How Smart Growth and higher density policies made San Jose housing less affordable. (part 2 of 3)

Contrary to urban planners' promises, growth boundaries and the push for greater densities ended up making local housing more expensive and lowered tax revenues. Randal O'Toole explains in the second of his three-part exclusive to Opportunity Now.

by Randal O'Toole

High housing prices were the unforeseen result of San Jose’s slow-growth plan. Any economist would predict that limiting the supply of land for development would increase housing prices, but urban planners insist to this day that Silicon Valley’s high housing prices are a function solely of demand and have nothing to do with the growth boundary.

A standard measure of housing affordability is the value-to-income ratio, that is, median home prices divided by median family incomes. Under standard mortgage rules that limit loans to 30 years and limit spending on housing to 30 percent of a homebuyer’s income, a family can easily afford a house that costs three times their income. Depending on the interest rate, it is more difficult to buy a house that costs four times the family’s income. A family cannot get a mortgage for a house that costs five times its income unless it pays for most of that house with a hefty down payment.

In 1969, the median price of homes in the San Jose urban area was 2.2 times median family incomes, which was actually more affordable than the national average of 2.4. The establishment of the urban-growth boundary drove prices to more than 4.0 times incomes in 1979 and more than 5.3 times incomes in 1989. By 2005, when the city was considering expanding the growth boundary to include Coyote Valley, they were nearly 7.6 times incomes. 

Although prices fell after 2006, they have since recovered and today are more than 7.6 times incomes.

Density advocates say that the solution is to build more high-density housing. This will increase the housing supply and make housing more affordable. The problem with that is that mid-rise (four- to six-story) and high-rise (seven stories or more) housing projects cost a lot more per square foot than single-family homes. Such projects require a lot more structural steel and concrete. They also require large common areas including hallways, recreation areas, and in some cases parking structures whose costs must be counted against the residential units. They also require elevators, whose cost per housing unit can be especially high in mid-rise developments.

In testimony before the San Francisco Bay Area Metropolitan Transportation Commission, Bay Area developer Nicholas Arenson reported that mid-rise housing cost three to four times as much to build per square foot as low-rise, while high-rise cost 5.5 to 7.5 times as much per square foot as single-family homes. Contrary to claims that there is a pent-up demand for such housing, he also noted that such high-density housing “sells at a discount” to single-family dwellings.

The Los Angelization of San Jose

In 2001, former San Jose Mayor Janet Gray Hayes, who was on the city council when it adopted the 1974 growth boundary, called proposals to expand the urban-growth boundary “the Los Angelization of San Jose.” The Sierra Club has called Los Angeles “the epitome of sprawl” and considers its growth pattern as something to be avoided.

In fact, close scrutiny reveals that Los Angeles is the epitome of smart growth, not of sprawl. The Los Angeles urban area, which includes Pasadena, Anaheim, and other parts of Orange County, is the densest urban area in the United States, with more than 7,200 people per square mile. The New York urban area, which includes northern New Jersey, most of Long Island, and parts of Westchester County, has less than 5,500 people per square mile.

Rather than preventing the Los Angelization of San Jose, the smart-growth plans for the region ensured it. In 1970, the San Jose urban area had less than 3,700 people per square mile while the San Francisco-Oakland urban area had under 4,400 people per square mile and the Los Angeles region, at 5,300 per square mile, was less dense than the New York urban area. Since then, urban-growth boundaries and similar policies have boosted the density of San Francisco-Oakland to more than 6,800 people per square mile and San Jose urban area to more than 6,300 people per square mile, making them the second- and third-densest urban areas in the United States.

There is no evidence that this density has made housing more affordable. Instead, affordability has declined as these and other regions became denser. On a nationwide basis, there is a strong inverse correlation between densities and price-to-income ratios: denser regions are less affordable. Among the nation’s sixty largest urban areas, the correlation coefficient between density and the price-to-income ratio is 0.82, where 1.0 is perfectly correlated and 0.0 is totally random.

In addition to the higher cost of building mid-rise and high-rise housing, urban-growth boundaries increase land costs. A 2017 study published by the National Bureau of Economic Research found that the average cost of an acre of land in the San Jose urban area was $2.3 million, compared with costs of around $300,000 per acre in fast-growing regions with no growth boundaries such as Atlanta, Dallas, and Houston.

The growth boundaries also allow cities to impose huge impact fees and an onerous approval process on home construction since they know developers can’t go elsewhere. The highest impact fees are in Fremont, which charges $75,000 per dwelling unit, but fees are also high in San Jose and other cities in the region. By comparison, the only fees in Houston are about $2,000 per unit to hook up to sewer and water systems. In Houston, it’s also possible to buy land, get all permits, build a home, and move in within a few months; in San Jose, it can take years.

A 2002 study quantified each of these costs by comparing home construction in San Jose vs. Dallas. A 2,400-square-foot lot in San Jose cost $200,000 more than a 7,000-square-foot lot in Dallas. Impact fees in San Jose at the time were $24,000 per home than in Dallas. Because it cost more to live in San Jose, labor costs were higher, adding another $143,000 to the cost of the house. Finally, the additional cost of getting a permit—and the risk that permit applications would be denied—added $90,000 to the cost in San Jose. All of this explains why the median home price in San Jose at that time was about $500,000 compared with under $120,000 in Dallas.

All of these factors have only gotten worse since 2002. The Census Bureau estimates the median home price in the San Jose urban area was $1.2 million in 2019 while in the Dallas-Ft. Worth urban area it was under $250,000. Zillow reports about the same median home values in 2020. To add insult to injury, San Jose now charges an “affordable housing fee” of $17 per square foot of new residential construction, thus making existing housing that much less affordable.

High housing costs have greatly limited the growth of Silicon Valley. Despite the fact that it is one of the wealthiest urban areas in history, the population of the region has grown by only about 25 percent since 1990. In the same time period, affordable regions such as Atlanta, Charlotte, Dallas-Ft. Worth, Houston, and Phoenix have grown by 80 to 140 percent or more.

Density and Taxes

The Santa Clara County tax assessor bases the value of residential properties on the value of the land plus the value of the improvements. Under proposition 13, which was passed in 1978, property values are reassessed whenever the property is sold. Between sales, assessed values can only increase at the rate of inflation but not more than 2 percent per year. Since urban-growth boundaries have led to prices increasing much faster than 2 percent per year, people who have recently purchased homes end up paying much higher taxes than people who have owned their homes for many years.

To see how density influences tax revenues, I compared the assessed values of single-family homes and condominiums of similar sizes that were most recently sold in the same years. In general, I found that, for properties sold in the same year, the assessed value of the improvements per square foot was about the same whether it was a single-family home or a multifamily condominium. However, the values of the land under the single-family homes were much greater than the land values for the condominiums. Thus, single-family homes bring in more revenue per dwelling unit than condominiums, but condominiums bring in more revenue per acre.

I also found that more than 200 multifamily housing projects, with more than 21,000 apartments or condominiums, have been built in Silicon Valley with low-income housing tax credits and other affordable housing subsidies since 1990. While I haven’t checked them all, the ones that I checked were almost completely exempted from property taxes as a part of their affordable housing subsidies. Many of the housing units in these developments are open to any family whose income is less than the median family income of the region. In 2019, the median family income in the San Jose urban area was more than $150,000, so this tax-exempt housing is not necessarily serving people with very low incomes.

Follow Opportunity Now on Twitter @svopportunity.
Image taken by Will Buckner.