Since the mid-1990s, most new market-rate housing in California has been of the single-family detached (SFD) variety. Under traditional assumptions, these units should be generating school-age children at a predictable rate — because they would be occupied by families with a predictable number of children. But when single-family homes represent the only housing being built, those assumptions might go out the window. All types of households might occupy the new houses. And that raises some obvious questions about student generation rates. Under such circumstances, do K-12 students really live in the new houses? And, conversely, are those students rarer in multi-family dwellings (MF)? A quick comparison we did between the Central Valley towns of Clovis and Fresno and the coastal cities of Ventura and Oxnard found that the traditional assumptions might be faulty. Yes, the new houses have kids. But far more children appear to live in apartment buildings and townhouses these days than in years past. A May 2001 facilities needs assessment by the Clovis Unified School District — a Central Valley district of about 150,000 people with a considerable amount of new housing — follows the traditional assumptions. The district calculated student generation rates of 0.7822 students per single-family home, and of 0.330 students per multi-family dwelling. These rates imply that there are nearly 2 1/2 times more K-12 students in single-family houses than in multi-family units. These rates are not unusual for California, but they are on the high side nationally. We explored the "kid rate" for new housing developments, and we compared current demographics with 20-year-old data to see if the number of students in different types of housing has changed. We also compared inland (Clovis and Fresno) to coastal (Ventura and Oxnard). This is not a rigorous analysis — more of a quick look — and we only looked at the percentage of households with children younger than 18. Our new housing areas are in Census 2000 blocks that were built between 1990 and 2000 and are nearly all single-family and owner-occupied. For the past-versus-present study, we selected Census 2000 tracts that were either at least 90% single-family or at least 60% percent multi-family. The most interesting thing we found is that the percentage of multi-family units with children appears to have jumped significantly. Both the inland and coastal multi-family kid rates increased — nearly doubling inland from 32% to 60%, and increasing from 40% to 53% in the coastal area. The MF population and persons-per-unit number increased accordingly. In the single-family neighborhoods, we found that children are far more likely to live in new houses than in pre-1980 structures. We found that about 60% of households in newly built areas have children younger than 18, whether those areas are coastal or inland. This figure is a bit below Clovis Unified's calculated rate. We also found that coastal households are slightly larger, but, given that this is not a scientific sample, we cannot say for sure. In the established inland neighborhoods, the single-family-dwelling kid rate was 34% in both 1980 and 2000. In the coastal area, the SFD kid rate was lower yet (about 22%) and also unchanged from 20 years earlier. Ours is a quick look at only a few census tracts. But we may be seeing a pattern in which young households first move into multi-family housing (little of which is getting built these days) and remain there longer with their school age children before moving into newly built houses. Meanwhile, older single-family houses remain occupied predominantly by households with no children. This pattern appears to be more pronounced inland. Christopher Williamson is senior research associate at Solimar Research Group. Mivelia Andika is an intern at Solimar.