New apartment construction in American cities tends toward large buildings with small units, leaving fewer new options for households with more than 2 members.

Complaints abound about the lack of apartments suitable for a family with kids. In Milwaukee, I often hear questions like, “Who is going to live in these?” and “Where will families go?”

The truth is, even the robust growth of studio and 1-bedroom apartments has failed to meet the even more rapid rise in the number of these households. Over the past decade, the city of Milwaukee added around 9,300 1-person households versus about 2,200 new studio or 1-bedroom housing units.

Surely, not all these 1-person households will want to live in a small unit, but judging by current dwelling behavior, many clearly do. The growing gap between potential residents of small units and available inventory is larger than that for any other market segment.

Bigger buildings, smaller units

The next two graphs show annual construction statistics for the Midwest. In 2022, 63% of new apartment units came from buildings of at least 50 units, compared to just 21% in 2005. Apartment buildings with 2-4 units are hardly built anymore. They contributed just 1% of new apartments last year, down from 11% in 2005. Likewise, small apartment buildings with 5 to 9 units contributed 22% of new units in 2005 and just 4% in 2022.

As apartment buildings have grown larger, units have grown smaller by every measure. Seventy-eight percent of apartments built in 2005 contained at least 2 bedrooms, falling to 50% in 2022.

From 2000-2009, studio apartments made up about 1.5% of annual apartment construction. During the first three years of the 2020s, studios comprised 12.3% of new construction.

In 2005, 28% of new apartments were under 1,000 sq. ft. and 10% were 1,800 sq. ft. or larger. In 2022, 51% of new units were under 1,000 sq. ft. and just 2% were at least 1,800 sq. ft.

Growing demand for small units

Often, reasons for this dearth of “family-sized” apartments are given from the supply side. Zoning codes often make it hard to build the kinds of small apartment buildings that would fit in residential neighborhoods. Where room allows, bigger buildings are both more profitable. U.S. fire codes require large apartment buildings to be built around a central hallway. These “double-loaded” floor plans limit the number of multi-bedroom apartment units common elsewhere.

In cities like Milwaukee, there is also tremendous demand for small apartment units.

Even as Milwaukee’s population has fallen over the past 50 years, the number of households (or occupied housing units), has remained stable due to the growing number of people living by themselves.

Just over the past decade, the city added 9,300 1-person households and 1,500 2-person households, while losing 11,500 households with 3 or more members.

Milwaukee’s housing stock has, overall, not caught up to these changes.

Recent census estimates count about 87,000 single-member households in Milwaukee, compared with an available supply of 52,000 studio or 1-bedroom apartments.

In sharp contrast, Milwaukee holds 23,000 more 2-unit dwellings than 2-person households and 53,000 more 3-unit dwellings than 3-person households. Four-person households are closely matched to the number of 4-bedroom homes, and the number of households with 5 or more members (though shrinking) still outstrips the number of 5+ bedroom houses by about 16,000.

Of course, households do not match house sizes in the neat way those numbers suggest. Children in large families have always shared bedrooms and small households sometimes want more space. As I began writing this piece, I wondered if the relationship might run in reverse. Perhaps parents have less money to spend, and so wind up in smaller houses, while couples without children might be able to afford the largest houses.

As it turns out, this is not the case. Within the constraints of Milwaukee’s housing supply, people generally wind up choosing homes that closely match their household size.

The most common situation, involving 37% of households, is when the number of bedrooms exactly matches the number of residents. Another quarter of households have one extra bedroom. Fourteen percent have two extra bedrooms, and another 14% have 1 more person than they do bedrooms. Only 9% of households have at least two fewer bedrooms than people.

Forty percent of 1-person households live in either a studio or a 1-bedroom apartment. Another third live in a two-bedroom unit. The modal 2-person household lives in a 2-bedroom unit (41%), and just over half of 3-person households also live in a 3-bedroom home.

More of a mismatch exists for larger families, those with 4 or more members. Half of 4-member households live in a 3-bedroom house, a quarter have 4 or more bedrooms, and another quarter live somewhere with fewer than three.

Among households with 5 or more people, just 9% live in a house with 5 or more bedrooms, reflecting the lack of supply in this category. Large families are still the most likely group to live in one of these uncommonly large houses.

In a city like Milwaukee, with little available land for new construction, it comes as no surprise that most new construction is tailored to the only market segment with significant growth—adults without kids, usually living alone. Three-bedroom units abound, albeit primarily in aging housing stock.

The big challenge for builders and policymakers will be figuring out how to build enough smaller units that are appealing and affordable to the growing numbers single-member households. The large apartment buildings that dominate the current construction landscape are an important part of that. Milwaukee also needs to advance strategies like accessory dwelling units and corner quadplexes that accommodate shrinking household sizes in popular, already built-out residential neighborhoods.

Data Note

The Census Bureau does not publish a table counting households by members and bedrooms, so I computed this data from the most recent batch of census microdata. Due to limitations with the dataset, these statistics include Milwaukee County’s north shore suburbs.

Additional statistics are drawn from the 2017-21 American Community Survey, the 2008-12 ACS, decennial censuses from previous years, and the Survey on Construction.

The number of houses sold in Milwaukee fell by 32% during the second quarter of 2023, compared to the same period last year. While prices continued to rise, they did so more slowly than the inflation rate.

In “real,” inflation-adjusted dollars, the median sale price was about $10,000 below the peak in 2021, albeit still well above pre-pandemic prices.

My analysis of preliminary state transaction records shows about 1,950 “arm’s length” home sales during April, May, and June of 2023. That is roughly 900 fewer than either of the past two years and close to the total sold in the second quarter of 2020, when much of the economy was still frozen by the COVID-19 pandemic.

In my last housing market report, I described first quarter home sales as falling to “pre-pandemic levels.” Now, it’s fair to say that the housing market is even cooler than that. Sales were down about 6% through the first three months of 2023 compared to 2019. That downward trend has continued, and sales in the 2^nd quarter of 2023 were fully 18% lower than in 2019.

The median nominal price paid for a home during April-June has increased each year since 2019, but recent price growth has trailed inflation.

This year’s median price in the second quarter was $170,000. When adjusted for inflation to the value of the June 2023 dollar, the median price was $172,000 in 2022 and $181,000 in 2021.

The number of home sales declined by at least a fifth in every aldermanic district over the past year. The smallest drops came in 4^th district (covering downtown) and the 12^th (the eastern half of the near south side). The sharpest declines occurred on in the north side’s 1^st district (-45%) and the 14^th district (-39%), which covers Bay View and surrounding areas.

Inflation-adjusted prices grew, year-over-year, in 7 districts and declined in 8. The greatest decline came in the 4^th district, where the median price fell 14% between the 2^nd quarters of 2022 and 2023. The largest proportional increase occurred in the 6^th district, which includes Brewer’s Hill, Halyard Park, Harambee, and much of Riverwest. The median price jumped 39%.

As in the first quarter of this year, out-of-state landlords have switched from net buyers of Milwaukee houses to net sellers. During April, May, and June of this year, an out-of-state owner was the seller in 12% of transactions and the buyer in just 7%. A year ago, the relationship moved in the other direction, as owners outside Wisconsin bought 15% of transacted houses and sold 9%.

The behavior of Milwaukee’s out-of-state, private-equity backed landlords gives some insight into this change. After several years of rapid acquisitions, VineBrook Homes and SFR3 have both switched to selling Milwaukee houses. Transaction records show that VineBrook sold 20 houses during April, May, and June of this year. SFR3 sold 18, and neither company purchased any.

Data collected through mid-May showed that VineBrook was generally losing money on sales in 2023, while SFR3 was selling houses for far more than they originally paid. The rest of the 2^nd quarter did not alter that finding.

I was able to make direct comparisons between the original purchase and final sale prices in 17 VineBrook transactions during April, May, and June. Records show that the company sold the houses for $1,267,500 collectively, after buying them for $1,326,140. That is a loss of $58,640, or 4.4%, not including any other expenses beyond the purchase price.

I likewise identified comparable acquisition records for 16 of SFR3’s 2023 second quarter sales. Collectively, SFR3 sold these houses for $2,671,800 after buying them for $1,274,009—a profit of $1,397,791, or 101% before additional expenses.

Flippers remained active in the market. The most prolific, Rentalvest LLC, bought 17 houses and sold 12, according to state transaction records. California-based Ace Property Acquisitions LLC bought 15 and sold 16. Smart Home Solutions LLC bought 9 and sold 10.

After several years of rapid owner-occupancy growth, homebuyer activity has cooled, but the trend remains positive. Through the first half of the year, parcel records indicate that the city has enjoyed a net increase of slightly more than 200 owner-occupiers.

A note on data

The statistics in this article are derived from a custom dataset matching state transaction records with city parcel data. Due to delays in the reporting process, the 2023 statistics are preliminary, and the final totals will likely be slightly higher than at present. About 1.5% of transactions could not be matched and are not included in this analysis.