota to 2.3% in Arizona and 1.7% in Florida; the median state swing is 10.1%.
• Winter temperature tracks the state gap closely: the correlation between January average temperature and seasonal swing is -0.86 across 49 states, and the coldest 13 states average a 23.7% swing against 3.1% for the 7 warmest.
• Seasonality has narrowed since 2020: 48 of 51 state swings were smaller in 2021 to 2025 than in 2014 to 2019, and the national employment swing fell from 10.0% to 7.1%.
• The output price index for new nonresidential buildings shows steps in January, April, July and October (mean change 1.13% vs 0.09% in other months) but no warm or cold season pattern (0.4% in January to June vs 0.47% in July to December).

Frequently asked questions

Related guides and services

• How to Compare Construction Bids Without Getting Burned

Methodology

Source selection: Tier 1 federal datasets only for every number in the index: Census Construction Spending (C30) not seasonally adjusted private series, BLS Current Employment Statistics national and State and Area flat files, BLS JOLTS and Current Population Survey series for construction, BLS PPI flat files, and NOAA NCEI Climate at a Glance statewide temperatures. Trade association material (AGC, ABC) is used only for headline context and is marked Tier 2.

Inclusion and exclusion: complete calendar years only. The spending index uses 2014 to 2025; the employment and state indexes use the latest ten complete years, 2016 to 2025. 2020 is excluded from primary results and tested separately. 2026 months are preliminary and are reported only as recent context. October 2025 Current Population Survey data were not collected because of the 2025 lapse in appropriations and are skipped.

Derived metrics: monthly spending index = mean of each month's share of the annual total x 12 x 100; employment index = mean of each month's level divided by a centered 2×12 moving average, rescaled to average 100 (a calendar-year version, level divided by the annual mean, is shown as a check); amplitude = (peak index / trough index – 1) x 100. Spending checks: a per-calendar-day index, a ratio to a centered 2×12 moving average, and the Census implied factor (not seasonally adjusted value / (seasonally adjusted annual rate / 12)). All calculations are in scripts/compute.py and derived_metrics.csv.

Conflicting numbers: the share method, moving-average method and Census factors give different swing sizes (from 10.4% to 22.5% for spending). The report states the share method as the brief's primary measure, labels it an upper bound, and shows the alternatives next to it. Seasonally adjusted figures quoted by trade groups are not mixed with unadjusted figures except to show the gap between them.

Limitations: monthly construction spending is not published by state, so the state ranking uses employment for all construction, not commercial building alone. Delaware, the District of Columbia and Hawaii use mining, logging and construction. Employment measures jobs in the pay period including the 12th, not hours worked or backlog. Climate correlation does not establish cause.

Last updated 2026-10-07 using the Census release of 2026-10-01 (data through August 2026), BLS national data through September 2026, BLS state data through August 2026, JOLTS through August 2026 and PPI through August 2026. Not advice: the report describes historical data patterns and does not recommend when any project should bid or start.

This report presents published data with its limitations. It is not legal, financial or investment advice.

Sources, ranked by quality

Tier 1: primary and official sources (20)

• [1] U.S. Census Bureau. Construction Spending: Value of Private Construction Put in Place, Not Seasonally Adjusted (privtime.xlsx), October 1, 2026 release. January 1993 to August 2026 (2014 to August 2026 used). Accessed 2026-10-07.
• [2] U.S. Census Bureau. Construction Spending: Value of Private Construction Put in Place, Seasonally Adjusted Annual Rate (privsatime.xlsx), October 1, 2026 release. January 1993 to August 2026. Accessed 2026-10-07.
• [3] U.S. Census Bureau. Construction Spending: Value Put in Place Methodology. current methodology. Accessed 2026-10-07.
• [4] U.S. Census Bureau. Construction Spending: Release Schedule. 2026 to 2027 release dates. Accessed 2026-10-07.
• [5] U.S. Bureau of Labor Statistics. Current Employment Statistics flat file ce.data.20a.Construction.Employment. through September 2026. Accessed 2026-10-07.
• [6] U.S. Bureau of Labor Statistics. Current Employment Statistics flat file ce.data.20b.Construction.AllEmployeeHoursAndEarnings. through September 2026. Accessed 2026-10-07.
• [7] U.S. Bureau of Labor Statistics. State and Area Employment flat file sm.data.62.Construction.Current. January 2007 to August 2026. Accessed 2026-10-07.
• [8] U.S. Bureau of Labor Statistics. State and Area Employment flat file sm.data.61.MiningLoggingConstr.Current. January 2007 to August 2026. Accessed 2026-10-07.
• [9] U.S. Bureau of Labor Statistics. CES State and Area: Seasonal Adjustment methodology. current methodology. Accessed 2026-10-07.
• [10] U.S. Bureau of Labor Statistics. JOLTS flat file jt.data.2.JobOpenings (series JTU230000000000000JOL). December 2000 to August 2026. Accessed 2026-10-07.
• [11] U.S. Bureau of Labor Statistics. JOLTS flat file jt.data.3.Hires (series JTU230000000000000HIL). December 2000 to August 2026. Accessed 2026-10-07.
• [12] U.S. Bureau of Labor Statistics. Producer Price Index flat file wp.data.80.Construction (series WPU801). June 2009 to August 2026. Accessed 2026-10-07.
• [13] U.S. Bureau of Labor Statistics. Producer Price Index flat file wp.data.80i.ConstructnInputs (series WPUIP2312001). June 2010 to August 2026. Accessed 2026-10-07.
• [14] U.S. Bureau of Labor Statistics. Seasonal Adjustment in the PPI. current methodology. Accessed 2026-10-07.
• [15] U.S. Bureau of Labor Statistics. Unemployment Rate, Construction Industry, Private Wage and Salary Workers, not seasonally adjusted (LNU04032231). January 2016 to December 2025 (October 2025 not collected). Accessed 2026-10-07.
• [16] U.S. Bureau of Labor Statistics. Schedule of Releases for the Employment Situation. 2026 schedule. Accessed 2026-10-07.
• [17] U.S. Bureau of Labor Statistics. Schedule of Releases for the State Employment and Unemployment (Monthly). 2026 schedule. Accessed 2026-10-07.
• [18] NOAA National Centers for Environmental Information. Climate at a Glance: Statewide Time Series, January average temperature, 1991 to 2020 (Florida file shown; one file per state). January 1991 to January 2020. Accessed 2026-10-07.
• [21] U.S. Bureau of Labor Statistics. Job Openings and Labor Turnover Survey home page (August 2026 release, September 29, 2026). August 2026. Accessed 2026-10-07.
• [22] U.S. Bureau of Labor Statistics. Producer Price Index series catalog wp.series. October 2026 catalog. Accessed 2026-10-07.

Tier 2: trade bodies, market research, company data (2)

• [19] Associated General Contractors of America. Construction Employment Increased By 33,000 In January As Nonresidential Hiring Drives Modest Growth (February 11, 2026). January 2026. Accessed 2026-10-07.
• [20] Associated Builders and Contractors. Weather, Additional Issues Lead to 3.2 Percent Decline in Construction Backlog Indicator. First quarter 2015. Accessed 2026-10-07.