One year after Stargate, Port Washington sellers are taking less than they ask. Three nearby cities are not.

First, what is a sale-to-list ratio

The sale-to-list ratio is the final sale price divided by the home's original list price, usually expressed as a percentage. It is the single clearest measure of who has negotiating power in a local housing market.

If a home lists at 350,000 dollars and sells for 357,000, the ratio is 102 percent. The buyer bid above asking; the seller had the leverage. If that same home lists at 350,000 and sells for 340,000, the ratio is 97 percent. The seller accepted a concession; the buyer had the leverage.

Average ratios above 100 percent mean a seller's market. Below 100 percent is a buyer's market. Watching the ratio move over time, in one city compared to its neighbors, tells you whether bargaining power is shifting -- even when absolute home prices are still rising with the broader market.

What changed in Port Washington

In February 2025, OpenAI and Oracle announced Port Washington, Wisconsin as one of five new Stargate sites. On December 18, 2025, Vantage Data Centers broke ground on a 672-acre, 902 MW first phase. Total project cost: 15 billion dollars. On April 7, 2026, voters passed the first anti-data-center ballot measure in United States history, 66.4 percent in favor, requiring public approval for any tax increment financing district over 10 million dollars.

In the months between those events, the Port Washington housing market changed shape in two specific ways.

First, inventory rose 36 percent above its 2024 baseline. More homes sat on the market. Second, the sale-to-list ratio dropped from 101.4 percent in 2024 to 98.6 percent in early 2026. Port Washington sellers were no longer getting asking price. They were taking concessions.

Absolute sale prices kept rising. Median sale price went from 343,500 dollars in 2024 to 440,000 in early 2026, a 28 percent jump in line with regional appreciation. Homeowners who sold did not lose money; they gained with the market. What changed was how much of what they asked for they actually received.

In November 2025, one month before Vantage broke ground, Port Washington recorded 45 active listings, 5 homes sold, and a sale-to-list ratio of 95.1 percent. No other month in the eleven years of available Redfin data for the city showed that combination.

We checked the three nearest cities in the same county for comparison. They did not move the same way.

Three cities 7 to 15 miles away

Grafton (7 miles south, population 12,000), Cedarburg (10 miles south, 12,000), and Mequon (15 miles south, 25,000) are all in Ozaukee County, all sharing the same regional economy, the same labor market, and the same Federal Reserve interest rates. We used them as control cities.

From 2024 to early 2026, Grafton's monthly inventory went from 54.9 to 60.0 to 39.0. Cedarburg's went from 19.3 to 19.6 to 12.5. Mequon's went from 63.8 to 62.2 to 47.5. All three held flat or fell. Port Washington, over the same period, went from 22.5 to 30.5 to 33.0.

Sale-to-list ratios in the three control cities held steady above 100 percent -- sellers still getting asking price or a bit more, even as 2026 opened. Port Washington and Saukville, the two cities closest to the Stargate campus, fell below 100 percent.

On a 340,000 dollar home -- roughly the Port Washington median at the announcement -- a sale-to-list ratio of 98.6 percent versus 101.4 percent the prior year means the seller takes home about 9,500 dollars less relative to what they asked. The home might still have appreciated in absolute terms. What shifted was the final negotiation at the closing table.

The chart and the full methodology are at poweredbywho.com/research/port-washington-housing.

We tried to replicate it in Lebanon, Indiana. The result surprised us.

Meta is building a 1,500-acre, 1 gigawatt data center called Project Domino in the LEAP Research and Innovation District in Lebanon, Indiana -- roughly the same scale as Stargate in Port Washington. We ran the same check on Lebanon against three Indianapolis-metro control cities (Brownsburg, Whitestown, Zionsville).

Lebanon's sale-to-list ratio did not move. It was 98.0 percent in 2024, 98.0 percent across the post-announcement period, and 99.0 percent in early 2026. Port Washington dropped 2.8 percentage points. Lebanon did not drop at all.

But Lebanon's inventory jumped 75 percent, from 26.8 homes on the market in 2024 to 47.0 during the post-announcement period. That is a bigger inventory shift than Port Washington saw. And Lebanon's median price held up at plus 2.4 percent from 2024 to 2026 while the three control cities fell anywhere from 6 to 16 percent.

Our interpretation: the same underlying behavior may be showing up in both cities, just in different metrics. The sale-to-list ratio is sensitive only when a market was previously tight. Lebanon was already a buyer's market in 2024 at 98 percent sale-to-list; there was no further downside for that metric to register. Inventory is the cleaner signal in soft markets. In Lebanon, inventory is where we see it.

We want to be honest that this is an interpretation. Other possible explanations: Lebanon's LEAP district was a state-designated innovation zone that broadcast industrial use in advance, reducing the announcement shock. There has been no ballot measure, no organized ballot campaign, and less national news coverage than Stargate got. Lebanon's market is deeper and more liquid than Port Washington's, which can dampen signal in any one metric. The Indianapolis metro has been a stronger housing market than the Milwaukee metro, absorbing some of the pressure.

There are a number of factors it could be, and we are continuing to research. The point is not that every data center announcement moves every metric the same way. The point is that running this check for your city is quick, free, and might tell you something worth acting on.

How we looked, so you can look too

If a data center is being proposed in your city, you can run this same check. You do not need a data team.

Step one: pull Redfin data. Go to redfin.com/news/data-center. Filter to your city. Download the CSV. It has monthly sale-to-list, inventory, homes sold, and median sale price going back about a decade. Do the same for three to five control cities within your county or metro -- ideally ones similar to yours in size and housing stock.

Step two: pick a before period and an after period. We used the 12 calendar months of 2024 as the before (12 months is enough to average out seasonality). For the after, start one month after the announcement to give the market time to react. Compute the period averages for each city. Do not cherry-pick months.

Step three: compare three metrics, not one. Sale-to-list, inventory, and median price tell you different things. Sale-to-list moves when negotiating power changes. Inventory moves when more sellers list or buyers pause. Median price reflects both mix and underlying demand. If any one of those moves in your treatment city while the controls do not, you have found something worth asking about.

Step four, optional but powerful: pull parcel data for the population directly exposed. Most states publish their statewide parcel database through an ArcGIS REST API. Wisconsin's is hosted by the State Cartographer's Office. We queried every residential parcel within 1 mile of the Stargate campus coordinates and got 1,581 parcels. Median assessed fair market value: 261,200 dollars. Total residential value: 423 million dollars. Those are the families whose primary household wealth is tied to what happens next.

Why run this check in the first place

Developers are required to disclose a long list of project characteristics -- megawatts, acreage, water use, tax abatement requests. They are not required to disclose what might happen to nearby home values. Neither is the city council voting on the rezoning.

In every community where a data center has been proposed, we have seen the same set of pre-decision claims from developers: the facility will be invisible, the water use will be managed, the noise will be minimal, property values will be unaffected. None of those claims are backed by the kind of specific, city-level data that residents could actually use to pressure-test them.

Redfin has monthly data for every US city for about a decade. State parcel databases have every residential property, every assessed value, every owner. The information is already public. The only thing missing, usually, is someone willing to pull it and compare.

This sort of study may be helpful in your data center fight. If you live near a proposed project and want to run this check for your city, we will help -- send the zip code and the announcement date to george@poweredbywho.com and we will send back the same table we built for Port Washington. Bring it to your next planning commission meeting.

What we are watching next

The Port Washington market has not yet stabilized. Two months of 2026 data is not enough to tell whether the shift continues, reverses, or plateaus. We will keep updating this page as Redfin publishes new months.

We are also setting up the same comparison for the Saline Township, Michigan Stargate campus (1.4 gigawatts, 575 acres, under construction, $16 billion in financing being finalized as of April 2026) and for the Augusta Township, Michigan project (1 billion dollars, on the August 4, 2026 ballot). Both are in Washtenaw County.

If the Port Washington sale-to-list pattern appears again somewhere, that is one data point toward generalization. If it does not but inventory moves instead, as in Lebanon, that is another finding worth saying out loud. Either way, we will show the numbers.