Where do housing unit need projections come from? If you answered, "from population projections," you are only partly correct. The little-understood and unappreciated "headship rate" translates population projections into units. Small changes in headship rates could result in significant changes in housing need projections. Housing element law, first enacted in 1969, mandates that local governments work to provide a safe and livable environment for Californians. That mandate evolved to include adequately planning to meet the existing and projected housing needs of all economic segments of a community. During the latest round of state-mandated housing element updates, many cities and counties have struggled to plan for the number of units required by the Regional Housing Needs Allocation (RHNA). The RHNA number is determined by the headship rate — the probability that individuals, families, or groups will rent or buy a housing unit. Headship rates are worth a closer look because they estimate how many households are formed from the population projections. The Southern California Association of Governments (SCAG) currently is focused on the headship rates and realizes that small changes in the rate will make large differences in future housing unit projections. The RHNA allocation is a two-part number: the existing housing need and future need based on population growth. The existing need figure is the number of units necessary to reduce overcrowding and provide an economic choice range that, coupled with a decent vacancy rate, enables an efficient and equitable housing market to operate. The growth-based housing need estimate is based on population projections (a combination of net natural increase and net migration) translated into numbers of households via the headship rates. Like most demographic methods, headship rates reflect how the population has been carved up into housing units in the past. The Department of Finance (DOF) and SCAG headship rates are essentially probabilities that a person of a specific age and race will be listed as the "Head of Household" ("Reference Person" in census jargon) or the person listed first on a lease, mortgage or property title. Chart No. 1 contains the Department of Finance's headship rates for San Bernardino County. It shows that about 22.5% of males age 18-24 are a household head. That means the remaining 77.5% of age 18-24 males are living with their parents, at school, in jail or with someone else who is listed as the "household reference person" in a census or survey. You can also see that male heads of household tend to outnumber female, and that the highest headship rates are for people older than 55. Headship rates also vary by race, income and geography, reflecting the various factors that influence people's desire and ability to form households. In San Bernardino County, the expected number of households for males age 35 to 44 differs by race, as shown in Chart No. 2. "Others" include Asians and Pacific Islanders and American Indians. Here you see the general pattern that more than half of all males in this age group are renting or owning an apartment or house, more likely renting than owning. These "starter households" are the most likely to be affected by a tight housing market because of cost and low housing supply. Most people older than 35 already have a home, and they may choose to remain in it pending a good opportunity to move. So, when they do not move, or move less frequently, the pipeline "jams up" and affects the starter households the most. We are already seeing in some areas more young adults continuing to live with their parents. As a result, young adults are less able to move to take entry-level jobs, or the jobs have to offer higher wages. You may infer that because non-Hispanic whites generally have higher incomes than the other groups, headship is a function of income. You would be right. That raises the question of other ways to estimate households that may better anticipate future household formations in an increasingly expensive housing market. Perhaps sex and race-based headship rates are not the best way to anticipate future household formation. When Census 2000 income data are fully released by September, this is one topic worth examining. Maybe in the next round of fair-share housing allocations, a different headship rate methodology would yield results more attuned to the interests of local businesses and residents. In our increasingly diverse population, one wonders if race-based headship rates still make sense. Perhaps the rates should be based on income and/or education. The existing rates reflect the housing choices of the 1980s and early 1990s, which is another reason to explore more accurate ways to create the rates.