Welcome to the Department of Housing and Community Development

Right Column

Chapter 2: California Housing Production Needs, 1997-2020

How many people will call California home in 2010, and how many housing units will they need? This chapter draws on recently updated population projections from the California Department of Finance to develop 2010 and 2020 household and tenure projections by county, metropolitan area, and region.

California Population And Household Projections

After a slow start in the early 1990s—the result of the worst economic conditions since the Great Depression—population growth in California has again picked up. Between January 1995 and January 1999, California's population increased by 1.8 million. If present trends continue, California's population will likely reach 35 million by 2000, 40 million by 2010, and 45.5 million by 2020 (California Department of Finance, 1998; see Exhibit 1). On a yearly basis, California's population is expected to grow at a rate of 1.6 percent per year between 1997 and 2010, and 1.3 percent per year between 2010 and 2020.

Age Group Trends

Most of California's projected population growth will occur through new births. Net migration, which accounted for more than half of the State's population growth during the 1980s, is expected to account for a significantly smaller share of 1997-2020 statewide population growth (California Department of Finance, 1998).1 All but 5 percent of California's projected population growth will occur in metropolitan areas.

The age groups projected to grow the most will be those under 18 and those over 65 (see Exhibit 2):

  • 4.6 Million More Children: Between 1990 and 2020, the number of California children (those 18 and under) is projected to increase by 4.6 million. Children will account for almost 30 percent of the State's population growth between 1990 and 2020, and comprise 27.4 percent of the state's population by 2020.
  • 3.2 Million More Seniors: At the other end of the age spectrum, California's senior population (those 65 and older) is projected to increase by 3.2 million persons between 1990 and 2020. Seniors will account for 21 percent of the State's 1990-2020 population growth, and 14 percent of its 2020 population.
  • 55- to 64-Year-Olds: The number of Californians 55- to 64-years-old is projected to increase by 3.1 million between 1990 and 2020. As of 2020, 55-to-64 year olds will account for 12 percent of California's population.

While the number of Californians 25- to 34-years-old will increase in absolute terms (by nearly three-quarters of a million people between 1990 and 2020), their overall share of the State's population is projected to decline from 19 percent in 1990 to just over 14 percent in 2020. The number of 35 to 45-year olds will also grow in absolute terms but decline in share.

Race and Ethnicity Trends

Population growth increments and rates will also vary widely by race and ethnicity (see Exhibit 3):

  • Hispanics—both immigrants and native-born—will, it is projected, account for 64.5 percent of the state's population growth between 1990 and 2020. Just under 40 percent of Californians in the year 2020 will be Hispanic, up from 26 percent in 1990.
  • Asians and Native Americans will, it is projected, account for the next largest share—24.6 percent—of California's population growth between 1990 and 2020. Asians and Native Americans will comprise just under 15 percent of California's population in 2020, up from 9.8 percent in 1990.
  • Whites, it is projected, will account for 6.4 percent of California's 1990-2020 population growth. Statewide, the number of white Californians will increase by approximately one million persons between 1990 and 2020, even as their share of the State's population declines from 57.2 percent to 39.9 percent.
  • African-American Californians will account for 4.5 percent of the state's projected 1990-2020 population growth. California's Black population will likely increase by 700,000 persons between 1990 and 2020, even as their share of the state's population declines from 7 percent to 6.2 percent.

Household Projections and Housing Production

These numbers and percentages are projections, not facts, and projections are frequently updated. The 1998 Department of Finance (DOF) population projections, for example, are significantly lower than those issued in 1993. Those projections, as well as history, suggested that California's population would grow at an average annual rate of 1.7 percent between 1990 and 2020. The 1998 projections peg California's projected average annual growth rate at 1.4 percent.

Percentages aside, California will grow a lot during the next 20 years. In terms of households, the basic building block of housing demand, the state will likely add just over 3 million households between 1997 and 2010, and just over five million households by 2020.

These population increases will put significant pressure on California's urban housing markets. To meet this level of demand, California homebuilders would have to construct an average of 220,000 additional housing units every year for the next 23 years. Producing this amount of new housing on a sustained basis will be a stretch. Since 1987, annual non-manufactured housing production in California (as measured by the number of permits) has averaged only 141,000 units per year.2, 3 Annual production since 1990 has been much lower, averaging just over 100,000 units per year. Since 1980, statewide housing permit levels have exceeded 220,000 per year only four times—in 1985,1986, 1987, and 1988 (see Exhibit 4).

Even if such production levels are possible, it is unlikely they could be sustained. Housing production is cyclical. It declines when the economy contracts or interest rates increase, and expands when interest rates fall or the economy booms. Rarely do the boom years fully compensate for the lean years. Particularly in coastal markets, California homebuilders are forever playing catch-up with demand. If current tenure preferences hold, two-thirds of California's new households will want to be homeowners.4We say want, because if current housing supply and price trends continue, fewer than half of these new households will be able to afford home ownership. Homeowners currently make up 57 percent of California households. (see California Housing Markets 1990 û 1997, Statewide Housing Plan Update Phase II, 1999.) Californians' preference for homeownership and single-family housing is hardly new. Since 1987, single-family homes have accounted for more than 80 percent of annual statewide new home production.

Regional Population Projections

As always, population growth will vary from one part of California to another (see Exhibit 5). Comparing different urban regions:

  • Greater Los Angeles: The six-county Greater Los Angeles Metropolitan Region (including Los Angeles, Orange, Riverside, San Bernardino, Ventura, and, Imperial counties) will, it is projected, account for almost half of California's 1997-2020 population growth. Altogether, the Greater Los Angeles region is projected to grow from 16.1 million persons in 1997, to 19.2 million persons in 2010, to 21.8 million persons in 2020—a 23-year increase of more than five-and-a-half million!

    Among individual counties, Los Angles County alone is projected to add two million more residents by 2020. Further to the east, Riverside and San Bernardino counties will likely add 570,000 new residents between 1997 and 2010, and 1.1 million new residents by 2020. Orange County, which by some measures is already nearly built-out, is projected to add 458,000 new residents by 2010 and 726,000 by 2020. Projected growth increments in Ventura and Imperial counties will be considerably smaller.
  • San Francisco Bay Area: California's second-largest metropolitan region, the nine-county San Francisco Bay Area, will, it is projected, add 1.1 million residents between 1997 and 2010, and another 600,000 by 2020. Home to 20.4 percent of the state's population in 1997, the San Francisco Bay Area will account for a disproportionately-low 13.3 percent of California's future population growth.

    More than half of the Bay Area's population growth will occur in just two counties: Santa Clara and Alameda. Santa Clara County's population is projected to grow by 525,000 persons between 1997 and 2020, while Alameda County's population will likely grow by 395,000. Elsewhere in the Bay Area, projected 1997-2020 population growth will exceed 200,000 persons in Contra Costa County, and 140,000 in each of San Mateo, Solano, and Sonoma counties. Even "built-out" San Francisco is projected to grow by 5,000 residents between now and 2010. Napa and Marin counties are expected to grow by 37,000 and 25,000 persons, respectively.
  • The San Joaquin Valley: The eight-county San Joaquin Valley Metropolitan Region (consisting of Fresno, Kern, Kings, Madera, Merced, San Joaquin, Stanislaus, and Tulare counties) was home to just under ten percent of California's population in 1997. By 2020, that share is projected to have increased to 11.2 percent. According to the California Department of Finance, the San Joaquin Valley Metropolitan Region will likely grow by just over one million residents between 1997 and 2010, and by 900,000 between 2010 and 2020.
    Within the San Joaquin Valley region, population growth will be fairly evenly distributed. Kern and San Joaquin counties are projected to add the most new population; and Madera, Merced, and Kings counties, the least. Projected population growth rates in the San Joaquin Valley will vary from a low of 43 percent in Fresno County to a high of 98 percent in Madera County.
  • San Diego County, California's fourth largest metropolitan region, will, it is projected, grow by 678,000 residents between 1997 and 2010, and by 476,000 between 2010 and 2020. With 8.4 percent of the state's population in 1997, San Diego County will account for 9.2 percent of projected state population growth between 1997 and 2020.
  • The Sacramento Metropolitan Region—consisting of El Dorado, Placer, Sacramento, Sutter, Yuba, and Yolo counties—is currently home to 1.8 million people, or 5.6 percent of the state's population. DOF projections suggest that this region will add another 544,000 residents by 2010, and another 923,000 residents by 2020.

    Among individual counties, projected 1997-2020 population growth increments will vary from a high of 504,000 in Sacramento County to a low of only 23,000 in Yuba County. Population growth in foothill counties of Placer and El Dorado is projected to exceed 175,000 during the 1997-2010 period, and 275,000 during the 1997-2020 period.
  • The Central Coast Metropolitan Region is less an identifiable urban region than a combination of several small and mid-sized cities strung along Highways 1 and 101. From north to south, it includes Santa Cruz, San Benito, Monterey, San Luis Obispo, and Santa Barbara counties. Its northern portion, including Santa Cruz and San Benito counties, borders on and is sometimes considered to be part of the San Francisco Bay Area. Its southern county, Santa Barbara, is sometimes considered part of Southern California.

    In 1997, this amalgam of coastal counties was home to 1.3 million residents, or four percent of the state's population. By 2020, its population will be just under two million. Population growth will be evenly distributed throughout the region, with Monterey County likely to add the most additional population (+197,000), and San Benito likely to add the least (+36,000). Altogether, the Central Coast Metropolitan Region will account for 5.3 percent of the state's 1997-2020 population growth.
  • The Northern California Non-metropolitan Region consists of five mostly rural counties (Butte, Colusa, Glenn, Shasta, and Tehama), each having one or two small cities. The population of this region is expected to grow from 460,000 in 1997, to 612,000 in 2010, to 722,000 in 2020. Currently home to 1.4 percent of California's population, the Northern California Metropolitan Region will account for 2.1 percent of the state's 1997-2020 projected population growth. Nearly three-quarters of this region's projected population growth will occur in just two counties: Butte and Shasta. The populations of California's 21 non-metropolitan counties will, DOF projects, grow from 1.1 million in 1997, to 1.7 million in 2020.5 The non-metropolitan counties projected to grow the most during this period are Imperial (+156,026), Kings (+68,864), Nevada (+48,037), Lake (+38,024), San Benito (+36,125), Mendocino (+32,848), Tuolumne (+25,199), and Calaveras (+24,794).

From Population To Household Projections

Headship Rate Trends

Population projections can be converted into household projections—the building block of housing demand—by applying what are known as headship rates. Headship rates measure the statistical tendency of a given population to form households. Headship rates are calculated as the ratio of the number of household heads of a given age, gender, and race (as identified on census forms) to the population having the same characteristics. The higher the headship rate, the more new households are formed from a given population. A headship rate of .75 means that a population of 1000 persons will form 750 households. With a headship rate of .5, that same population of 1000 will form only 500 households.

Headship rates vary by age—they tend to rise as people age; by gender—men have historically had higher headship rates than women; and by race and ethnicityùwhites generally have higher headship rates than non-white Hispanics and Asians of similar age (see Exhibit 6).

Headship rates change over time, but not always in predictable directions. Among male Californians, headship rates declined consistently during the 1980s, albeit more for younger men than for older ones, and more for Hispanic and Asian men than for whites and blacks.

The story is different for women. For women over 25, headship rates increased consistently between 1980 and 1990, regardless of race or ethnicity, with the largest increases occurring among women aged 35 to 54. Headship rates among white women aged 18-24 increased between 1980 and 1990, but fell for Black, Hispanic, and Asian women of comparable age. Headship rates among black women over 65 rose during the 1980s, but declined among Hispanic and Asian women of comparable age. Headship rates also vary by location. They tend to be higher in metropolitan areas and lower in rural ones.

To accommodate these possible variations, age-, gender-, race-, and county-specific headship rates were estimated from the 1990 census (See Appendix A). Next, we multiplied DOF's most recent population projections (see Exhibit 5), organized by county into gender, race, and age cohorts, by their corresponding headship rates, to yield age-, gender-, race-, and county-specific household projections. The resulting projections are summarized by county in Exhibit 7 and Maps 1, 2, 3 and 4, and by metropolitan region in Exhibits 8 and 9.6

Household Growth Patterns

The pattern of household growth in California will more or less follow the pattern of population growth. The Greater Los Angeles Metropolitan Region (including Imperial County) will gain 1.4 million new households by 2010, and 977,000 additional households between 2010 and 2020. This region by itself is projected to account for nearly half of the State's household growth through 2020. By 2020, if these projections hold true, the six-county Los Angeles Metropolitan Region will be home to more than 7.5 million households.

Because its population is older and wealthier than that of the rest of California, the San Francisco Bay Area will add proportionately fewer new households per increment of population growth than the rest of the state. Altogether, the nine-county Bay Area will, we project, add 460,000 additional households between 1997 and 2010, and 248,000 additional households between 2010 and 2020. Bay Area household growth will account for just under 14 percent of the state's projected household growth between 1997 and 2020. By 2020, the Bay Area will, it is projected, be home to just over three million households.

The largest percentage household growth between 1997 and 2010, and then again between 2010 and 2020, will occur in the eight-county San Joaquin Valley Metropolitan Region. This region, which extends from San Joaquin County in the north to Kern and Kings counties in the south, will, we project, add 420,000 additional households between 1997 and 2010, and 320,000 more households between 2010 and 2020. By 2020 the number of households in the San Joaquin Metropolitan Region will, it is projected, have increased more than 70 percent over 1997 levels.

The number of households in San Diego County are projected to grow by 265,000 between 1997 and 2010, and by 165,000 between 2010 and 2020. These are increases of 28 percent and 46 percent, respectively. By 2020, we project that San Diego County will be home to nearly 1.4 million households.

The number of households in the six-county Sacramento Metropolitan Region are projected to grow from 668,000 in 1997, to 890,000 in 2010, to 1,042,000 in 2020. These are increases of 33 percent and 17.3 percent, respectively. Household growth in the Sacramento Metropolitan Region between 1997 and 2020 will likely account for 7.3 percent of statewide household growth.

The number of households in the state's two other metropolitan regions—the five-county Central Coast Metropolitan Region and the five-county Northern California Metropolitan region—will, it is projected, increase by 67,000 and 115,000, respectively, between 1997 and 2020. Although not as large in absolute terms as in other parts of California, these increases amount to percentage gains in excess of 58 percent and 63 percent, respectively.

California's 21 non-metropolitan counties (including those in metropolitan regions), will, we project, gain 136,000 additional households by 2010 and 91,000 more households by 2020. These are increases of 35.1 percent and 17.3 percent, respectively, and will account for just over four percent of the state's total projected household growth.

Because population and household growth will both occur in rough proportion to their current numbers, the distribution of population and households within California is not expected to shift all that much. Assuming current trends continue, by 2020 the Greater Los Angeles Region's share of the state's population is expected to remain at about 47 percent. The San Diego, Sacramento, Central Coast, and Northern Central Valley regions will also retain their current household shares well into the next century. By virtue of its slower population growth, the Bay Area's share of California households will decline from 22 percent in 1997 to 19 percent in 2020. The share of California households living in the San Joaquin Valley Metropolitan region will rise from 9 percent in 1997 to 11 percent in 2020.

Small in proportion but large in absolute number, there will also be some significant population and household shifts within metropolitan regions. In the Los Angles County Metropolitan Region, for example, the share of households living in Los Angeles County will likely decline (from 59 percent in 1997 to 54 percent in 2020), while the shares living in San Bernardino and Riverside counties will likely grow (from 19 percent to 24 percent).

In the Bay Area, San Francisco's share of regional households will likely decline from 13 percent in 1997 to 11 percent in 2020, while Santa Clara's share will increase from 23 to 24 percent. Fresno County's share of San Joaquin Valley households will, we project, decline slightly (from 23 percent in 1997, to 21 percent in 2020). Sacramento County's share of its region's households will also decline slightly, as growth favors the more eastern counties of Placer and El Dorado. Among Central Coast counties, Santa Barbara's share of regional households will decline from 31 percent to 28 percent, as household growth favors other counties, particularly San Benito.

As large as these intra-regional changes are, they should not obscure the more important point that every metropolitan region and urban county in California will experience significant population and household growth during the next quarter-century. The number of households—and thus the demand for housing—will grow at a faster rate than the population.

Alternative Household Projections

Implicit in all of these projections is the assumption that 1990 headship rates (as estimated using 1990 Census) will remain unchanged through 2010 and 2020. This is probably not a realistic assumption. Headship rates can and do change, often in unpredictable ways. Forecasting with alternative headship rates is thus a tricky business, but also a necessary one.

To test the sensitivity of the prior household projections to possible changes in headship rates, the following additional assumptions regarding future headship rate trends were made:

  • For the purposes of generating short-term household projections (2000 and 2003), 1990 headship rates were used, disaggregated by age, race, and gender. This method assumes that short-term household formation rates will not change from 1990 Census levels.
  • For the purposes of generating longer-term household projections (2010 and 2020), a trend-line based on 1980 and 1990 headship rates (as always, disaggregated by age, race, and gender) was developed. For some groups, headship rates during this period trended upward; for others, downward. Across all groups, headship rates declined slightly.

Once identified, the various headship rate trend lines were extended to 2010 and 2020, and used to estimate future headship rates (included as Appendix C). The resulting headship rates were then separately applied to DOF's 2010 and 2020 population projections. This method assumes that changes in headship and household formation rates observed during the 1980-90 period represent fundamental and embedded demographic and social trends, and that such trends will continue well into the future, albeit at a somewhat moderated pace.

These revised headship rates were applied to DOF's 2010 and 2020 population projections, shown in Table 5. We compared the resulting household projections (included as Appendix D) with our original household projections as presented in Exhibit 7. These comparisons are presented in Exhibit 10.

The principal effect of applying alternative headship rates is to reduce the number of projected 2010 and 2020 households by between three and nine percent, depending on the county. Statewide, the effect of using alternative headship rates is to reduce total projected 1997-2010 household growth from 3.1 million to 2.5 million. For the 1997-2020 period, total projected household growth would fall from 5.1 million to 4.4 million. Instead of needing 220,000 new housing units every year, the state would need 190,000. This is a significant reduction to be sure, but one which would still require a substantial increase in housing production activity over current levels.

Additional Caveats

Several additional caveats also merit mention. Household growth is typically the largest and most important source of the demand for housing, but it is not the only source. Other sources of housing demand include: (i) the replacement of demolished or unserviceable units; and, (ii) the addition (or subtraction) of units required to normalize local vacancy rates. When local vacancy rates are below normal levels, typically regarded as being in the range of three to five percent, owners and builders are able to make extraordinary profits. Conversely, when vacancy rates rise above normal levels, new construction typically becomes unprofitable.

More important than vacancy rates and replacement demand, especially in the short-term, are mortgage interest rates and lender qualifying ratios. When mortgage rates rise, fewer households qualify for homeownership, causing a decline in demand. The converse occurs when mortgage rates fall—housing demand rises.

Qualifying ratios also matter. Qualifying ratios are an underwriting measure used by mortgage lenders. They are calculated as the ratio of annual Principal-Interest-Taxes-Insurance (PITI) payments to gross or after-tax income. The higher the qualifying ratio, the larger the mortgage loan a household of a given income can afford. The lower the qualifying ratio, the smaller the mortgage. For conforming loans, (those which may be sold in the secondary market) qualifying ratios are set by the Federal National Mortgage Association (FNMA, or Fannie Mae) and/or the Federal Home Loan Mortgage Corporation (FHLMC, or Freddie Mac). Homeownership affordability rises with increasing qualifying ratios, and declines with decreasing qualifying ratios.

Another caveat concerns the usefulness of counties as the basis for reporting demand projections. Counties are administrative units, not housing markets. Sometimes county and housing market boundaries match each other fairly well, as in the cases of San Diego and Fresno counties. More frequently, housing market boundaries overlap county boundaries, leading to inter-county "spillovers" of housing demand. Projected housing demand which can not be accommodated in Alameda or Contra Costa counties for example, will spillover into San Joaquin and Stanislaus counties, even though the former and latter counties are in different metropolitan regions. In much the same way, unaccommodated housing demand will spillover from Orange County (and even San Diego County), into Riverside and San Bernardino counties; and from Los Angeles County, into Ventura, Riverside, and San Bernardino counties.

Like water in a channel, spillover housing demand seeks the path of least resistance, and the lowest housing prices. Recent job growth in Santa Clara County, for example, has led to an increase in housing construction in faraway-but easy-to-develop Tracy, but had little effect on production levels in nearby, but-harder-to develop, Santa Cruz County. Thus, while counties are a useful unit of analysis for summarizing information on housing demand and supply, they are less useful for understanding the subtleties and dynamics of the housing marketplace.

From Household To Tenure Projections

Like headship rates, long-term homeownership rates are slow to change. If one assumes that future homeownership rates—disaggregated by age and ethnicity—will mirror current rates, then it becomes a straightforward operation to convert future household projections (see Exhibit 7) into future homeownership and renter projections (Exhibit 11).

The demographic characteristics of California households (e.g., the aging of the population) will also constitute a strong bias toward homeownership. Statewide, the increase in the number of homeownership-oriented households will be double that of the increase in the number of renter-oriented households. All else being equal, and solely on the basis of anticipated demographic change, California's homeownership rate should rise from its current level of 57.1 percent to 59.1 percent by 2010. Statewide, the demand for owner-occupied units in 2010 will be just over 8.3 million, while the demand for rental units should exceed 5.8 million.1

Of course, all else is rarely equal, and California's high housing prices have long served to depress actual homeownership rates below expected levels, especially in coastal markets. If future housing production levels continue to lag demand, instead of rising, California's homeownership rates will fall.

Homeownership preferences will play out differently in different metropolitan regions. In the Greater Los Angeles Metropolitan Region, for example, the incremental (demographic) demand for homeownership2 will be strongest in Los Angeles and Orange counties, and weakest in Imperial County. Region-wide, and based on projected demographic characteristics alone, the incremental demand for homeownership will exceed the incremental demand for rental units by a ratio of 2.0 to 1. By 2010, if the appropriate supply of housing were made available, the homeownership rate in Riverside and Ventura counties could well exceed 66.4 percent, while the homeownership rate in Los Angeles County could top 50 percent. Region-wide, the demand for ownership housing in 2010 will exceed 3.8 million. The demand for rental housing in 2010 will approach 2.8 million units.

Among Bay Area counties, and based solely on projected demographic characteristics, the incremental demand for homeownership will be strongest in San Francisco, San Mateo, and Santa Clara counties. Region-wide, the incremental demand for homeownership units could exceed that of rental units by a 2.8-to-1 ratio. By 2010, given adequate production, the homeownership rate in Contra Costa County could top 70 percent. Even in San Francisco, where only one in three households is currently a homeowner, the incremental demand for homeownership will likely exceed the incremental demand for rental units. Region-wide, the demand for ownership housing in 2010 will exceed 1.7 million.

Based on projected demographic characteristics, the incremental demand for homeownership in the San Joaquin Valley Metropolitan Region will exceed the incremental demand for renting by just over fifty percent. Comparing individual counties, the incremental demand for homeownership will be strongest in Madera and Stanislaus counties, and weakest in Kings and Merced counties. Overall homeownership rates will rise only marginally by 2010, and range from a low of 52.2 percent in Kings County to a high of 64.3 percent in Madera County. Region-wide, the demand for ownership units in 2010 will exceed 844,000, while the demand for rental units will approach 590,000.

In San Diego County, the incremental demand for rental units will be just slightly less than the incremental demand for homeownership, based on projected demographic characteristics. Region-wide, the demand for ownership units in 2010 will exceed 660,000, while the demand for rental units will be just over 550,000.

Ownership units currently outnumber rental units in the Sacramento Metropolitan region by a ratio of nearly 2 to 1—a disparity which will likely continue through 2010. Between 1997 and 2010, the incremental demand for owner-occupied units will exceed 145,000 while the incremental demand for rental units will be just under 75,000. Comparing counties, the incremental demand for homeownership (based on projected demographic characteristics) will be strongest among the foothill and Sierra Nevada counties of Placer and El Dorado. Sacramento County by itself will account for more than half of the incremental demand for both ownership and rental units.

Among Central Coast counties, the 1997-2010 incremental demand for owner-occupied homes will exceed the incremental demand for rental units by 65 percent. Comparing counties, the incremental demand for homeownership over renting will be strongest in Santa Cruz and Santa Barbara counties, and weakest in Monterey County. Region wide, the demand for ownership units in 2010 will exceed 340,000, while the demand for rental units will be just over 243,000.

In the five-county Northern California Metropolitan Region, the 1997-2010 incremental demand for owner-occupied housing will be just over 45,000 units, while the incremental demand for rental units will reach nearly 23,000. The demand for homeownership will likely be strongest in Butte County, while the demand for rental units will be strongest in Shasta County.

As with any set of forecasts, care must be taken to treat these tenure projections with the appropriate caveats. They assume that current relationships between household demographic characteristics—especially age and race—and tenure preferences will remain constant through 2010, and that they will not be affected by short-term housing market or economic conditions. Should those conditions change on a permanent basis, then the various tenure projections would also be expected to change. Other changes that would likely impact the demand for homeownership would include: (i) changes in the availability of construction or mortgage credit and investment capital; (ii) significant shifts in U.S. housing, mortgage, or tax policy; (iii) changes in underlying housing preferences; (iv) the adoption of local land use policies that substantially limit housing production; and, (v) changes in the income distribution of the population.

In all the discussion over homeownership, it's easy to overlook the fact that California will also need many hundreds of thousands of additional rental units. Statewide, we project that the number of renter-oriented households will increase by more than one million between 1997 and 2010 (see Figure 11). To meet this level of demand, California builders would need to produce an average of 75,000 new additional rental units every year between now and 2010. By way of comparison, between 1990 and 1997, California builders produced an average of only 27,000 new multi-family units each year.3

Among individual counties, the largest increments of new rental construction between 1997 and 2010 will be needed in Los Angeles (+216,000), San Diego (+125,000), Riverside (+76,000), San Bernardino (+73,000), Orange (+57,000), Sacramento (+40,000), and Santa Clara County (+36,000).

Finally, we note the substitutability of for-sale and rental housing production. The greater the production of for-sale housing (relative to demand), the fewer the number of households displaced downward into the rental market. As recent trends in the Santa Clara and San Francisco rental markets so vividly indicate, shortfalls in the production of either for-sale or rental units quickly translate into rapidly escalating rents.

Chapter Summary

  • State's Population to Grow: Barring some unforeseen natural or economic calamity, California's population will likely grow from 34 million people in 1999 to just under 40 million in 2010, to 45.4 million in 2020. The number of households will grow at a faster rate than the population, reaching 14.2 million in 2010 and 16.3 million in 2020.
  • High Growth in Cities, Southern California: Following past trends, almost all of the State's population and household growth will occur in metropolitan areas; and more than half of it will occur in Southern California. Those counties that now have the lion's share of population—Los Angeles, Orange, San Diego, San Bernardino, Santa Clara, and Alameda—will also experience the lion's share of population growth. Thirteen of California's 58 counties are projected to add 100,000 or more new households between 1997 and 2020. Los Angeles County alone is projected to add more than two million people and one million households between now and 2020.
  • Varying Growth Rates: The Greater Los Angeles Metropolitan Region and the San Diego Region are each projected to add population and households at the same rate as the state. The San Francisco Bay Area will likely grow at a slower rate, while the San Joaquin Valley and Sacramento Metropolitan Regions will grow at a faster rate.
  • Some Rural Growth: Although not the focal points of California's growth, some rural counties (e.g., Calaveras, Lake, and Nevada) will see their populations increase by 50 percent or more between now and 2020.
  • More Homeownership: Most new California households will be strongly inclined toward homeownership. Statewide—and only if sufficient homes and apartments can be built—the demand for owner-occupied housing in 2010 among current and new households will be just over 8.3 million units, while the demand for rental housing should exceed 5.7 million units.
  • 220,000 Units a Year: To meet demand, California homebuilders would have to construct an average of 220,000 additional housing units each and every year through 2020. Since 1987, new single- and multi-family home production has averaged just 141,000 units per year.

    These projections assume constant (1990) household formation rates. A decline in household formation rates (consistent with trends observed during the 1980s) would reduce projected 1997-2020 household growth from 5.1 million to 4.4 million. Instead of needing 220,000 new housing units every year, the state would need 190,000.

References

State of California, Department of Finance. 1998. County Population Projections with Race/Ethnic Detail. Sacramento. December.

State of California, Department of Finance. 1998. Historical County Population Estimates and Components of Change, July 1, 1970-90. Sacramento. December.

State of California, Department of Housing and Community Development. 1999. Statewide Housing Plan Update, Phase II. Sacramento.

U.S. Bureau of the Census. 1990. 1990 Census of Population and Housing. Washington, D.C.: Department of Commerce.

U.S. Bureau of the Census. 1980. 1980 Census of Population and Housing. Washington, D.C.: Department of Commerce.

Endnotes

  • Migration proportions were developed for the two decades between 1970 and 1990 by a survived-population method. The 1970 population was aged forward in time to 1980 by adding recorded births to form new cohorts and subtracting deaths from existing cohorts. The survived population was compared to the 1980 population and differences were assumed to be migration. The ten-year migration was annualized and divided by the total to derive a proportion. The same process was used for the period 1980 to 1990. The migration proportions for the two decades were then averaged and smoothed using a five-year moving average.
  • In enacting the Tax Reform Act of 1986, Congress eliminated almost all of the tax shelter benefits associated with investing or owning multi-family housing. This change in federal tax laws caused multi-family development activity to collapse. It also complicates direct comparisons of pre- and post-1987 multi-family housing production levels.
  • This estimate excludes new manufactured housing units put in place. Manufactured housing accounted for approximately 3.7% of new California homes between 1990 and 1996.
  • This estimate is based on a demographic determination of projected homeownership rates; specifically, on the observation that homeownership rates rise with age. Thus, as California's population ages, all else being equal, homeownership rates should also rise. Homeownership does not require ownership of a detached-single-family home. As of 1990, 14 percent of California homeowners lived in attached structures. Thus, accommodating an increased demand for homeownership will likely require an increase in the production of both single- and multi-family housing units.
  • These estimates include the populations of Imperial, Kings, and San Benito countiesùthree non-metropolitan counties which partially fall within metropolitan areas.
  • To account for systematic deviations between DOF population projections and census counts, we developed and applied a household projection correction factor. This factor was the ratio of projected 1997 households (using DOF 1997 population projections, series E-5) and 1990 census headship rates) to DOF's 1997 household projections.
  • Our tenure projections extend only to 2010. Because we were reluctant to extend 1990 tenure rates beyond 20 years, we did not develop comparable projections for 2020.
  • The incremental demand for homeownership measures the demand for homeownership among newly formed households.
  • Not all rental units are multi-family units, and not all multi-family units are rentals. Nonetheless, for historical purposes, the comparison between apartment construction levels and rental demand is a reasonable one.

Copyright ©2000. All rights reserved.