Modeling System The regional forecasting model produces employment forecasts by industry that are, in turn, disaggregated into employment for individual counties. The county demographic component produces individual county level population forecasts directly. County level economic and demographic data serve as control totals for the sub-county model, Projective Optimization Land Use Information System (POLIS). Finally, the results from the POLIS system are used as controls for the Subarea Allocation Model (SAM). The potential for economic and demographic growth and its distribution are based on the availability of space (vacant land, under-utilized existing developed areas, and building sites that can be reused or redeveloped). Space must be adequate to allow for the forecast levels, whether it is the total region, county, zone, or census tract. Therefore, because of the importance of available space, land use data is continuously collected from all jurisdictions in the Bay Area through ABAG's Local Policy Survey. Data collected through the Local Policy Survey include the availability of vacant land, timing of future development, type of future development, density of development, transportation, land use policy constraints on development, and other land use related factors that could affect development. This data is used in subsequent parts of ABAG's modeling system and acts as a key determinant of the geographic disaggregation of the regional, county, and census tract forecasts. Regional / County Forecast (HENRY) Regional Economic Component Regional economic relationships drive the employment projections for the San Francisco Bay Area. The core element of the regional economic component is a regional input/output (I/O) table that provides a flexible framework for analysis of the area's economy. The input/output table represents the economy as the flow of goods and services between sectors. These sectoral flows show the inputs from each industry needed to produce a good or service for a single economic activity and the corresponding sales of goods and services to other industry sectors and to consumers. While the input/output relationships show how the region supplies each unit of goods and services, the level of demand determines the amount of economic activity and total employment. Demand in each year is driven by assumptions about consumer spending, government spending, capital investment, and exports. County Employment Disaggregation The disaggregation of employment and income for each of the nine counties in the Bay Area is calculated using a statistical model to estimate separate industry employment levels for each of the nine counties, consistent with the employment forecasts produced by the regional input-output model. Prior to the forecasting process, predictive equations are developed. These equations recognize that employment in a particular industry in an individual county is related to factors like employment in related industries, population growth, income growth, and competition from other areas. Employment data comes from the U.S. Census Bureau's County Business Patterns, and population and income data are derived from the California State Department of Finance. County Level Demographic Component The demographic component of the HENRY model uses the cohort-survival method to develop population forecasts. This method works by starting with a beginning population for each county, identified by age cohorts, and growing it over time using age-specific information about birth rates, birth timing, death rates, and net migration. Few alternatives to the cohort-survival method are available. Statistical trending is often used in conjunction with, but not as an alternative to, the cohort-survival method. The cohort-survival method's superior flexibility and theoretical foundation make it an unsurpassed choice in population forecasting. Since data is available for individual county populations, ABAG estimates county populations directly. This is in contrast to employment estimates where regional data is used to estimate total regional employment and then allocated to individual counties. Migration/ Commuting Component The interaction of the economic and demographic forecasts occurs through migration rates. The population forecast is used to generate a labor force forecast. A primary driver of in-migration occurs when a tight labor market causes people ("economic migrants") to relocate to obtain employment. Migration is also composed of (although to a lesser extent) social migration and retirement migration, which is dependent on employment, income, and the cost of living. The number of people migrating into each county is also subject to available housing. A related component of the population forecast is inter-regional commuting. People working in the Bay Area, but living outside the region, are motivated by factors similar to economic migrants. However, housing costs and opportunities cause them to make different choices. Zone Level Allocation (POLIS) Travel to work, the availability of and attractiveness of housing, and the current levels of nearby employment help to determine residential choice, i.e. where individuals choose to live. The locational patterns of industries are influenced by the accessibility to labor supply, the proximity to other similar industries (agglomeration economies), the availability of land for development, and housing opportunities for employees. The basis of POLIS is a random utility theory that describes the behavior of individuals (employees) selecting from a set of alternatives in order to maximize their utility (or benefit). The POLIS model is able to capture a measure of the factors that attract firms and individuals to particular parcels of land. The individual attractiveness variable incorporates factors like education, which is highly correlated with income, and housing prices, which make newer, more outlying areas more attractive. As a dynamic model, the values for attractiveness change over time. The constraints of the POLIS model are land supply, as described by the housing and development policies of the different cities, and the total employment and housing to be allocated to the areas within each county as derived from the HENRY Model. Census Tract Allocations (SAM) The final step in ABAG's modeling system is to allocate jobs and housing to approximately 1,400 census tracts in the region. This is accomplished through the Subarea Allocation Model, or SAM. Employment is distributed to census tracts according to a weighted combination of a tract's share of employment potential and its previous share of its zone's employment. Employment potential is indicated by the amount of land available for development as determined in POLIS. Similarly, population (measured by occupied housing units) is allocated according to the weighted census tract share of residential unit potentials and the base year number of units. Frequently Asked Questions about the Projections Process and the Modeling Is the modeling a top-down or bottom-up system? Neither. The model does not fit either of these categorizations, for the heart of the model is statistical data, from the national census to local sources. However, before the projections are released to the public, local planners have the opportunity to comment on the projections. Changes resulting from specific local information can result in overall changes. Are the subregional projections a floor or a ceiling to anticipated growth levels? Neither. The projections are estimates of an area's development potential based upon current zoning, general plans, and other local development policies, in conjunction with economic and demographic trends. Does the Projections forecast assume that each county can be totally built out? No. A variety of factors determine whether a vacant parcel is suitable for development. These factors include proximity to job sites and local land use designation. Do the regional housing need numbers differ from the Projections forecast numbers? Yes. The State Department of Housing and Community Development (HCD) sets a regional goal for housing production in ABAG's region and regions throughout the state. Since HCD's goal is to increase the amount of available housing, its number is higher than ABAG's forecast of expected housing levels under existing policies. Does HCD assign housing need numbers to individual cities and counties? No. The Regional Housing Needs Numbers are allocated by using a combination of information from the State of California Housing and the Community Development Department (HCD) and from ABAG's Projections forecast on primarily expected housing growth and expected employment growth. Does ABAG use the estimates provided by local planners regarding anticipated new housing units and/or jobs? Sometimes. ABAG forecasts growth based on macro-economic trends using data that aggregate many projects. Estimates by local officials may not be used by ABAG if they are counting a specific project, even a large one, as basis for data. However, whether or not ABAG accepts the data provided by local agencies, all comments by local agencies are circulated to the elected officials on ABAG's Executive Board as part of the process. Do the jobs that ABAG estimates equal the jobs reported by the State of California Employment Development Department (EDD)? No. ABAG generally projects more jobs for the Bay Area over a given period than EDD. ABAG and EDD use somewhat different data as inputs. Specifically, EDD uses information records provided by employers. ABAG uses employer records, as well as information on military personnel, self-employed persons, and contract laborers, all of which are not included in the EDD input data. Do ABAG's population estimates differ from the State of California Department of Finance (DOF)? Yes. DOF and ABAG use the same type of model -- the cohort survival model. However, ABAG assumes levels of migration to the Bay Area based upon the availability of land and job growth. ABAG's model consequently makes more modest assumptions about population growth than DOF. Does ABAG assume existing policies and laws, or does ABAG assume that local policies will change? ABAG assumes that local policies will be consistent with current political agendas in the community. However, ABAG assumes that policy makers will respond to problems as they arise and that specific policies will be modified to respond to new problems in the long term. PROJECTIONS 2002 ABAG HOME