The blog of the Urban Institute
November 14, 2019

Four Ways Technology Is Addressing the Housing Affordability Crisis

November 14, 2019

We’re in the midst of a housing affordability crisis that will worsen if increases in housing supply remain anemic while demand increases as expected.

Brian Brooks, chief legal officer at the cryptocurrency company Coinbase (and former Fannie Mae general counsel), outlined the housing sector’s basic supply and demand mismatch during his keynote speech at the Urban Institute’s Housing Finance Policy Center’s sixth annual housing symposium.

Brooks offered a hopeful vision of four ways technology companies are stepping up to help address this crisis: reducing the cost of supplying new housing, expanding access to capital, improving our risk-prediction methods, and easing the costs and hassles of loan administration.

The supply and demand mismatch

As Brooks laid out in his speech, the rate of home construction today is less than half of what it was before the 2008 crisis, with just seven homes being built for every 1,000 adults. At the same time, the household formation rate is twice what it was immediately following the 2008 crisis, with about 1.2 million new households forming over the next few years, compared with the 600,000 households formed between 2010 and 2015.

Millennials have also delayed homebuying for six to seven years compared with previous generations, even though they are entering their prime homebuying years.

Essentially, we are seeing half the historic rate of household construction and double the rate of household formation. That’s why even middle-income residents end up as renters in high-cost cities such as San Francisco, where the average home price is $1.3 million, or in Washington, DC, where the average price is $800,000.

Technology may provide solutions.

Reducing the cost of supplying new housing

Technology can make it cheaper to generate new units. Homebuilding has benefitted from 3D printing and modular housing construction by companies like Blokable, which builds the parts of a house in a factory and then ships the parts to the site for assembly.

This reduces the cost and time of construction by working within a controlled environment, allows for productivity regardless of the weather, and decreases the amount of physical labor used for a traditional, stick-built home. (Company names are examples of what the broader industry is doing.)

Katera is putting technology to work at all levels of design and construction, including ways that cut out the middle men in plumbing, electrical assembly, and other activities that add costs.

Reducing the cost of supplying new housing may come from unexpected places. Researchers at the Massachusetts Institute of Technology found that self-driving cars may reduce the amount of land used for highways, driveways, and garages, opening opportunities for more green space and home construction.

New living arrangements may also help reduce the cost and increase the supply of housing. Nesterly is basically a matchmaking service for older and younger people that helps them share costs and alleviate loneliness. Starcity and Common offer “coliving” where the companies supply and furnish the shared space in a home or building, providing utilities, cleaning, and social opportunities while renters lease individual rooms.

Expanding access to capital

Homebuyers need access to capital to access homeownership. Today’s tech companies are making this access easier and, in some cases, turning the mortgage model on its head.

Companies like EasyKnock and NerdWallet make shopping for the best mortgage easier by taking on and automating most of a borrower’s work, even filling out the application, and delivering competitive options that fit a borrower’s unique situation. Divvy is helping renters turn their rental payments into down payments toward homeownership.

Some are challenging the idea that credit is the only way to get into homeownership, with companies like Point using shared-equity financing to help borrowers buy a home. Instead of going to the bank for a loan, borrowers purchase a home with a financial technology partner who pays part of the down payment in exchange for an equity stake.

If the home is sold for a profit, the borrower shares the earnings with the partner, but the partner contribution isn’t subject to monthly debt repayment like a traditional mortgage.

Improving risk predictions methods

Our current risk prediction models are imperfect. As Brooks explains, “[Although] they’ve gotten better over time, those credit models only predict, at best, 60 percent of credit performance, and the rest is really hard to capture.”

This means we’re excluding many people, disproportionately those with moderate incomes and communities of color, because we can’t capture why they are creditworthy.

Artificial intelligence (AI) and blockchain technology can help us better predict risk by allowing us to create more accurate models and by delivering more data to plug into these models.

Using artificial intelligence, we can better know what is actually relevant to loan performance. We need to recognize AI’s potential risks and fair housing challenges, but we have the same dangers and challenges with the current manual process.

Current risk prediction is based on a handful of manually weighted factors, one of which is the borrower’s credit score. Today, credit scores are based on only a few categories of payments, such as the borrower’s history of paying mortgages, credit cards, and other consumer loans.

Most recurring payments—cell phone bills, Netflix subscriptions, rent payments, utility bills, and the like—are not captured in the traditional credit bureau system.

Blockchain entrepreneurs are now working to build networks of payment information that capture more categories of recurring payments, all of which may predict a borrower’s future ability and willingness to repay. Although some traditional players in the credit score ecosystem are doing this to some extent, such as FICO with their XD product, which includes telecommunications payments, the current system does not holistically incorporate these other payments.  

Reducing the cost of loan administration

Loan administration is costly. Originating a loan, payment collection, title searching, and default servicing can add 100–150 basis points to the cost of a loan.

Blockchain technology can reduce the costs of loan administration by automating more of the process. Provenance offers an electronic mortgage application process that creates a fully electronic mortgage note recorded on blockchain, without the need to register the note with an analog registry platform.

All payments and transfers are recorded “on chain.” All due dates and servicing information are in one place, the borrower’s bank account is connected to the mortgage so payments are never missed accidentally, and even default servicing happens without having to engage and pay a default servicer. All of these cost savings can be shared with the borrower.

In addition, an internal token process allocates loan origination opportunities to improve the match between individual borrowers and the lenders and loan products most suited to their particular credit profile.

In all the ways described here, tech innovation holds promise for making housing and homeownership accessible to more people.

Brooks warned, however, that technology can do little to address zoning laws, environmental and other regulations, and other compliance mandates, which currently add $80,000 to $100,000 to new home construction. He urged industry stakeholders to continue to work to align these costs with the need to address the continuing supply and demand mismatch.

Brian Brooks delivers a keynote address at the Urban Institute's sixth annual housing symposium: Reimagining Housing: Closing the Equity and Supply Gaps, in Washington, on Oct. 23, 2019. (Ting Shen for The Urban Institute).

 

SHARE THIS PAGE

As an organization, the Urban Institute does not take positions on issues. Experts are independent and empowered to share their evidence-based views and recommendations shaped by research.