Home » Posts tagged 'Jobs'
Tag Archives: Jobs
Construction Briefs NOV 2025
After some delayed data released recently, we now have August Construction Spending and Sept Jobs.
In the past 12mo, Rsdn construction jobs fell 46k (-1.4%). Nonres Bldgs jobs increased 59k (+1.6%) and Nonbldg jobs increased 24k (+2.1%). Rsdn spending is down 5% (-$39bil) Nonres Bldgs is down 3.4% (-$25bil) Nonbldg is up 3.1% (+$15bil) Expect total spending in 2025 down 2.1%, jobs UP 1%
Construction Spending Forecast Total spending varies less than 1% from current through 2026 Jobs YTD up 16k. Only times job growth that slow was 2020 or recessions. 2011 through 2024, even with losses in 2020, avg jobs growth was 200k/yr. Don’t expect job openings (see JOLTS) in near future.
Environment for construction jobs looking difficult. Constant $ spending in 2026 is down just less than 1%. But Volume of work (spending minus inflation) available is down just over 4% and is declining all through 2026. Biggest declines by far, Manufacturing and SF Rsdn.
The decline in Manufacturing construction spending is due to having passed the peak in the scheduled project timelines for the large volume of mega-projects that started in 2021-22-23-24. (I wrote about it in more detail in Nov’24 in the article linked.) Peak spending is typically just past the midpoint of project construction. From Apr 2024 to Nov 2024, Mnfg spending averaged $240bil., the highest rate of spending on record. In 2025 it started the year at a rate of $$230bil but will end the year at $210bil. By the end of 2026 the rate of spending drops to $190bil.
The Manufacturing Spending Taper
My forecast has not changed much overall in the last few months. Residential has gained in revisions added to June and July and Aug posted a very strong 1.3% gain.
We will wait a little longer before we see any meaningful changes in construction materials input costs. September data (released 11-26-25) reported here. Also remember, PPI does not track imports, only domestic producers. Therefore, any implied increase in PPI being related to tariffs would be a domestic reaction to an import tariff. We can expect that.
INPUTS thru Sept up ~2% from avg2024. Final Demand for Nonres Bldgs is up 1.2% ytd vs avg 2024. However, Oct is the revision month for Q3 Final Demand data, so Final Demand data not finalized for Q3.
Construction Briefs Aug 2025
The biggest story in construction data right now is jobs.
Average construction jobs growth through July, last 25 years, excluding recessions, +130,000.
Average construction jobs growth through July, last 10 years, excluding recessions, +140,000.
2025 Construction Jobs growth through July, +21,000.
Not so surprising, as the Constant $ construction spending through July is down -5.7%, (compared to same months previous year), steepest decline since 2011, which was the end of the great recession. Constant $ (inflation adjusted) construction spending is now back to early 2022 level.
Construction Spending is down 5 of the last 6mo, now down a total -3.0% from Dec. Over that period spending is down most significantly in Residential, Manufacturing and Commercial w/o Warehouse. It’s up the most in Data Centers, Highway and Public Utilities. Overall, business is declining.
Construction Spending inflation adjusted is forecast to drop slightly every month for the rest of the year. Expect constant $ spending at year end down -6.3% from 2024. Uncertainty over tariffs and funding subsidies has slowed decision making on planning and moving new projects forward to construction. This is not an environment to expect jobs growth.
If jobs were to move at the same rate as business volume, with 2025 construction spending in constant$ expected to fall -6.3%, then jobs would be expected to fall -6.3%. That’s 500,000 jobs. The only times we’ve ever lost 500,000 jobs in a year was in both 2009 and 2010. In those years, after falling 17% in the previous 3 years, business volume dropped another 12% and 10% respectively.
Residential construction jobs peaked last September and are now down 1% or 35,000 jobs since then. In Constant $, residential spending is down 10% since last September. Jobs never move at the same rate as spending. This has a significant impact on productivity.
Data Centers are the bright spot in construction spending, up 17% since December and forecast to finish the year up 30%, an increase of +$10bil.
Biggest forecast declines in Current$ construction spending: Residential -5.8% (by far largest $ decline, -$55bil); Manufacturing – 6.5%, -$15bil; Warehouse -10%, -$7bil; Office (ex Data Centers) -10%, -$7bil; Comm Retail (ex Warehouse) -8%, -$6bil.
Manufacturing spending is now receding from an astronomical high. From 2019 through 2021 spending was averaging $80bil/yr. In 2024 it reached an average of $235bil, and peaked in Oct at $244bil. In June it was only $223bil and it’s expected forecast for 2025 is down -6.5% from 2024, but that is still a very high $220bil.
Construction Forecast Update May 2025
The total construction spending forecast, now at $2,237bil, +3.7% vs. 2024, has been lowered a bit since the Outlook in Feb. ($2,272bil, +5.5%). Most of the reduction is in Residential, from $997bil, +7.2% down to $958bil, +3.0%. Nonres Bldgs was reduced by $11bil and Nonbldg increased by $7bil. Compared to the average for the year 2024, current total spending YTD is up 2.6%, but is expected to pick a little up throughout the year across all sectors.
Constant $ growth is forecast down 0.7% in 2025 and down more after that. Inflation Index shows annual percent and index to base 2024 = 100. All years, current spending / index = reported as constant 2024$, as if all constant years are the same 2024$.
Residential Single Family spending has been in a range +/- 2.5% for the last 7 months. With only a few months lower than that, this range extends back 18 months. Take out 3%/yr inflation and you can see that SF volume has been flat to down. Multi-family spending, for the last 12 months, is down 15% off recent highs in Q1’24. Reno/repair spending extended recent highs across 5 months in mid 2024, before falling off 10% in Q4, and now, in Q1’25 it has gained back most of that drop. Any spending gains in the near future would be driven by multi-family.
Manufacturing is still the largest $ contributor ($222bil/yr) to nonresidential bldgs total spending ($772bil/yr), but has fallen 6% in the last 5 months. The trend is down in most months for 2025 and down in 2026. This decline is entirely expected and you can read about it in my article, The Manufacturing Taper.
Data Centers shows the largest % spending growth for 2025, forecast +39%, continuing on a phenomenal streak of +45% in 2023 and +56% in 2024. Data Center starts increased 300% over the previous 3 yrs. and still increase in 2025 and 2026, but at a slower rate of growth. Manufacturing spending peaked in Oct’24. Data Center spending continues to increase for the next few years. Both of these forecasts take into account some cancelations or delays announced recently (see May Briefs), however do not account for any major stoppages due to recent trade impacts.
The largest $ increases in spending are Power, forecast to increase +$16bil, and Highway +$13bil. Educational spending is forecast to post the largest nonres bldgs $ growth in 2025 (+$12bil, +9%). Data Centers increase $10bil.
This same scenario that looks to occur in Manufacturing will occur also in Highway/Bridge. Normal starts were about $100bil/yr, with slow growth. But for the last 3 years, actual starts totaled closer to $500bil for the 3 years or $167bil/yr. This strong growth in starts is expected to continue at least into 2025, totaling near $650bil for 4 years. Again, consider that part of that is inflation, but the remainder is government investment growth. So a decline from the taper back to normal for Highway/Bridge may not show up at least for the next few years. But once the taper begins it will have the same effect on Nonbldg Infrastructure spending that we will see from Mnfg in Nonres Bldgs.
Typical jobs growth is 2.5% to 3.5% per year, even though spending can sometimes far outpace that. But jobs growth doesn’t track spending, it tracks volume growth. Volume is spending minus inflation. For example, for the 4 years 2021 thru 2024, spending increased 44%. But inflation increased 33%. Volume growth was only 11%. Jobs increased 12.5%. For 2025, Inflation will outpace spending growth by 1%, but jobs are still expected to increase by 1.5%. Rarely do jobs decrease. That could change.
There are 870K construction jobs in TX. 500K are immigrants and 300K are undocumented.
New starts are forecast up in 2025, but I would caution there are a lot of headwinds that could slow new starts growth. Many economists predict the current trade impacts will slow overall economic growth. That in turn could slow capital expenditures, which, in this case, is new construction starts. So far year-to-date, spending is outpacing new starts growth. That means backlog is decreasing, mostly in nonresidential buildings. We haven’t seen a decrease in Nonres Bldgs backlog since pre-2011. Residential backlog is down slightly. Nonbuilding is increasing.
See Also Construction Briefs May 2025
and Construction Briefs Apr 2025
5-15-25 all ppi DATA UPDATED TO APR See Construction Inflation 2025
I’ve increased the inflation outlook since Feb to a range of 4.5% to 5%. Inflation may be the most uncertain of all factors affecting construction this year. We have yet to see any significant impact from tariffs, and there is sure to be impacts to many construction inputs. We may not know the total impact for several more months. But I would expect, if anything, inflation will go up from here, not down.
The bees are swarming the flowers outside. At least some lifeform is content and at peace with this world.
Construction Briefs May 2025
For the 9th consecutive year, I will be speaking at Advancing Preconstruction. I will be opening the program May 22 to the plenary session with a summary of the current and expected economic conditions affecting everyone involved in construction, all geared towards one word, RISK.
Construction Spending Q1’25 vs Q4’24 notable Q/Q increases: Education, Healthcare, Amusement/Recreation and Communication are all up 2% to 3%. Highway is up+4.9%, Data Centers +5.4%, Warehouse +7.5% and Lodging +8.3%.
Construction Spending for March is down 0.5% from Feb, but that’s because Feb was revised UP by 0.5%. Jan also revised up 0.66%. YTD Total vs Jan-Mar 2024 is up 2.8% YTD. Data Centers vs Jan-Mar 2024 is up 40%.
Construction Spending Q1’25 vs Q4’24 is UP in every category except Residential, Commercial/Retail w/o Warehouse and Manufacturing (Mnfg was expected). Residential and Comm/Rtl are down only a slight 0.2% and 0.4%. Manufacturing is down 4.7% Q1vQ4. This is the beginning of the Manufacturing spending taper as early projects come to an end. I described that taper here. The Manufacturing Spending Taper
Not seeing any major indications in spending due to tariffs yet. Still early in the data (thru Mar) for that.
Construction Jobs increased 11,000 in April. However, hours worked dropped by 0.6%. Total workforce hours worked declined by an equivalent of 50,000 jobs. Jobs are now at 8,316,000, an all-time high. Jobs are up 27k year-to-date, the slowest growth for the 1st 4 months since 2012 (excld 2020). Although hours worked fell in April, total workforce hours worked increased 2.1% over same 4mo 2024. Average yr/yr growth for Jan-Apr hours worked is 3.7% for the last 10 yrs (ex 2020).
J P Morgan expects imports from China to fall 75%-80% in the 2nd half of the year. Total all imports from all sources are expected down 20%. Some products are going to become unavailable.
The U.S. imports about 30 million metric tons, about 30% of total steel used, of all types of steel annually. The U.S. imports about 6 million metric tons of steel pipe annually. Approx 2/3rds of steel pipe used annually in the U.S. is imported. If the U.S. loses its imports of steel pipe, we can’t support as many building projects. Pipe here refers to pipe and tube. That includes things like gas and oil pipelines, water pipe, steel conduit and structural square/rectangular tube sections (Trump’s Wall).
What’s frustrating this week is all the latest construction spending and jobs data just came out, and everyone wants to know, What’s the impact on the forecast?, and none of the data reflects tariff impacts or potential slowdowns. Spending is thru Mar31 and jobs are thru Apr12. Some of the inflation data is 1 to 2 quarters behind.
I am expecting, when I prepare the Midyear Forecast, that spending projections will go down, perhaps 1% to 3%, and inflation projections will go up. Currently, I’m carrying inflation between 4%-5%. Owner’s may slow or even cancel capital expenditures and material prices are broadly expected to increase.
When PPI data is released May12, that will be thru April. But remember, PPI data is domestic products only. So any inflation in the PPI data is domestic suppliers adjusting pricing to reflect pricing similar to expected increases to match imports. We might begin to see our first clues of tariff impacts/demand when the next construction starts data gets released around the end of May. How much in previous starts have been canceled/delayed? We already know of some chip plants and data centers canceled/delayed.
Construction – What to Watch: Cost to build going up; Cost to finance is up; Product availability in question; Product delivery schedule delays; Margins pressured; Small/Midsize firms squeezed; Labor let go/disappearing; Projects in planning, delayed; Project ROI not met; Projects planned, canceled.
I recommended (going back 6 yrs ago, but still relevant today) that every construction cost estimator is going to need to identify in every estimate/budget presented to an owner for every upcoming project, all items subject to price revision due to tariff. If you don’t you stand to lose your already meager profits.
I can’t even begin to know what to tell construction cost estimators to carry in budgets for increased cost due to tariffs and supply issues. Best I could suggest at this time is to carry an agreed allowance (IMO, better than contingencies), which can be visited at a later date and adjusted to actual cost. Contingencies are for unknown, unexpected, unidentified issues. Allowances are described in the basis of estimate for identified cost issues, but at unknown cost amounts. All allowances in any estimate/budget should be identified at conception with intent to revisit at later date to adjust to actual cost. (The most common allowance you may be familiar with is a rock allowance). Identify allowances up front and reach agreement on budgeted cost with all parties. This will make your contract administration go a lot smoother than trying to negotiate how much of the contingency you can use for a cost increase that was foreseen. The only unforeseen here is actual cost.
ABI – DMI – CBI Leading Construction Indicators
With exception of residential, which has short durations and for which backlog is always only about 30%-35% of previous yr revenues, for all other work, never (since 2010) was backlog shown to be less than the previous yr spending. https://edzarenski.com/2021/05/01/abi-dmi-cbi-leading-indicators/
Construction Backlog, all work under contract yet to be put-in-place, usually extends out 2 to 3 years. Backlog changes only IF new starts are greater than spending in the month, backlog goes UP. If new starts are less than spending, backlog goes DOWN. Subtract canceled projects from starts causes backlog to go down, but delays are are just moved out in time, so are still in backlog.
PPI INPUTS Q1 vs avg 2024: to Nonres Bldgs +0.9%, to Residential +1.15%, to Highway +1.0%. All these being near 1% for Q1, if growth is constant, would be near 4% for the year. Big IF! Paving mixtures +11% in Q1, Lumber Plywood +4.5%, Fab Str Steel +0.03%, Fab Str Stl Bridges -1.1%, #2 Diesel Fuel -9.6%, Steel Pipe and Tube -3.85%, Nonferrous Wire and Cable +1.8%, Copper and Brass Mill Shapes +4.7%, Aluminum Mill Shapes +7.5%.
PPI Final Demand 1st 3mo vs avg 2024: Avg Nonres Bldgs +1.3%, Educational +1.6%, Healthcare +2.7%, Roofing Contractor + 2.8%, Avg 4 trades +1.7%. Your monthly reminder, although this index is posted monthly, it is corrected quarterly. April data is the correction month for Q1.
New home construction costs have risen about 3% in the last year, from lumber down 4% to concrete up 6%, per JBREC. The US Census Constant Value Rsdn Index is up 3.5% for the 1st 3 months 2025.
The Biden admin supported the construction $200 billion in new manufacturing facilities that began in 2022 and is now tapering down. It will take a lot of jobs to fill those facilities. But will jobs grow in the current economic environment?
Just about anything that can be considered a leading indicator is pointing down. Layoffs, container ship projected offloads are down and falling, China cut shipping to US, supply chains disrupted, immigrant fears affecting labor. Expect costs up, workload down, labor tight.
I’ve been asked, Why don’t you use AI to develop economic analysis? Artificial Intelligence sometimes gets analysis really wrong. There is some percentage (40%?, 60%?) of end results that AI creates that is literally just made up. If you were to use AI to develop forecasts and analysis of construction data, without having a thorough knowledge of the data and an ability to recognize when it’s meaningful, or garbage, then how would you know when AI is right or wrong. Understand your data well enough to know when your analysis makes sense. For my part, I’d rather spend my time understanding the data and the analysis then to spend it verifying if AI is producing realistic and meaningful output.
Summer is just around the corner. The Hummingbirds returned last week.
Construction Analytics Outlook Feb 2025
2-22-25 — A PDF of this entire Outlook article has been attached at the bottom of this post. 32pages, watchout. The Outlook has quite a bit more than in the post here.
Construction Spending
for 2024 vs 2023, as of Dec 2024 data, is up 6.5%. All sectors gained between 6% and 7% over 2023. Growth is forecast at 5.5% in 2025. While Residential and Non-building Infrastructure will both gain 7%, Nonresidential Buildings will only see growth of 2% in 2025.
In February of 2025, with the Dec 2024 data in hand, my forecast for 2025 spending is $2,272 billion, 5.5% higher than my current 2024 forecast of $2,154 billion. There is strength in most markets, but Manufacturing is starting a downward slope in spending after three years of blockbuster performance.
This is the first report of a full 12 months of data from 2024. This number gets revised in Mar and Apr and again in July, when any/all months for the last two years get revised.
Last year at this time, leading into 2024, many of the Nonres Bldgs and Non-bldg line items showed Nov-Dec spending was already several points higher than the 2023 average. This year, many markets show very small gains or a decline in the rate of spending from the 1st half of the 2024 into the 2nd half. Some notable declines are Warehouse (-2.7%), Office w/o Data Centers (-3.6%) and Highway/Bridge (-2.9%). All begin 2025 down from the average in 2024.
However, Data Center spending is already up 16% in Oct-Nov-Dec vs the average of 2024, so begins 2025 on a high note, up 16% from 2024. Data Center spending increased 45% in 2023 and 56% in 2024. With spending increasing at an average 3%+ per month in 2024, and starting out at that pace in 2025, it’s easy to predict Data Center spending may reach 40% growth for 2025.
As we begin 2025, the current rate of spending (SAAR) for Nonresidential Buildings in Q4’24 is $761bil, only 2% higher than the average for 2024 ($746bil). If spending stays at the current level and no additional growth occurs, Nonresidential Bldgs spending will finish 2024 up 2%. Spending would need to have more monthly declines than increases to finish the year up less than 2%. The current forecast shows the monthly SAAR rates for Manufacturing, Warehouse and Office w/o DC are driving the downward pressure on overall spending.
Non-building Infrastructure current rate of spending is only 1% higher than the average for 2024, however the forecast is indicating steady growth of 1.5%/mo for all of 2025. Highway, Transportation and Public Utilities are all contributing to that growth in the spending rate.
Residential current rate of spending is a bit less than 2% above the 2024 average. Growth of 1%/mo will occur in the 1st half 2025, then reverse to a slight decline in the 2nd half.
My construction spending forecast for 2025 Nonres Bldgs is only an annual gain of 2%. Low growth is driven by projects ending in Manufacturing and Warehouse. In the last 3 yrs, there were $230bil Mnfg starts, most in 2022, $130bil above normal, now some are ending. Without Manufacturing, Nonres Bldgs 2025 spending would be up 6%. So while outward appearance may be that Nonres spending is declining, in large part it is due to mega spending on Manufacturing buildings (and Warehouse) tapering down upon completion, creating very large annual declines, but normal. See The Manufacturing Spending Taper
Whenever we get an unusually large increase in new construction starts and spending, the tapering off of those projects leads to a decline in spending in the 2nd half of the scheduled construction. Mnfg new starts peaked in late 2022 – early 2023, so some of these projects would now be in construction for anywhere from 24 to 30 months, well past the midpoint or peak spending. Here’s what the manufacturing spending taper may look like.
In three years, 2020-2022 Manufacturing new starts increased by over 200%. Now, even projects that started in late 2022 are more than 24 months into construction. Peak spending occurs at the midpoint of a project, so peak spending may already be behind us. The current rate of spending in constant$, with exception of Oct. which jumped 2%, remained near flat for the last 8 months. Mnfg new starts peaked in late 2022 – early 2023, so some of these projects would now be in construction for anywhere from 24 to 30 months, well past the midpoint or peak spending. It appears the slowdown in the 2nd half 2024 may be an indication that spending peaked. Although 2025 spending begins 4% higher than 2024 average, my model indicates the rate of spending drops 10% by midyear and by year-end is down 25% from current spending. I’m forecasting 2025 spending average falls 8.5% below 2024.
See The Manufacturing Spending Taper
This same scenario will occur in Highway/Bridge. Normal Highway starts have consistently been about $100bil/yr, with slow growth. But actual starts were closer to $500bil total for the last three years. This strong growth in starts is expected to continue at least into 2025, totaling near $650bil for four years. Again, consider that part of that is inflation, but the remainder is government investment growth. So a decline from the taper back to normal for Highway/Bridge may not show up at least for the next few years. Once the taper begins it will have the same effect on Non-bldg Infrastructure spending that we will see from Manufacturing in Nonres Bldgs.
Residential spending is 45% SF, 15% MF and 40% Reno. So, only 60% of the total is spending on new housing units. The other 40% is Renovations. Single family construction spending reached a post-2006 high in Q4’21 thru Apr’22. From Apr’22 to the low-point in May ’23 spending dropped 25%. By year-end 2023 it had recovered most of that drop. It fell again in mid-2024, but has since recovered again to the year-end’23 level. Single Family spending is up 7% in 2024 over 2023. Multi-family in Q4’24 is down 10% from Q4’23. Renovations is up 15% from Q4’23 to Q4’24.
The Non-Building Infrastructure spending forecast for 2025 will be most affected by the fact that Power starts for the last 3 years range from 9% to 11%, Highway new starts range from 8% to 11% and Public Works new starts range from 13% to 17%. Starts have been greater than spending each of the last three years. Backlog increased 10% each in 2023 and 2024 and 9% in 2025. Power spending will increase $13bil (8.5%) in 2025, supported by 22% growth in starts the last 2 years. Public Utilities (Sewage and Waste, Water Supply and Conservation-Rivers-Dams) will increase $12bil (13%), with 30% growth in starts the last 2 years.
The Nonresidential Buildings spending forecast for 2025 is most affected by declines in Manufacturing and Warehouse, and increases in Educational and Data Centers. Starts in 2021 and 2022 increased at an average 14%/yr. But new starts in 2023 were only 5.5%, and in 2024 were only 2.6%. Starts in 2023, and especially 2024, have the most impact on 2025 spending. Backlog is increasing, but the rate of backlog growth has slowed from 15% in 2022 to 5% in 2024 and 2% in 2025. Although Data Center spending is expected to increase 42%, that is $12bil. Educational spending will also increase $12bil, even though that is only 9%.
Educational SAAR rate of spending begins 2025 4.5% higher than the average for 2024. Starts are up 18% over the last 2 years. Spending finished 2024 3% higher than it started. The rate of spending is increasing at 0.50% to 0.75%/month for 2025. My forecast is for 9% spending growth in 2025.
Healthcare SAAR rate of spending begins 2025 0.5% lower than the average for 2024. Starts are up 22% over the last 2 years. The rate of spending, increasing since June, is flat in Q1’25, then increases at an average of 1.5%/month for Q2 and Q3 before slowing. My forecast is for 6% spending growth in 2025.
Amusement /Recreation SAAR rate of spending begins 2025 3.0% higher than the average for 2024. Starts are up 18% over the last 2 years. The rate of spending, up 9% from Q1’24 to Q4’24, increases at an average 1.25%/month for 2025. My forecast is for 13% spending growth in 2025.
Lodging SAAR rate of spending begins 2025 3% lower than the average for 2024. Starts are up 14% in 2024 and 20% over the last 2 years. The rate of spending, lower in the 2nd half 2024, increases at an average of 1.25%/month for 2025. My forecast is for 9.5% spending growth in 2025.
Office Bldgs w/o Data Centers SAAR rate of spending begins 2025 3.6% lower than the average for 2024. Starts are down 15% over the last 2 years. Spending in the 2nd half of 2024 is down 5% from the 1st half, in fact from Q1’24 to Q4’24, the rate of spending fell 10%. The rate of spending increases 0.5%/mo in Q1’25, but then falls at 0.50% to 0.75%/month for Q2 thru Q4. My forecast is for a 4% spending decline in 2025.
Data Centers SAAR rate of spending begins 2025 16% higher than the average for 2024. Starts are up 120% over the last 2 years. From beginning to end of 2024 spending increased 20%. The rate of spending increases at near 2%/month for 2025. My forecast is for 42% spending growth in 2025.
Commercial / Retail Bldgs w/o Warehouse SAAR rate of spending begins 2025 only 1.5% higher than the average for 2024. Starts are down 0.5% over the last 2 years. In 2024, the rate of spending fell 3% over the year. The rate of spending begins 2025 with a slight drop then is flat for the rest of the year. My forecast is for 1% drop in spending in 2025.
Warehouse Bldgs SAAR rate of spending begins 2025 2.7% lower than the average for 2024. Starts are down 17% over the last 2 years. Spending in the 2nd half of 2024 is already down 13% from the 1st half 2024. The rate of spending in 2025 drops at 0.25%/month. My forecast is for 4% spending decline in 2025.
Highway SAAR rate of spending begins 2025 2.9% lower than the average for 2024. Spending fell 7% over the course of 2024, but still finished the year up 4%. Starts are up 18% over the last 2 years. Monthly spending is up and down, but averages 0.8%/mo for the year. My forecast is for 3% spending growth in 2025.
Power SAAR rate of spending begins 2025 2.0% higher than the average for 2024. Starts are up 22% over the last 2 years. 2024 spending finished level with where it started and begins 2025 up 2% over avg. 2024. The rate of spending starts strong at 2%/mo in Q1, then falls back to 0.3%/mo for the remainder of 2025. My forecast is for 8.5% spending growth in 2025.
Public Utilities SAAR rate of spending begins 2025 1.7% higher than the average for 2024. Starts are up 29% over the last 2 years. From Q1’24 to Q4’24, spending increased 9%. The rate of spending slows from 1.5%/mo at the beginning of 2025 to 1%/mo by year-end. Sewage/Waste and Water Supply provide the greatest $ growth. My forecast is for 13% spending growth in 2025.
Transportation SAAR rate of spending begins 2025 4.0% higher than the average for 2024. Starts are up 9% over the last 2 years. Spending finished 2024 6% higher than it started. The 2025 rate of spending starts strong just above 1%/mo, but averages only 0.75%/mo for the whole year. My forecast is for 8.7% spending growth in 2025.
STARTING BACKLOG
Starting backlog is the estimate to complete (in this analysis taken at Jan 1) for all projects currently under contract. The last time starting backlog decreased was 2011. If new construction starts in the year are greater than construction spending in the year, then for the following year starting backlog increases. It’s when new starts don’t replenish the amount of spending in the year that backlog declines.
80% of all nonresidential spending in any given year is from backlog and could be supported by projects that started last year or 3 to 4 years ago. Residential spending is far more dependent on new starts than backlog. Only about 30% of residential spending every year comes from backlog and 70% from new starts.
The table below, Forecast Starting Backlog, is model generated by Construction Analytics. Adjusted starts are spread over time to generate cash flow. Spending each year is subtracted from starting backlog and new starts are added to get starting backlog in the following year.
Construction Backlog leading into 2025 in total is up 53% from Jan 2020. Even though several markets are down for the year, every sector (Res, Nonres, Nonbldg) is at all-time high. Since 2020, construction starts have been greater than construction spending, therefore backlog is increasing, but the spread is decreasing. From 2015 through 2019, new construction starts were steady at 4% greater than spending. In 2020, 2021 and 2022, starts exceeded spending by 8%. In 2023 it fell to 6% and in 2024 it was 5%. For 2025, starts are expected to exceed spending by only 3%, indicating the rate of backlog growth is slowing.
Reaching new highs in Backlog could mean not enough labor to support advancing growth so quickly. Future workload from new starts is piling up faster than the current workforce can complete. And the labor force has limitations to how fast it can grow. Nonresidential Bldgs and Non-building posted higher than average starts growth vs spending.
Residential new starts average only 2% greater than spending. Residential new starts in 2023 were lower than spending, so, in 2024, for the first time in 10 years, residential backlog decreased. Even then, residential backlog beginning 2025 is up 61% since 2020. However, these backlog numbers are not inflation adjusted. 45%, or about ¾’s of the 61% residential backlog growth over 5 years is inflation. Real residential backlog growth since 2020 is 16%.
Nonresidential Bldgs starting backlog for 2025 received a boost from all the starts in 2021, 2022 and 2023. Backlog is up 55% from 2020. After 36% inflation, real backlog growth over 5 years is 19%.
Non-building Infrastructure starting backlog was also boosted by strong starts in 2021- 2024. For 2025, backlog is up 49% from Jan 2020. After 38% inflation, real backlog growth over 5 years is 11%.
Manufacturing backlog increased 130% from 2020-2025, from $117bil going into 2020 to $270bil beginning 2025. Prior to tracking Data Centers separately, no other market has ever been close to that growth. But, Manufacturing is 6 times the dollar value of Data Centers. Manufacturing was responsible for 60% of all the Nonres Bldgs spending $ growth in 2023 and 85% in 2024. It was also responsible for 33% of the Nonres Bldgs Backlog growth from 2020 to 2025.
Nonres Bldgs has a total 3.7 million jobs and has never increased by more than 150,000 jobs in one year. Manufacturing is 30% of all Nonres Bldgs spending, so we can assume 30% of Nonres Bldgs jobs. That’s 1.1 million jobs supporting just Manufacturing projects. So Backlog of $270bil, at 5000 jobs per billion per year, would need 1,350,000 jobs for a year. With a 1,100,000 jobs share of the workforce, $270billion in backlog would provide support for 15 months. Of course, new starts add to support throughout the year, but the calculation of how long backlog would support that market segment is valuable.
Likewise, Highway/Bridge has 2025 starting backlog of $240billion and represents 30% of Non-bldg Infrastructure spending, so may occupy 30% of Non-bldg jobs, or 345,000 jobs. But Non-bldg work requires fewer jobs, more like 3,000 jobs/bil/yr, so $240 billion at 3000 jobs/billion/year would need 720,000 jobs for 1 year to complete. Therefore, with a pool of only 345,000 jobs supporting highway work, the backlog of $240billion would provide support for 25 months.
Backlog at the beginning of the year or new starts within the year does not give an indication of what direction spending will take within the year. Backlog is increasing if new starts during the year is greater than spending during the year. An increase in backlog could immediately increase the level of monthly spending activity, or it could maintain a level rate of market activity, but extended over a longer duration. In this case, there is some of both in the forecast. It takes several years for all the starts in a year to be completed. Cash flow shows the spending over time.
NEW CONSTRUCTION STARTS
Total construction starts for 2024 are up 5.3%. Residential starts increased 5.7%.Nonresidential Buildings starts gained 2.6% and Non-building Infrastructure starts are up 8.3%.
Total construction starts for 2025 are forecast to increase 3.8%. Residential starts are forecast to increase 6.1%. Nonresidential Buildings starts are expected to gain 2.5% and Non-building Infrastructure starts will be up only 1.6%.
Residential starts increased 5.7% in 2024. Only about 30% of the spending in 2025 comes from 2024 starts. Most of the spending (70%) in 2025 comes from projects that starts in 2025. This is a result of short duration single-family and renovations projects. Residential starts are expected to increase 6.1% in 2025.
Nonresidential Buildings, starts fell 20% in the 1st half 2023 but still posted the 2nd highest 6-mo average ever. Some of these starts will still be adding to spending into 2025. Nonres Bldgs starts for 2024 gained only 2.6%, due to large declines of 7% to 8% in Manufacturing, Office w/o Data Center and Warehouse, and a moderate 1.2% decline in Commercial / Retail w/o Warehouse. Office w/o Data Center has fallen each of the last 5 years and is down 15% in the last 2 years. Manufacturing starts in 2025 are expected to drop -13%, Office w/o Data Center -4%. Strong growth in new starts in 2025 is expected from Data Centers, Lodging, Healthcare and Public Bldgs (80% of Other Nonres Bldgs). The forecast for Nonres Bldgs new starts in 2025 is +2.5%.
Non-building starts for the 6-mo period Mar-Aug 2023 posted the best 6 months on record, up 30% from the average of 2022. For 2024, Power, Highway/Bridge and Public Works have the strongest gains. These same three markets had the strongest gains in 2023. Power starts are up 22% the last two years. Highway starts are up 19% the last two years. Environmental Public Works are up 29% the last two years and up 50% the last three years. Total Non-building Starts for 2024 are up 8.3%. Non-building starts for 2025 are forecast up only 1.6%.
Data Center starts are up 300% since 2020, up 53% in 2023 and up 44% in 2024, and are expected to increase 27% in 2025. In 2014-2015, Data Centers was less than 5% of total Office+DC construction spending. Today it is approaching 30%. Next year it will approach 40%.
Warehouse starts have dropped 17% in the last two years. Warehouse spending will now slow after climbing 100%+ since 2019. In 2015, Warehouse was 25% of total Commercial spending. By 2022 it had climbed to 54%. In 2025, it will fall back to 45%. But spending will remain near the current level at least for the next three years. Warehouse starts will remain flat in 2025.
Manufacturing starts, the market with the largest $ movement, fell 8.7% in 2024, but still gained 115% from 2019 to 2024. Manufacturing projects can have a moderately long average duration because some of these are multi-billion $ projects and can have schedules that are 4 to 5 years, so some of these projects are still contributing a large volume of spending in 2025. However, as earlier projects begin to taper off, spending will begin to decline.
Manufacturing starts hovered near $80bil/yr from 2014 through 2019. By 2023 new starts had increased to $206bil/yr. Starts fell 9% in 2024 and are predicted to fall by $20bil/yr to $30bil/yr (10% to 15%) over the next three years. Spending is predicted to decline by approx. 10%/yr for the next three years.
Public Works project starts have increased on average 15%/yr for the last four years. Project starts are up 75% in the last four years. Spending is predicted to climb for the next three years.
CURRENT $ SPENDING / INFLATION / CONSTANT $ VOLUME
Volume = spending minus inflation. Spending includes inflation. Inflation adds nothing to volume.
Many construction firms judge their backlog growth by the remaining estimate to complete of all jobs under contract. The problem with that, for example, is that Nonresidential Buildings spending (revenues) increased 14% in 2022, but after adjusting for 12% inflation the actual volume of work was up only 2%. By this method, firms are including in their accounting an increase in inflation dollars passing through their hands. Spending includes inflation, which does not add to the volume of work.
Total volume for 2024 is up 3.1%. Residential +2.7%, Nonres Bldgs +3.3%, Non-bldg +3.5%.
Total volume forecast for 2025 is +1.3%. Residential +2.9%, Nonres Bldgs -1.9%, Non-bldg +3.0%.
Since 2019, spending has increased 55%. Volume has increased only 10%. The difference is inflation.
Inflation adjusted volume is spending minus inflation, or to be more accurate, spending divided by (1+inflation). Inflation adds nothing to volume growth. The following table shows spending, inflation and volume (spending without inflation) for each year. Spending is current to the year stated. The values in the constant table are indexed to a constant value year, now using 2024. This shows business volume year to year, can be a lot different than spending would indicate. When inflation is positive, volume is always less than spending by the amount attributed to inflation.
Caution: the following table, showing Constant$ analysis, now shows Constant$ with base year at 2024. Since Q1-2020 I have used the base year at 2019. This update changes the Constant$ amount, but not the Constant$ percent growth. Slight changes in prior years inflation resulted in some minor changes in Constant$ growth.
Spending during the year is the value of business volume plus the inflation on that volume. When inflation is 12%, volume plus 12% = total spending. Revenue is generally measured by spending put-in-place during the year. Therefore, Revenue does not measure volume growth. In 2022, Nonresidential buildings inflation was 12%, so business volume was 12% less than spending, or 12% less than revenue. Residential volume was 15% less than spending.
Construction spending includes inflation, which adds nothing to the volume of work put-in-place. Construction Volume is a measure of business activity. It eliminates inflation as a variable and shows Constant$ growth. As an example, 2021-22 posted some of the biggest spending increases we’ve seen in 20 years, up over 25% in two years. But, if you look at the bottom row in the table above, we see constant$ growth, or volume, increased only 1.6% in those two years. The inflation rates for those years confirms that almost all of the spending increases were inflation, not added business volume. Construction volume, (spending minus inflation) will finish 2025 up 1.3%, but up only 12% since 2019.
Compare this following Spending by Sector Constant$ plot to the Sector Current$ plot at the beginning of this article.
CONSTRUCTION JOBS
Construction Jobs should not get compared to construction spending. Spending includes inflation, which adds nothing to business volume. Compare Jobs growth to Volume growth. If volume is declining, there is little to no support to increase jobs. And yet, we’ve seen historically that jobs increase at an approximately 3.5%/yr, even when volume does not increase.
Construction Jobs increased 204,000 jobs or 2.5% in 2024. There are currently 8,291,000 construction jobs reported by BLS. The largest annual increase post 2010 is 321,000 jobs (+4.6%) in 2018. The average jobs growth post 2010 is 200,000 jobs per year.
From 2012-2019, we added an annual average 245,000 jobs/yr (+3.9%). From 2021-2024, we added 240,000 jobs/yr (3.1%). In 2024 we added only 204,000 jobs (2.5%). Since 2011, there have been only 3 years in which we added fewer than 200,000 jobs, 2012, 2020, 2021. Also, since 2011, there have been only 3 years in which we added more than 300,000 jobs, 2015, 2018, 2022. Seldom do jobs increase by 400,000 or by 5%/yr. Excluding down years, since 2012, average annual growth is 3.6%/yr.
Here’s an enlarged view of just Jobs/Volume. Removing Spending enhances the vertical scale.
From 2012-2019, we added an annual average 245,000 jobs/yr (+3.9%). From 2021-2024, we added an average of 240,000 jobs/yr (3.1%). In 2024 we added only 204,000 jobs (2.5%). Since 2011, there have been only 3 years in which we added fewer than 200,000 jobs, 2012, 2020, 2021. Also, since 2011, there have been only 3 years in which we added more than 300,000 jobs, 2015, 2018, 2022. Seldom do jobs increase by 400,000 or by 5%/yr. Excluding down years, since 2012, average annual growth is 3.6%/yr.
January jobs report shows we added 4k jobs in Jan, but the unemployment rate went up from an average 4.7% in Q4 2024 (5.2% in Dec), to 6.5% in Jan.
3-7-25 Construction gained 19k (+0.2% mo/mo) jobs in Feb, BUT total hrs worked fell -0.3%. Total jobs have increased but hrs worked has gone down in each of the last 5 months. We’ve posted minor jobs gains in both Jan and Feb, and yet unemployment has gone up from 5.2% in Dec to 7.2% in Feb.
The unemployment rate in construction goes UP in the 1st qtr every year, by at least 2% to 3% (data since 2011). Now, your 1st thought may be, if unemployment is increasing, that is probably because jobs are falling. Well, construction has ADDED jobs in the 1st qtr. every year since 2011 (excluding 2020), by an avg of nearly 30% of all jobs added annually. Construction unemployment is not going up in winter months because we lose jobs in winter. So how can the unemployment rate still go up? The numerator (jobs) is increasing. There’s only one number left in the equation, the denominator (workforce). The result goes up because the entire workforce increases. In this case, the workforce is increasing by greater than the number of jobs added. So, when the unemployment rate increases in Q12025, don’t assume it is because we are losing jobs.
The plot below shows how consistent jobs growth has been over the last 14 years. After the 2020 dip, the slope (annual rate of growth) of jobs growth is about the same as Jan2011-Jan2020. Jobs increase at about 3.5%/yr to 4%/yr., regardless of what volume is doing. In fact, since 2016, the last time volume increased by more than 4%, jobs are up 22%. Volume is up only 11%.
Since Jan 2011, average jobs growth is 3%/yr. Average volume of work growth since 2010 is 2.3%/yr. This plot above shows Jobs and Volume growth closely match from 2011 to 2018. With few exceptions for recession periods, this pattern can be seen throughout the historical data.
What’s remarkable about the growth is this; since 2016, spending has increased 77%, volume after inflation increased only 11% and jobs increased 22%. Volume and jobs should be moving together.
It takes about 5000 (Nonres) jobs to put-in-place $1 billion of volume in one year. It could easily vary from 3000 to 5000, depending on the type of work. So, on average, an add of $100 billion+ of Nonres Bldgs in one year would need 500,000 new jobs. Jobs should track volume, not spending growth. Volume = spending minus inflation. Normal construction jobs growth is about 250,000 jobs per year and maximum prior annual growth is about 400,000.
Non-building, over the next two years, could experience the same kind of growth spurt as Nonres Bldgs., a forecast increase in volume the next two years without an equal increase in jobs. Volume which was lower than jobs since 2021, is now increasing faster than jobs. Non-bldg volume is forecast up 6% to 8%/year the next 3 years. Jobs increase at an avg. 3.5%/year.
Residential volume has exceeded residential jobs all the way back to 2011. The recent decline in volume brings the two even, if the jobs hold the pace.
For as long as I can remember, the construction industry has been complaining of jobs shortages. And yet, as shown in the data mentioned above, jobs have increased greater than volume of work. With an exception for recession years, (2007-2010 and 2020), jobs increase at a rate of 2.5% to 3% per year. The greatest disparity between jobs and volume occurred in late 2022, when jobs growth had already resumed normal pace, but volume of work was still reeling from the effects of new construction starts that were canceled dating back to late 2020 and early 2021. Recent volume growth at a much faster rate than jobs growth is now closing the gap.
For the 4yr period 2021 thru 2024, jobs are up 13%, but volume of work put-in-place is up only 6%.
Don’t be surprised if 2025 construction jobs growth slows a bit. Jobs are slightly ahead of volume growth, particularly in the Non-building Infrastructure sector. Since 2019, both Jobs and Volume increased 10%. But that includes 2020, when volume increased 4% but jobs fell by 250k, or 3%.
When jobs increase without an equal increase in the volume of work, productivity declines. This recent increase in volume and the projected increase in volume in 2024, several points stronger than jobs, will offset some of the disparity which has been negative for a long time.
The Harvard Joint Center for Housing Studies recently posted that In Texas, California, New Jersey, and the District of Columbia, immigrants make up more than half of construction trade workers. This analysis will make no effort to discuss the potential impact of immigrants in the workforce, but we must recognize the makeup of who is filling these jobs. Keep in mind after this analysis reaches some conclusions that this aspect may weigh heavily on the outcome.
INFLATION
To properly inflation adjust the cost of construction, use a Final Cost Inflation Index.
General construction cost indices, that do not address labor, productivity or margins and Input price indices, don’t reflect whole bldg final cost and therefore do not capture the full cost of escalation in construction.
Final cost indices represent total actual cost to the owner and are usually higher than general indices. Producer Price Index (PPI) INPUTS to construction reflect costs at various stages of material production, may not represent final cost of materials to the jobsite and do not include labor, productivity or margins. That’s why a PPI Inputs index +20% for a material could be only a +5% final cost. PPI Final Demand indices include all costs and do represent actual final cost.
Construction Analytics Building Cost Index is a weighted average of eight final cost indices.
The following Construction Inflation plot (for Nonresidential Buildings only) shows three elements: 1) a solid grey bar reflecting the max and min of the 10 indices I track in my weighted average inflation index, 2) a solid black line indicating the weighted average of those 10 indices, and 3) a dotted red line showing the Engineering News Record Building Cost Index (ENR BCI). Notice the ENR BCI is almost always the lowest, or one of the lowest, indices. ENR BCI, along with R S Means Index, unlike final cost indices, do not include margins or productivity changes and in the case of ENR BCI has very limited materials and labor inputs.
The solid black line (above) represents the Construction Analytics Building Cost Index for Nonresidential Bldgs and is a final cost index.
This short table shows the inflation rate for each year. Useful to compare to last year, but you would need to mathematically do the compounding to move over several years. The plot below shows the cumulative inflation index, or the cumulative compounded effect of inflation for any two points in time.
30-year average inflation rate for residential and nonresidential buildings is 4.1%. But when excluding deflation in recession years 2008-2010, for nonresidential buildings the long-term average is 4.7% and for residential is 4.9%. For Non-bldg Infrastructure the 30-year average is 3.6%. When excluding deflation in recession years 2008-2010, Non-bldg long-term average inflation is 3.9%.
Since 2011, Nonres Bldgs inflation is 4.8%, Residential is 5.4% and Non-bldg is 4.3%.
Reference Inflation Data Construction Inflation 2024
Construction Analytics Nonres Building Cost Index is a weighted average of eight final cost indices. It is compared below to the PPI Inputs (not final cost) and PPI Final Demand (yes final cost).
Final cost indices represent total actual cost to the owner and are usually higher than general indices.
Producer Price Index (PPI) INPUTS to construction reflect costs at various stages of material production, generally do not represent final cost of materials to the jobsite and do not include labor, productivity or margins. Even with that, a PPI Inputs index +20% for a material could be only a +5% final cost. PPI Final Demand indices include all costs and do represent actual final cost.
We can’t always tell what affect changes in the cost of construction materials will have on the final outcome of total construction inflation. PPI materials index does not account for productivity or margins and varies on stage of input.
Jan’25 Inputs are up, +0.8% to Nonres, +1.2% to Residential and 1.2% to Highway, the largest increases since Jan’24. The largest Input item increases in January are Paving Mixtures (up 14.6%), Diesel Fuel (up 3.6%), Concrete Brick, Block & Pipe (up 2.2%), Copper & Brass Shapes (up 1.9%) and Flat Glass (up 1%). Steel Pipe & Tube is down 1.3%. Both Lumber/Plywood and Fabricated Structural Steel are down 0.5% or less.
The PPI Final Demand index (for Nonresidential Bldgs only) is one of several that does account for labor and margins, hence it is defined as a final cost index. In this plot, Jan’25 closes out Q4’24. Total growth in 2024 was only 0.4%.
A Check on Measuring Methodology
And finally, here’s one of the markers I use to check my forecast modeling, my forecasting performance tracking index. The light plot line is forecast predicted from my modeling. The dark plot line is actual construction spending. Even after any separation in the indices, the plots should move at the same slope. Almost without fail, the forecast model, estimated spending from cashflow, predicts the changes in direction of actual spending. The nonresidential buildings plots (and the residential plot prior to 2020) are remarkably close, providing an indication the method of analysis employed, cash flow of all construction starts to get spending forecast, is reasonably accurate.
Note the divergence of residential in Jul-Dec 2020. Actual residential spending finished much higher than predicted. Even the cash flow from an all-time high in new residential starts does not predict spending to increase so rapidly. But residential project spending was fully back to prior levels by August 2020, within 3 months from the May 2020 bottom. In 3 months, the actual spending pushed 15% higher than starts predicted. A part of the spending was the resumption of delayed projects, but another big part was renovations, which surged, and reno is 40% of all residential spending.
Construction Inflation 2025 – 2-21-25 PPI data, INDEX TABLES, Inflation History
Construction Briefs – As We Begin 2025 jobs and unemployment, PPI and tracking edz
Construction – Brief Thoughts Dec’24 thoughts on jobs and when spending mega ends
Compare 10 Construction Forecasts Jan 2025 Jan and Midyr Forecasts compared to actual at end of yr
The Next Forecast Challenge written Apr 2023. Nonres volume did climb much faster than Jobs.
This analysis does not take into consideration the inflation impact of a recession or significant new tariffs. Nor does it assume losing any portion of the workforce to deportation. These are all possible. Construction starts may be negatively affected, and if so, then construction spending will also be affected. Some business will be negatively affected. All will lead to higher inflation.
You can print the Summary report by selecting/printing just the first 5 pages.
Construction – Brief Thoughts Dec’24
Construction spending in 2024 will hit near $2.15 trillion, another new high, up 50%+ since 2019. (The last time constr spending declined was 2011. Construction volume, (spending minus inflation) will finish the year up only 10% since 2019.
Construction Spending for OCT is up 0.4% mo/mo. Aug & Sep rvsd up ~1%/mo, mostly Rsdn. Spending now essentially flat for last 4mo. YTD’24 vs YTD’23 now +7.2%, falling slightly as expected for last few mo, because spending was rapidly increasing in late months of 2023 and is flat in 2024.
Real Manufacturing Construction Starts over last 3yrs is just over $600bil, whereas normal starts without any influx of government investment would be about $300bil for the same 3yrs. Most of those starts were posted from Q3’22 thru Q4’23. By far the highest period of growth in new starts was the 2nd half 2022. About $100bil of that spending growth over the 3 years is inflation, leaving the remainder of about $200bil in excess (but welcome) spending growth attributed to government investment in new manufacturing facilities. Now some of these projects are ending or soon to end. My latest construction spending forecast for 2025 Nonresidential Bldgs is down 0.5%. But this decline is driven by projects ending in Manufacturing. Without Manufacturing, Nonresidential Buildings 2025 spending would be up 4%.
The same effect on spending by the tapering off of Manufacturing Bldgs ending will happen in Infrastructure spending due to the current increases in Highway/Bridge, only ending a few years later.
Construction jobs increased to 8.313 mil in NOV. Since SEP, last two months barely increased, only +12k jobs. But perhaps more important, hours worked, since SEP, decreased 0.75%, equiv of 60k jobs. However, ytd’24 vs ytd’23, jobs are up 2.9%, total hrs worked is up only 2.7%
The unemployment rate in construction goes UP in the 1st qtr every year, by 2% to 3% (data here since 2011). Now, your 1st thought is that if unemployment is increasing, that is probably because jobs are falling. Well, construction has ADDED jobs in the 1st qtr. every year since 2011, by an avg of 24% of all jobs added annually, and more recently by over 30% of total annual jobs. Construction unemployment is not going up in winter months because we lose jobs in winter. It goes up because the entire workforce increases by greater than the number of jobs added.
To properly inflation adjust the cost of construction, use a Final Cost Inflation Index. General construction cost and Input price indices that don’t track whole bldg final cost do not capture the full cost of escalation in construction.
Will vacant office space to residential conversions be the construction spending that ignites the 2025 fire? Residential is forecast now to spring ahead 8% in 1H’25, then hold steady thru 2H. If conversions jump, that percent goes up and we could see a shortage of residential labor. (not addressing deportations).
There are 870K construction jobs in TX. 500K are immigrants and 300K are undocumented.
Sweet potatoes should be steamed only just soft enough to slice (long way preferred) about 1/2″ thick, then griddle fried in butter and maple syrup until there is some blackening (important). This was my dad’s recipe. No other recipes are allowed.
Hurricanes & Residential Construction
Over the next few years there could be many billion$ of construction spending to “repair” hurricane damaged homes. That spending will NOT be reported in Census construction spending reports. According to Census definition, renovations to “repair” natural disaster damage are not recorded in construction spending.
Construction spending to “replace” homes entirely lost to hurricane damage (wiped completely off the foundation) IS reported in Census spending, but it is reported as renovations/repair, not new SF or MF construction. These are not “new” homes and do not add to the new home count. The work is to replace an existing home.
There will be billions$ of residential construction and the associated labor that will fall outside posted Census data. Some of that spending and labor will be diverted from the growth in residential construction tracked by the Census Rsdn$. So, it is conceivable that Census tracked residential spending could slip slightly from current projections, and yet the work is being done, only unaccounted for.
(All of this is true also for nonresidential).
Expect increased demand and potentially longer waits for all residential building products; lumber, doors, windows, roofing, siding, drywall, flooring, HVAC, plumbing, electrical and appliances, primarily or exclusively in the affected regions.
If your business is in a region that is currently outside the affected region and slow for residential construction, consider that anybody with a hammer and a pick-up truck may be traveling to the affected regions to work. That could mean that many slow regions may lose some available labor, as the labor will move to the affected area.
For every $1billion of work needed to replace damaged buildings it requires 4000-5000 construction workers for one year. That could also be stated as 1000 jobs for 4 to 5 years. Very early damage projections due to Helene and Milton range from $50-$175 billion, but not all of that is construction. If even $10 billion of work is construction to repair/replace damaged buildings, that would require 10,000 jobs for a period of 4 to 5 years (or 20,000 jobs for 2 to 2 1/2 years).
Can We Increase Residential Construction by 3 Million Units in 4 Years?
Here’s a look at the residential construction industry. Except for the first mention of Current $, All $ in this analysis are inflation adjusted Constant 2019$, so all years can be compared on equal terms. All numbers are rounded slightly.
Currently in 2024 $, residential construction spending is $940 billion/year, 43% of all construction. Only 60% of that ($560 billion) is spending on new housing units. The other 40% is Reno.
Currently for 2024, residential construction supports 3,350,000 jobs. Only 60% of the jobs (2,000,000) are for new units.
30 years ago residential construction spending was $550 billion/year and there were 2,200,000 jobs. On average over 30 years (including all up or down years) we’ve added $14 billion/year in spending and 38,000 jobs per year.
The 30 year (1995-2024) average growth in residential spending, counting only 21 up years, (eliminating all 9 down years) is +$36 billion/year. There were 2 years that we added +$70 billion/year in spending (one of those in 2021) and including those, there were only 7 years over +$50 billion/year.
The 30 year (1995-2024) average growth in residential jobs, counting only 21 up years, (eliminating down years) is +125,000 jobs/year. In 1999, we added 270,000 residential jobs. Since 2011, the most we’ve added in a year is 170,000 jobs in 2021. That’s for ALL residential construction.
In 30 years, we’ve never added more than $70 billion/year in residential construction spending nor more than 270,000 residential jobs in a year. Since 2011, the most residential spending increased was $70 billion in a year and the most residential jobs increased was 170,000 jobs in a year, both in 2021.
Only 60% of all residential spending and jobs is associated with building new housing units. The other 40% is renovations, not associated with new construction. That must be considered when reviewing all the historical data noted above. So the following is just to look at the data without renovation.
Reduce the data by 40% to eliminate spending and jobs associated with renovation and look at just the data associated with building new housing units: In 30 years, we’ve never added more than $40 billion/year (2004,2021) in residential construction spending on new units nor more than 160,000 residential jobs in a year (1999,2005). This is for New Units construction only.
Since 2011, the most spending increased for new units was $40 billion in a year (in Constant 2019$) and the most jobs increased was 100,000 residential jobs, both in 2021.
The last time residential jobs for new units increased by 100,000 or more in a year was 2005. Best growth ever was 2004-2005-2006, when we averaged +120,000 jobs/yr for 3 years, just for new units. In fact, the only times residential jobs for new units increased by more than 100,000 were 1999, 2004, 2005 and then finally again in 2021. But 2004-2006 was a time when the construction unemployment rate averaged 7.5%-8.5%. In 2021, construction unemployment averaged 6%. Now the unemployment rate is averaging 4.5%.
(It would be wise to use post-2011 data for the jobs analysis. The construction industry, especially residential and nonres bldgs, experienced a labor base reset in the period 2006-2010, a period from which we have never fully recovered. Notice in this plot that residential was nearly level at 4000 for 10 years, then from 2005-2009, it shot up to over 5000. It’s remained near 5000 since 2010.).
If we were to match the maximum growth rates since 2011 (and in the case of spending, since 1994) we would increase new residential units construction spending by $40bil/$390bil = 10% and we would increase new residential jobs by 100,000/2,000,000 = 5%. Spending maxed at 10%, but jobs max growth at 5%.
The number of new housing units built in a year varies, but for this post I’ll use 1,500,000. Currently it’s a bit less than that. Increasing new units by 5% to 10% per year on a base of 1.5 million units/year would net a growth between 75,000 to 150,000 added units per year. Let’s use 100,000 new units in the first year. That would be 6.7% growth, already more than the 5% jobs. We would continue the growth, so we increase that by another 100,000 in the 2nd year, and again in the 3rd and 4th years. Leave no doubt, growth of 100,000 units/yr would mean adding jobs to match or exceed the best single-year growth rate in the residential new units construction industry since 2005, and keep matching that growth rate for four consecutive years.
CORRECTION: 8-21-24 The Corrections (now fixed) above raise the upper limit of potential growth based on spending from 7% to 10%, or 150,000 units/yr, but only if we take the result to be dependant on the upper limit % spending growth. I had divided by current $ when I should divide by constant $. Spending can vary a lot more than jobs. While some years of jobs growth reach 6%-7%, average residential jobs growth is just over 4%. Jobs never grow at 10%/year.
This plot shows the steady growth in residential jobs, 75% over 15 years. Compounded that’s about 4%/yr. So, it seems much more likely the limiting factor here is jobs, not spending.
In this analysis, the number of units was chosen at 100,000/yr., an increase of 6.7%/yr., well below potential spending growth but already more than jobs growth. The jobs increase in 2021 was 5%/yr. In 2004-2005, jobs growth was 8%/yr. So, this 6.7% is near the high end of potential jobs growth. This implies an add of 130,000 jobs.
So we would increase new units by 100,000 + 200,000 + 300,000 + 400,000 or 1,000,000 total new affordable residential units over the next 4 years.
By this rough analysis, to increase new residential units by 3 million over current production over the next 4 years would require that we increase residential construction and jobs by 3x greater than the best performance achieved in the construction industry in the last 30 years, and keep it up for the next 4 consecutive years. That’s more jobs than the entire construction industry gains in one year and the new units subset of construction is only 25% of all construction.
Although there has been times past when jobs increased by 120,000/year, that was a time when unemployment was much higher and there were more workers available to readily go back into the workforce. It’s not impossible that jobs can grow faster, and therefore also the number of new units, but it is unlikely.
If a substantial portion of the current 1,500,000 unit production (500,000 units a year for the next 4 years) were shifted over to affordable units, the picture changes considerably. But I wouldn’t expect that current demand to just go away and shift to lower cost units. So, this leaves current demand as is and adds jobs.
A Proposal: If some amount of the funds to be invested in this program is directed towards trades training programs, within two years, and every year after, we could potentially be turning out some of the extra 260,000 additional workers we would need every year to achieve 3 million new housing units. This approach would create both good jobs and new affordable homes. Incentivize the training program somehow. Paid internship. Perhaps with a guarantee in return to work for 2-3yrs. If we don’t invest in training workers for this effort, we cannot build more than the labor we have. Here’s a thread on my training proposal. https://x.com/EdZarenski/status/1828778915055468816
Construction Analytics 1,000,000th view, Apr data Briefs 6-10-24
Updates to Forecast, spending, starts, inflation, jobs will follow next few days, so revisit this post for Apr data updates.
SEE ALSO Construction Analytics Outlook 2024
Today, 6-10-24, this blog recorded the 1,000,000 view. That doesn’t count scrolling from the home page, where a visitor might scroll down to read 3 or 4 articles, and there are 500 landings on the home page every week. So, the counter hit 1,000,000 but including scrolling, the actual total views could be higher. Nearly 500,000 people read on average 2.1 articles every visit. Inflation articles draw the most attention, with a read rate of about 1000 times a week on a slow week and 2,000 on a busy week.
Thank you to all my visitors, and to the 1,000,000th viewer. Keep reading! edz
April construction data update
Total construction spending for 2024 is forecast up 9%, down slightly since last update. Residential was reduced, expect now only +4%. Nothing else changed much. Nonres Bldgs forecast up 11%. Nonbuilding up 15%.
Jobs growth from May’23 thru May’24: Total Jobs +3.2%; Nonres Bldgs + 4.7%; Nonbldg Infra +3.6%; Residential +2.9%. In all cases most of the growth was in 2023.
Residential construction spending cash flows indicate a drop of 4% over the next 6 months, but then a slow steady climb of +10% in the following 12 months.
Nonresidential Buildings construction spending cash flows indicate we have settled in a temp flat top for the next 6 months, down 2% from the Feb’24 peak. Then we get a 5% annual rate of increase for the next two years.
Non-building Infrastructure cash flows are not showing any monthly pullback for the next two years. Spending adds 8% from now until year end and then adds 10% in 2025.
Baseline Spending Adjusted to Constant 2019$
Since the end of 2019, (in Dec 2019 spending hit $1,464T) total construction spending as of APR’24 is up 46%. The real Volume of Business is spending minus inflation. Inflation adds nothing to the volume of business. After inflation, Volume is up only 6%. Total jobs are up 9%.
Since the end of 2019, Residential volume is up 7%. Rsdn jobs are up 15%. Nonres Bldgs Volume is up 12%. Nonres Bldgs jobs are up 5%. Nonbuilding volume is down 8%. Nonbldg jobs are up 5%.
A follower commented to me they thought total adjusted spending would be higher. At the time, we were referencing Constant2019$. When we set base year to 2024, all other years would be compared in Constant2024$. All years would appear higher, but the percent change yr/yr would be the same. Here’s the same data as the table above, except the Index year has been changed to Constant 2024$.
Baseline Spending Adjusted to Constant 2024$
Nonresidential Bldgs volume (spending after adjusting for inflation), increased 28% in the last 18 months. Nonres Bldgs jobs increased only 7%, and hours worked (OT) did not increase. That leaves 21% of Nonres Bldgs volume that was put-in-place without a balancing increase in jobs. The added work got absorbed into the existing workforce. That’s a very hard fact left unexplained by the argument that there are jobs shortages. As one reader commented, there are some reasons why more $ value of work is counted at the jobsite with less jobs, i.e., prefab, but not 21% of all Nonres Bldgs work.
This plot below removes spending, shows a better scale comparison of Jobs to Volume.
The above plot is taken from this 2011-2025 plot below. Looking back to 2011 shows how consistent jobs and volume growth was from 2011 thru 2019. This plot shows Construction Jobs growth since mid-2020 occurring at the same slope (rate of growth) as 2011-2019. Construction jobs grow avg 3.5%/year, even if volume growth varies. The Pandemic set jobs growth back by almost 2 years.
The next three plots Jobs vs Volume Growth, are set to zero start at Dec 2019. Last month, MAR data, I showed these same plots going back to start Jan 2011. Here’s a link to Mar data for those 2011 plots.
I’ve recently read comments on Twitter X that a slowdown in new residential starts is causing a drop in residential employment, and that leads a recession. Well, residential jobs are up 1% year-to-date, up 2.5% from May’23 to May’24, and right in line with average jobs growth since 2011. There is a slowdown in the rate of growth of constant $ volume. Although May starts are the lowest since November, that slowdown from May to December is only 3.5%. This may not be a big enough drop in volume to initiate a drop in jobs. We’ve seen other years (2018) where a small volume drop is not followed by a drop in jobs and never does the drop in jobs equal the magnitude of the drop in volume.
Construction Data MAR Briefs 5-5-24
Updates to Forecast, spending, starts, inflation, jobs
SEE ALSO Construction Analytics Outlook 2024
A side note, before I begin with the economic data, sometime within the next few weeks, I expect by May 31st, this blog will record the 1,000,000 view. Nearly 500,000 people read on average 2.1 articles every visit. Inflation articles draw the most attention, with a read rate of about 1000 times a week on a slow week and 2,000 on a busy week. Thank you to all my visitors. Keep reading!
2024 construction spending will be measured to the avg of 2023, $1980 bil. The average Seasonally Adjusted Annual Rate (SAAR) for 2023 is the total spending for 2023, but is was lower in Jan and higher by Dec. By Dec the SAAR was already 6% higher than the average for 2023. So we began 2024 with Dec spending at a SAAR 6% above avg 2023.
As of MAR, the total SAAR is 8.1% above 2023. Rsdn is +5.4%, Nonres Bldgs +10.0%, Nonbldg +10.8%. If growth stalls at the current level for the rest of the year, meaning, if we were to end the year with the SAAR unchanged from today, then we would finish with these gains for 2024. The trend in most cases is up, so I expect end of year we will be a little higher than today.
5-24-24 Construction Starts $. Even though Nov and Mar were low, the last 9 months of Nonresidential Bldgs new starts $ was by far the 2nd strongest period of new Nonres Bldgs starts on record, averaging an annual rate (SAAR) of $435 bil. Best since 2nd half 2022 (avg $490 bil.). The last 9 months of Residential $ starts (annual avg $395 bil.) was the best since 1st half 2022 (avg. $438 bil.).
2024 construction spending for Nonres Bldgs, as of MAR, measured to the 2023 avg, is now up +10.0% and trending up. We began 2024 with Nonres Bldgs Dec spending at a SAAR 6% above avg 2023. The American Institute of Architects (AIA) Consensus for Nonres Bldgs, published at the beginning of the year, averages +4% growth over 2023. Only one of the 10 forecasts for Nonres Bldgs spending in the AIA 2024 Consensus is still above the current reading. So, I think it’s safe to say, the AIA Consensus was low right from the very start.
The trend in Nonres Bldgs construction spending is up 18 of the last 19 months and continues up for the next 12 months. To fall to the AIA Consensus average of +4% for the year from the current SAAR, up +10.0%, the remaining 9 months of 2024 would need to fall from the current +10% to average only +2% higher than 2023. It may not be apparent, but that is a continuous decline of more than 1.5% every month for the next 9 months. That’s like falling off a cliff next month and not being able to get up. That’s unrealistic. Unless something sets off a deep recession similar to 2009, that will not happen.
At the beginning of 2007-2010, the first sign of recession for construction was a decline in 2007 of 25% in residential starts. Then in 2008 residential starts fell 40%. In 2009, both residential starts and nonresidential buildings starts fell 30%. Nonbuilding starts fell only 6%. By 2010 starts were increasing. But spending lags starts. Residential spending fell 60% from 2006 to 2009. Nonresidential buildings spending fell 33% from 2008 to 2010.
Although nonres bldgs starts fell 18% in 2020 and residential starts fell 11% in 2023, neither led to a devastating drop in spending as recovery occurred quickly. There is nothing in the current outlook to indicate recession, on any horizon. This forecast does not anticipate a recession.
Since the end of 2019, (in Dec 2019 spending hit $1,464T) total construction spending as of MAR’24 is up 46%. After inflation, Volume is up only 6%. The real Volume of Business is spending minus inflation. More than 85% of the spending growth since Dec. 2019 is inflation. If current projections hold, the total business volume through year 2024 will have grown 10% since 2019. ALL business plan forecasts and labor demand should be based on this 10% growth. Inflation adds nothing to business volume.
Over the last 9 months, residential new starts (as reported by Dodge) averaged the highest since the peak high in the 1st half of 2022. For Q1’2024, residential starts are 27% higher than Q1’2023. Currently, residential starts for 2024 are averaging 8% higher than the total in 2023. Residential spending peaked at an all-time high in Q2’22. Spending has been level or increasing the last few months at a rate 6.5% lower than the peak, but at a rate 55% higher than Dec 2019. Due to the short durations in residential building, fluctuations in starts are more quickly apparent in spending. Expect both nominal and real (inflation adjusted) spending to continue increasing thru the 1st half 2024, then drop back slightly in the 2nd half 2024. Spending is expected to increase 7% in 2024 over 2023. Volume after inflation should grow 3%.
Single Family spending YTD through Mar. is up 16% from Q1’2023. Single Family rate of spending through Mar. is up 11% over the average (total) spending in 2023. Multi-family spending for Q1’24 is up 6% from Q1’23 and is 2.5% lower than peak spending in Aug ’23, however it’s still up 1% over the avg spending in 2023. These are all nominal values, so real growth is lower. But residential inflation for 2023 was only 3%, so not much lower.
An avg spending curve for long-duration Non-bldg Infra is 15:30:30:20:5. The greatest spending impact does not show up until year two and three after the year in which the projects start. Example: If 2024 posts $100bil in new starts for Infrastructure, only $15bil of that gets put-in-place in 2024. $30bil would get put-in-place in 2025 and 2026.
Plots below compare volume growth to jobs growth. Notice the slope of the increase in jobs is fairly constant, regardless of changes in volume growth.
In the past 18 months, Nonresidential Buildings construction spending increased 37%. Nonres Bldgs JOBS increased only 7%. Normally, this would be explained by inflation, but in this case after adjusting for inflation volume still increased 28%. 18 months, +28% volume, +7% jobs.
Jobs and volume of work should be moving together, evenly. The construction industry has been saying jobs shortages, and yet over an 18mo period, the nonresidential bldgs sector added 20% more volume of work than added jobs. Seems to me that would indicate that volume was absorbed by existing jobs. If there were a significant jobs shortage, either the existing crew would need to work overtime, hours worked would have increased, or the work would not have been put-in-place and would potentially have been delayed or postponed. Neither happened. The fact that the work was put-in-place would indicate that the existing workforce readily absorbed the excess workload.
Since 2016, TOTAL construction spending has increased 63%, but after inflation, business volume increased only 6%, or 1%/yr. From 2016 to 2023, jobs increased 2.5%/yr. When jobs are increasing at a greater rate than the volume of work, productivity is declining. That is shown on these plots when the jobs line is above the volume of work line. Volume and jobs should be moving together.
In 2024, construction volume may increase 6%. Don’t expect jobs to increase 6%.
Since 1980, the fastest rates of growth in construction jobs were 1983-85 avg 6.0%/yr. and 1994-99 at 5.4%/yr. All other plus years averaged +3.2%, with only six years above 4%.
Since 2000, (excluding negative yrs, all associated with recessions) construction jobs growth is 3.3%/yr. and average real volume growth is 3.4%. I would expect future jobs growth to remain within the historical averages, somewhere in the 3%-5% range.
Construction Analytics 