Higher productivity to start 2026 just means more profits, not better wages

Whether it's added AI investment or just worker improvements, the US saw higher productivity to start off 2026.

Labor productivity, or output per hour, is calculated by dividing an index of real output by an index of hours worked by all workers, including employees, proprietors, and unpaid family workers. During the current business cycle, starting in the fourth quarter of 2019, labor productivity has grown at an annualized rate of 2.1 percent, reflecting a 2.5-percent rate of growth in output and a 0.4-percent rate of growth in hours worked. The 2.1-percent annualized rate of nonfarm business productivity growth in the current business cycle thus far is higher than the 1.5-percent rate of the previous business cycle, from the fourth quarter of 2007 through the fourth quarter of 2019, and just below the long-term rate of 2.2 percent since the first quarter of 1947.

This graph from the Bureau of Labor Statistics explains a lot, showing that productivity has been responsible for virtually all increases in nonfarm output for 3 years.

And it’s even more so in the manufacturing sector, where hours works have consistently declined over the last 3 years, including at the start of 2026.

Manufacturing sector labor productivity increased 3.6 percent in the first quarter of 2026, as output increased 3.3 percent and hours worked decreased 0.4 percent. In the durable manufacturing sector, productivity increased 5.3 percent, reflecting a 5.4-percent increase in output and a 0.1-percent increase in hours worked. Nondurable manufacturing sector productivity increased 2.0 percent, as output increased 0.9 percent and hours worked decreased 1.0 percent. Total manufacturing sector productivity increased 1.7 percent from the same quarter a year ago.

Here’s an extra kicker – workers didn’t get financial benefits from being more productive in the first three months of 2026.

BLS calculates unit labor costs as the ratio of hourly compensation to labor productivity. Increases in hourly compensation tend to increase unit labor costs and increases in productivity tend to reduce them. Real hourly compensation, which takes into account consumer prices, decreased 0.5 percent in the first quarter of 2026, and increased 1.4 percent over the last four quarters. The labor share, which is the percentage of output that accrues to workers in the form of compensation, was 54.1 percent in the first quarter of 2026, the lowest recorded value since the series began in 1947.

Oh? The lowest share of worker compensation in at least 79 years? It’s almost like owners have used technological advances to grab profits for themselves and keep it away from the workers that made it possible! Especially in the 2020s.

That loss of real worker compensation isn’t going to get better in Q2, as Friday's tepid increase of 0.2% isn't going to come close to keeping up with the inflation we had in April. And with gas prices going up another 25 cents a gallon nationwide in the first week of May, that declining situation for US workers doesn't seem to be ending any time soon.

Which makes you wonder when a lot of workers start asking why they should go along with productivity "improvements", since it doesn’t seem to give them much in return.

More jobs in April, but fewer Americans working, and wage growth still lame

Even as gas prices go through the roof and everyday Americans feel bad about the economy, the job market kept moving along in April. Looking at Friday's US jobs report, it was a second straight month of six-figure job growth, something that didn’t happen in all of 2025.

Total nonfarm payroll employment edged up by 115,000 in April, after showing little net change over the prior 12 months. In April, job gains occurred in health care, transportation and warehousing, and retail trade. Federal government employment continued to decline.

In April, health care added 37,000 jobs, in line with the average monthly gain of 32,000 over the prior 12 months. Over the month, job gains occurred in nursing and residential care facilities (+15,000) and home health care services (+11,000).

Transportation and warehousing employment increased by 30,000 in April, reflecting a gain in couriers and messengers (+38,000). However, employment in transportation and warehousing is down by 105,000 since reaching a peak in February 2025.

I was trying to figure out what these Couriers and Messengers jobs are, and why they grew by so much last month. Here’s how the Bureau of Labor Statistics defines it.

Pick up and deliver messages, documents, packages, and other items between offices or departments within an establishment or directly to other business concerns, traveling by foot, bicycle, motorcycle, automobile, or public conveyance. Excludes “Light Truck Drivers”

Looks like a lot of these jobs are in medical labs and hospitals, along with regular delivery services. And Wisconsin had a surprisingly large amount of these types of jobs 3 years ago, with the 3rd highest concentration in the country, and by far the most in the Midwest.

But this sector has barely above 1 million jobs nationwide, and its “growth” in the last 2 months has been entirely seasonally adjusted. The BLS's models counted on more layoffs in March and April, and when that didn’t happen, the seasonally adjusted numbers went up.

Couriers and Messengers job change, seasonal vs non-seasonal
March 2026
Seasonally-adjusted +21,700
Non-seasonally adjusted -14,700

April 2026
Seasonally-adjusted +37,900
Non-seasonally adjusted +1,400

Look, there are a lot of industries that have their job numbers deflated in April as the weather warms, so this isn’t a big deal and I’d call the overall jobs numbers for April “decent to good”. But I will also say that when 1/3 of your job “growth” is due to a seasonal adjustment, it’s not great.

And the household survey wasn’t good at all.

The unemployment rate was unchanged at 4.3 percent in April, and the number of unemployed people changed little at 7.4 million. Both measures changed little over the year….

Both the labor force participation rate, at 61.8 percent, and the employment-population ratio, at 59.1 percent, changed little in April. These measures edged down over the year after accounting for annual population control adjustments.

That 59.1% figure in the employment-population ratio is the lowest non-COVID rate since September 2014, and the 61.8% participation rate is the lowest non-COVID level since the start of 1977!

This "edging down" means the number of people in the US labor force and the number of people ID’ing as employed have each declined by more than 1 million in the last 12 months.

I've got the trendline up on these because the big drop at the start of 2026 is due to the benchmarking that happened in March 2026, where the BLS followed lower population growth figures by saying the US labor force and number of people employed had shrunk more than first reported.

That's due to immigration has been limited under Trump 2.0 and older workers keep retiring and not being replaced. And White House adviser Kevin Hassett tried to turn those declining labor force and participation numbers into a good thing when asked about it on Fox News.

FOX NEWS: The negative is that labor supply is shrinking. You've got baby boomers retiring, you've got less illegal immigration, you got more deportations. How do you factor that in? HASSETT: I don't think that's a negative. I think it's a positive. You've got more good jobs for legal Americans.

[image or embed]

— Aaron Rupar (@atrupar.com) May 8, 2026 at 8:38 AM

Except that’s not happening, Kevin! Nearly 1 million of the drop in the labor force are from native-born Americans, and more than 1.1 million of the drop in employment over the last 12 months and is from native-born Americans (with more than 1.0 million being native-born American men).

In addition, wage growth sucks for American workers.

In April, average hourly earnings for all employees on private nonfarm payrolls rose by 6 cents, or 0.2 percent, to $37.41. Over the year, average hourly earnings have increased by 3.6 percent. In April, av>erage hourly earnings of private-sector production and nonsupervisory employees rose by 11 cents, or 0.3 percent, to $32.23.

While slightly above the 3.4% year over year increase in March, it's still a lot lower than what we were seeing during the last bout of gas price spikes in 2022.

This wage increase is also barely above the 12-month change of 3.3% between March 2025 and March 2026 in the Consumer Price Index, which means wage growth will likely fall below the 12-month inflation rate that shows up in April’s CPI report, which comes out next week.

So sure, the payrolls numbers look good on the April jobs report, and there's no evidence that we are in any kind of overall recession. But a number of other employment indicators tell us that things are far from booming, and many Americans (working and otherwise) aren't able to keep up with the higher prices that have become an increasing part of life in May 2026. Which goes a long way toward explaining why American consumer sentiment is at its lowest point in decades. And what could come along to change that for the coming months?

Well, turn around consumer sentiment positively, that is. If we start seeing unemployment being closer to 5% than 4% and this bump up in job growth reverse....LOOK OUT!

Gas price spending patterns the latest evidence of an economy brought to you by the Letter K

With gas prices jumping above $4.50 nationwide, there was a timely release from the New York Fed on how high gas prices hit differently, depending on how much money you’ re making.

...We divide all households into low-income (earning less than $40,000 a year) households, middle-income (earning between $40,000 and $125,000) households, and high-income (earning over $125,000) households, as we have done in previous blogs on the divergence in retail spending growth by income level and potential explanations. The high-income households represent approximately a third of all households. In the panel chart below, we present how the rises and falls in spending and consumption varied for the three income types. We deflate gas station spending for each income group using a demographic-specific gasoline price deflator.

We see that the three income categories had very different experiences during the March 2026 energy price shock. Low-income households increased their nominal gas spending by the least (12 percent). However, this was accomplished because they cut their real gas consumption the most (7 percent). On the other hand, high-income households increased their nominal gas spending by the most (19 percent) in a large part because they reduced their real gas consumption the least (1 percent). Middle-income households had intermediate increases in nominal spending and decreases in real consumption at gas stations. Thus, the K-shaped consumption pattern in both nominal and real gasoline spending was strongly evident in March 2026.

These divergences in the response to an energy price shock are not unique to the month. Four years ago, energy prices rose and remained elevated during the spring and summer of 2022 when the Russia-Ukraine war disrupted energy markets. The magnitude of the initial Russia-Ukraine gasoline price shock was broadly similar to the current one, but it lasted longer to date and ramped up over time. As we see from the panel chart below, between January and July 2022, nominal gas spending rose more for high- and middle-income households than it did for low-income households, while real gas consumption declined less for high-income households than it did for middle-income and low-income households. Notably though, while directionally similar, the magnitudes of the gaps (both for nominal and real spending) were noticeably smaller than the corresponding gaps we see in March 2026. Nominal and real spending rebounded to their pre-shock levels after energy prices declined in late 2022.

Which tracks, as lower-income people are the ones most likely to be close to the edge and have to either dip into savings or cut back on spending if any expense goes up. Richer people may get annoyed by the higher prices, but don’t often have to change their spending or investing habits.

We also saw more evidence that economic activity keeps growing without necessarily seeing people getting hired in the process. After the Institute for Supply Management released information showing that manufacturing in America continued to increase output while shedding jobs, we saw the same thing happen in ISM’s report about the nation’s services industry.

In April, the Services PMI® registered 53.6 percent, 1.1 percentage points above its 12-month moving average of 52.5 percent. A reading above 50 percent indicates the services sector economy is generally expanding; below 50 percent indicates it is generally contracting.

A Services PMI® above 48.1 percent, over time, generally indicates an expansion of the overall economy. Therefore, the April Services PMI® indicates the overall economy is expanding for the 71st straight month. Miller says, “The past relationship between the Services PMI® and the overall economy indicates that the Services PMI® for April (53.6 percent) corresponds to a 1.7-percentage point increase in real gross domestic product (GDP) on an annualized basis.”….

Employment activity in the services sector remained in contraction in April for a second month in a row. The Employment Index registered 48 percent, up 2.8 percentage points from the March figure of 45.2 percent and 0.6 percentage point below its 12-month average of 48.6 percent. Comments from respondents include: “We are experiencing large capital investment projects, requiring an increase in sourcing and material purchasing as well as head count” and “Beyond layoffs from the end of the year, there has been an uptick in attrition.”

The “attrition” comment makes sense to me, as we have seen low job growth for the last year, but there’s also a remarkably low amount of layoffs, including the lowest amount of (seasonally-adjusted) new unemployment claims since 1969 at the end of April?

When you have an economy that's as two-track like this, you’re going to get a situation where the typical American thinks the economy isn’t going well and their personal situation is getting worse, but the overall numbers still look good and the stock market loves it (hey, the DOW was back at 50,000 at one point today!).

And as long as enough money is getting sent around and large amounts of people aren’t losing jobs or cutting their consumption all at once, the musical chairs continues. Until someone decides it's time for the music to stop.

Manufacturing orders up, but so are prices. And jobs are still going down

More data showing that economic growth continued in March, as orders and shipments in manufacturing kept increasing.

New orders for manufactured goods in March, up four of the last five months, increased $9.1 billion or 1.5 percent to $630.4 billion, the U.S. Census Bureau reported today. This followed a 0.3 percent February increase. Shipments, up five of the last six months, increased $8.8 billion or 1.4 percent to $633.9 billion. This followed a 1.7 percent February increase. Unfilled orders, up twenty of the last twenty-one months, increased $1.6 billion or 0.1 percent to $1,540.9 billion. This followed a 0.1 percent February increase. The unfilled orders-to-shipments ratio was 6.88, down from 6.92 in February. Inventories, up six consecutive months, increased $5.8 billion or 0.6 percent to $956.3 billion. This followed a 0.1 percent February increase. The inventories-to-shipments ratio was 1.51, down from 1.52 in February.

New orders for manufactured durable goods in March, up following three consecutive monthly decreases, increased $2.7 billion or 0.8 percent to $318.9 billion, unchanged from the previously published increase. This followed a 1.2 percent February decrease. Computers and electronic products, up eleven of the last twelve months, led the increase, $1.0 billion or 3.6 percent to $29.6 billion. New orders for manufactured nondurable goods increased $6.5 billion or 2.1 percent to $311.5 billion.

Shipments of manufactured durable goods in March, up six of the last seven months, increased $2.4 billion or 0.7 percent to $322.4 billion, unchanged from the previously published increase. This followed a 1.6 percent February increase. Machinery, up four of the last five months, led the increase, $0.9 billion or 2.3 percent to $41.6 billion. Shipments of manufactured nondurable goods, up four consecutive months, increased $6.5 billion or 2.1 percent to $311.5 billion. This followed a 1.9 percent February increase. Petroleum and coal products, up three consecutive months, led the increase, $5.6 billion or 9.9 percent to $62.1 billion.

Which indicates that manufacturing was still going strong at the end of Q1 2026, at least on the demand for products.

And last week, the latest ISM Manufacturing report said that growth of output kept going into April.

The report was issued today by Susan Spence, MBA, Chair of the Institute for Supply Management® (ISM®) Manufacturing Business Survey Committee.

“The Manufacturing PMI® registered 52.7 percent in April, the same reading as March. The overall economy continued in expansion for the 18th month in a row. (A Manufacturing PMI® above 47.5 percent, over a period of time, generally indicates an expansion of the overall economy.) The New Orders Index expanded for the fourth straight month after four straight readings in contraction, registering 54.1 percent, up 0.6 percentage point compared to March’s figure of 53.5 percent. The April reading of the Production Index (53.4 percent) is 1.7 percentage points lower than March’s reading of 55.1 percent. The Prices Index remained in expansion (or ‘increasing’ territory), registering 84.6 percent, a 6.3-percentage point jump from March’s reading of 78.3 percent. In the last three months, the Prices Index has increased 25.6 percentage points to reach its highest level since April 2022 (84.6 percent). The Backlog of Orders Index registered 51.4 percent, down 3 percentage points compared to the 54.4 percent recorded in March. The Employment Index registered 46.4 percent, down 2.3 percentage points from March’s figure of 48.7 percent,” says Spence….

Spence continues, “In April, U.S. manufacturing activity remained in expansion territory, growing at the same pace as the month before. Of the five subindexes that make up the PMI®, the New Orders and Supplier Deliveries indexes indicated faster growth compared to the previous month, the Production Index grew at a slower rate, and the Employment and Inventories indexes remained in contraction.

But did you notice that “prices paid” part? A number over 50 would indicate prices went above the previous month’s level. But 84.6% is a whole different level, back to the days of Spring 2022, which had the highest price increases in over 40 years. And the ISM report says the higher prices are happening in every industry of the sector.

So if prices are jumping by so much, is a 1.5% increase in the cost of new orders for March or a 1.4% increase in the cost of shipments all that impressive? And looking inside the ISM numbers, you can see evidence of more job losses in April and the prospect of slower overseas business along with those higher costs and prices.

Increased production with lower employment can only be possible with higher increases in productivity. But that's supposed to bring prices per unit down , and instead prices are going up at their highest amounts in 4 years...and 40+ years.

This is a recipe for higher profits for manufacturers on top of what was already back near record levels in the industry at the end of 2025.

Hey lookie there - profits are back at the peaks of the "inflation" times of 2022! What a coincidence!

Let's see if my instinct is right, where these higher prices lead to record profits for businesses, while everyday workers pay. And when do the increased orders and demands stop when no one can afford the combination of higher prices and lower wages and fewer jobs.

American consumers kept spending and paying more in March. With savings going down further

Not surprisingly, we got more data showing that inflation spiked up in the month after the bombs started falling in the Middle East.

Consumers faced escalating prices in March as the Iran war sent oil soaring and created a new level of challenges for the Federal Reserve, according to a batch of reports Thursday that showed economic growth slower than expected and a generational low in layoffs.

The core personal consumption expenditures price index, which excludes food and energy, accelerated a seasonally adjusted 0.3% for the month, pushing the 12-month inflation rate to 3.2%, the Commerce Department reported Thursday. The readings matched the Dow Jones consensus estimates. Core inflation hit its highest level since November 2023.

Including the volatile gas and groceries components saw higher readings, with the monthly gain at 0.7% and the annual rate hitting 3.5%, also in line with forecasts.

We can blow off the gas price increase if we want, but I bet typical American consumers aren’t, and businesses won’t be ignoring the extra transportation and input costs that results. So let’s also not shrug off a PCE monthly inflation number that would be over 8% on an annual basis.

The only time we’ve seen the PCE Price Index increase by this much in a month was during the peak post-COVID inflation times between October 2021 and June 2022, when it reached 0.7% two times and exceeded it two other times. And while gas prices leveled off in the first part of April, we've now seen another jump in prices in the last week, with an increase of 35 cents a gallon in the last 7 days, and 44 cents in Wisconsin.

Digging into the full income and spending report itself, it shows that while Americans spent more in March, a lot of it was due to the rising gas prices in that month.

Personal outlays—the sum of PCE, personal interest payments, and personal current transfer payments—increased $198.6 billion in March. Personal saving was $857.3 billion in March, and the personal saving rate—personal saving as a percentage of DPI—was 3.6 percent….

The $195.4 billion increase in current-dollar PCE in March reflected increases of $132.6 billion in spending on goods and $62.9 billion in spending on services.

Take out increased gas and health care spending, and the rest of PCE went up by 0.5% instead of the 0.9% top line (not adjusted for inflation). Not bad, but not amazing either.

And for most Americans, incomes didn’t even go up that much, because most of the jump in income came from non-working income.

That rise in non-working income reflects farm owners seeing their incomes nearly double in March, as a large swath of Farm Bridge Assistance payments went out that month, trying to make up for damage caused by Trump tariff policy. Take that out of the equation, and it’s only a 0.3% increase in incomes for March - well below inflation. In addition, disposable incomes have been inflated due to increased tax refunds in the first months of 2026. That’s not going to be the case after April, so that’ll be another headwind.

The savings rate also continued to fall in March, as the 0.6% increase in personal incomes didn’t match the increase in spending. This continues a general decline that we've been in since the start of 2024.

With costs staying high for the foreseeable future, don’t be surprised if the savings rate falls below 3% soon, which also would be a flashback to 2022’s peak inflation…. without the stimmy checks and higher incomes of 2020 and 2021 bulking up savings before that. The other time we fell below 3% saving? The mid-2000s, right before the debt-pumped Bubbles popped and the Great Recession began.

Put all this together, and add in the fact that unemployment claims are somehow going lower at the end of April, and I’d argue that we are in a “’flation” stage of the stagflationary cycle that is reality in much of America. While the Federal Reserve stayed pat this week and didn’t even change their wording about risks or plans for future rate cuts (albeit with some dissenters on the Board). I have to think that more declines in savings and continued inflation for Q2 2026 will lead to changes at the next Fed meeting in mid-June.

And while American consumers and Wall Streeters continue to shake off the higher prices and stagnant incomes, there has to be a point where they deal with reality and make some adjustments of their own. Which is going to trigger a lot of other economic effects.

Americans say things are bad because of what is, but especially what COULD happen

It's no secret that Americans have been increasingly unhappy about the US economy, with consumer sentiment at the lowest level measured and consumer confidence mired near the lowest levels in over a decade.

Not surprisingly, this gloominess shows up in Gallup’s annual survey of Americans about their finances. In that survey, less than half of Americans said their current financial situation was good or excellent for the fifth year in a row.

But worse is that the percentage of Americans saying their financial situation is poor is at its highest level in 15 years, and matches what they were saying during the Great Recession.

In addition, Gallup says that they are seeing the highest percentage of Americans saying their finances are getting worse in the 25 years that they’ve asked the question. And that percentage has gone up 8% in the two years since April 2024 – the last full year Joe Biden was President.

And yet the stock market has recovered all of its losses from March and reached new highs in this month. I know the stock market isn’t the real economy, but these traders have seemed especially disconnected from reality in recent weeks, trading up or down based on whatever BS our clueless President may post.

That said, it does seem like there has been a break in other parts of the financial markets this week, where there is a realization that the Hormusz blockades aren't going away tomorrow, which means disruptions to oil supply are going to be a continuing concern.

Oops, sorry. That’s the 10-year US Treasury Note yield over the last week, which shows that Wall Street is realizing that interest rates aren’t coming down any time soon, and that long-lasting inflation is a growing possibility.

Here’s the oil chart.

And on top of that, gas prices have now blasted past $4 a gallon nationwide with an increase of more than 5% over the last week. And while Wisconsin's average price isn't there yet, we've also had prices going up.

I saw my first pump price start with a $4___ in Madison today, so that average number likely will be significantly higher soon. But we have yet to see gasoline consumption (or most consumer consumption, for that matter) be affected by the bombs falling in the Middle East over the last 2 months. In fact, the numbers are slightly above what we’ve had over the last 3 years, despite the higher prices of today.

So if it isn't bad enough to change behavior, why are American consumers so depressed and/or angry? Another finding from the Gallup poll gave a lot of the explanation to me.

Notice that the most common concerns are about potential crises, not current ones. That is what you would see in a society where people are still maintaining most or all of their jobs, standard of living and spending levels, but also know they are teetering on the edge, where one disruption to their income stream or costs will set off a lot of distress.

It’s not surprising, is it? No costs seem to be going down, especially necessities like health care and housing, and overall inflation has generally exceeded wage increases since last Summer, A lot of people have their economic futures dependent on gig economy jobs or similar hustles, with their retirements tied to the stock market instead of being full-time employees with set-aside pensions. And there is not much of a way to get off of the hamster wheel of month-to-month financial existence for Americans in a wide swath of income levels.

So when’s the point when the average American truly gives up and cuts back? Or when businesses decide they can’t keep paying these prices and cut back on orders or employees? And if/when the actual downturn starts, how awful is that going to be, and how far is it going to spread?

Despite the bad signs everywhere and increasing strains on pocketbooks, real bad stuff in the economy hasn’t happened yet. But safe to say that in April 2026, Americans don't think things are going in the right direction, and are worried about when things take a more tangible turn for the worse.

Data center tax exemption could give away billions

The placement and construction of data centers are already a hot issue in Wisconsin due to the effects that they may have on communities and costs for local electric customers. But now there’s another way that data centers could prove costly – through a tax break that the state has to encourage this construction.

Wisconsin Watch has a good summary of what seems likely to become an exploding exemption for these projects, and the economic pros and cons of the situation.

Wisconsin is poised to forgo more than $2 billion in sales tax revenue to subsidize hyperscale data centers built by trillion-dollar companies such as Microsoft and Meta.

Data centers were granted a sales tax exemption in the 2023-25 state budget, which was approved by the Republican-controlled Legislature and Democratic Gov. Tony Evers as a way to attract economic development to the state…

State Sen. Jodi Habush Sinykin, D-Whitefish Bay, who requested the estimate, said lawmakers should discuss what the state can get in return for the exemption. She said the exemption could be tied to, for example, requirements to protect the environment.

Habush Sinykin wants the Legislature, which Republicans control, to convene what is known as an extraordinary session to discuss a variety of data center bills, rather than waiting until the next regular session in January.

Sen. Romaine Quinn, R-Birchwood, and Rep. Shannon Zimmerman, R-River Falls, introduced legislation in 2023 that led to the exemption and have proposed expanding it. They did not respond to requests for comment.

Here’s how the Legislative Fiscal Bureau describes who is eligible for the tax break, what types of expenses are able to be written off, and the type of building gets this tax break.

Under [the 2023-25 state budget], a business, and certain contractors contracting with that business, that purchases certain property for a qualified data center certified by WEDC is eligible for state and local sales and use tax exemptions for eligible costs. The exemption applies to three main "eligible costs." The first eligible cost includes the price of property used solely with building and operating the data center. The property includes the sales price of various items for computer servers, networking, energy systems, security systems, and electricity. The second eligible cost includes the price of property associated with building water cooling or conservation systems used exclusively to cool or conserve water for data centers. The systems include the price of various items such as chillers, refrigerant piping, water softeners, fans, ducting, and filters. The third eligible cost includes the price of property sold to a construction contractor that, in fulfillment of a real property construction activity, transfers the property to the owner when it becomes a component of the data center.

A "qualified data center" is one or more buildings or an array of connected buildings owned, leased, or operated by the same business entity (or its affiliate) and for which all of the following apply: (a) the buildings are rehabilitated or constructed to house a group of networked server computers in one physical location or multiple locations in order to centralize the processing, storage, management, retrieval, communication, or dissemination of data and information; and (b) the buildings create a “minimum qualified investment” in this state within five years from the certification date of at least $50,000,000, $100,000,000 or $150,000,000, depending on the population of the county in which the data center is constructed.

The business then enters into a contract with the Wisconsin Economic Development Corportation (WEDC), who certifies that the project is OK for the sales tax exemption. Here is the list of four projects that had received that signoff as of March 2026.

The Microsoft project is located in and around the part of Racine County that Foxconn was supposed to put in numerous facilities nearly a decade ago (yes, it's what replaced the Fox-con). The other 3 projects in the works are:

· Open AI, Oracle and Vantage with the beginning steps of what they claim would be a $15 billion data center campus in Port Washington, which is supposed to be completed in 2028. This is the project that caused such an uproar that the voters of Port Washington approved a referendum earlier this month that would require the voters to OK any tax incentives for future projects of this sort.
· Meta’s $1 billion data center in Beaver Dam, which is expected to be running next year.
· The $347 million Epic Hosting project in Verona that is supposed to get over the $150 million threshold for certification this year.

As of the end of 2024, only the Microsoft data center had enough expenses to be qualified. Then in 2025, Meta's new Beaver Dam data center drew enough expenses to also qualify for the tax exemption, and Epic's data center got close.

Of course, the other data center projects have a few more years to be built, so no major concerns that they won't reach the point of certification. But the sales tax exemption is already being cashed in, to the tune of $10.8 million for 2024, and at least $12.2 million in 2025 (final reports for last year still have yet to be fully submitted and certified).

As the LFB notes, Microsoft is claiming that they will eventually build over $20 billion in data centers at the Foxconn site. If that should all actually happen, a lot more than just the $10.5 million in 2024 and $12.2 million will end up being written off in sales tax for the supplies and work involved.

And if all of these projects do get built for the amount that the businesses say that they will, that’s where the LFB says the tax break could reach 10 figures.

The announcements and planned investments from these four certified companies combine for a total investment of more than $36.9 billion, spread over the life of each project (currently planned to occur between 2024 and 2028). It is estimated that an investment this size would result in $1.5 billion in initial foregone state sales tax revenue. Additional foregone state sales tax revenue of $369 million on an annual basis is estimated once these projects are completed. Expenditures eligible for the state tax exemption would also be exempt from local sales and use taxes, resulting in forgone county, city, and premier resort area tax collections, if applicable.

That would seem to blast a giant hole into tax collections for the future, a lot more than what was anticipated by LFB when this data center tax exemption was originally put into place in 2023.

[The Wisconsin Department of Revenue] indicates that, based on typical capital expenditures for the construction of a data center ($215.5 million) and adjusting for items that would otherwise be subject to sales tax, the exemption would decrease sales and use tax collections by $8,500,000 for the initial construction of facilities, with an annual sales tax reduction of $735,000 related to ongoing operating expenses of such a facility. Although there are no known facilities in the state meeting the minimum qualified investment amounts specified above, it is assumed that the provision is intended to apply to the proposed Microsoft facility, which is locating in the electronics and information technology manufacturing (EITM) zone (Foxconn). The estimated amount of foregone sales tax revenue as a result of this provision is unknown. Further, it is unknown whether WEDC would certify any new qualified data centers to be constructed during the 2023-25 biennium.

As we’ve seen, the investment into these data centers is anticipated to be a lot larger than $215.5 million. And maybe I’m missing it, but I don’t see where the sales tax numbers were adjusted down in the 2025-27 budget to account for the projected explosion of data center activity and the writeoffs that will come from it. Sure, building these data centers would likely result in a boost of construction jobs in the short term, and there is evidence that this is happening over the last year in Wisconsin.

But given that the effective state income tax comes out to something like 5% for the level of income that construction workers get, how do the additional jobs make up for the 5% sales tax exemption for the data center activities, which goes beyond worker salaries? I know, multipliers of economic activity happen with jobs, and that’s a good thing. However, we also know that the added data centers and jobs require added infrastructure costs.

And that gets to the center of the controversy comes with the building of these data centers. There are a lot of costs of added electricity and roads and other needs that may be getting shoved onto communities and local taxpayers, while the corporation gets their costs written off in this sales tax exemption, and is the main beneficiary of the infrastructure.

Which makes last week’s move from Wisconsin’s Public Service Commission all the more interesting.

The Public Service Commission of Wisconsin on Friday unanimously approved a special rate plan for data center-scale customers in We Energies’ service territory. But before doing so, they made a host of modifications to the proposal they said were aimed at protecting other customers.

The decision comes as data centers in Mount Pleasant and Port Washington could double We Energies’ energy demand by 2030. The company is preparing to spend $19.3 billion on new electricity generation over five years, according to testimony filed with the PSC.

A major concern around the country has been that some of the extra cost of establishing and using electrical service for these data centers gets pushed onto homeowners and other smaller residential customers, who end up paying more when they aren’t using more. So this PSC decision requires We Energies to lock in a 15-year agreement with large-scale users that also segregates the costs of those large-scale users.

The energy demand threshold for a customer to qualify for the rate structure was reduced from 500 megawatts to 100 megawatts, the level at which new generation projects typically require PSC approval. The commission opted to make it mandatory for eligible customers to subscribe.

Very large customers would fund and subscribe to portions of multiple new power generation projects, or entire projects. The commission removed a capacity-only subscription option that would have allowed data centers to cover 75 percent of the costs of certain generating facilities.

It’s a start, and we will see if a similar situation comes up as the PSC deals with the rates that Alliant Energy wants as part of the Meta data center project in Beaver Dam.

It’s also clear that what was expected to be a small tax break in 2023 for data centers is something that has grown into a potential budget handcuff for the coming years in our state. It may not mess things up too much this year, but if these data centers do end up getting finished and expanded, it starts to be a major tax exemption in 2027 and beyond. And it tells me that even if the AI companies come through with their promised investments in Wisconsin (instead of them being PR moves to encourage private equity suckers to keep floating them money), that tax break is yet another reason that any AI “boom” that happens is not likely to be something that has a good payoff for most of us.