|
Draft - Demographic and Economic Forecasting Model - DEFM Forecast 1993 to 2015, Vol. 1... - July 1993
Click HERE for graphic. Board of Directors SAN DIEGO ASSOCIATION OF GOVERNMENTS The San Diego Association of Governments (SANDAG) is a public agency formed voluntarily by local governments to assure overall areawide planning and coordination for the San Diego region. Voting members include the incorporated Cities of Carlsbad, Chula Vista, Coronado, Del Mar, El Cajon,Encinitas, Escondido, Imperial Beach, La Mesa, Lemon Grove, National City, Oceanside, Poway, San Diego, San Marcos, Santee, Solana Beach, Vista, and the County of San Diego. Advisory and Liaison members include CALTRANS, U.S. Department of Defense, San Diego Unified Port District, and Tijuana/Baja California. CHAIRWOMAN: Hon. Gloria McClellan VICE CHAIRMAN: Hon. Mike Bixler SECRETARY-EXECUTIVE DIRECTOR: Kenneth E. Sulzer CITY OF CARLSBAD Hon. Bud Lewis, Mayor (A) Hon. Ann Kulchin, Councilmember (A) Hon. Julianne Nygaard, Councilmember CITY OF LA MESA Hon. Art Madrid, Mayor (A) Hon. Barry Jantz, (A) Hon. Jay La suer, Councilmember CITY OF CHULA VISTA Hon. Leonard Moore, Councilmember (A) Hon. Tim Nader, Mayor CITY OF CORONADO Hon. Mary Herron, Mayor (A) Hon. Thomas Smisek, Councilmember CITY OF DEL MAR Hon. Elliot Parks, Mayor (A) Hon. Henry Abarbanel, Councilmember (A) Hon. Ed Colbert, Councilmember CITY OF EL CAJON Hon. Harriet Stockwell, Deputy Mayor (A) Hon. Mark Lewis, Councilmember (A) Hon.,Richard Ramos, Councilmember CITY OF ENCINITAS Hon. Maura Weigand, Councilmember (A) Hon. Gail Hano, Deputy Mayor CITY OF ESCONDIDO Hon. Jerry Harmon, Mayor (A) Hon. Lori Holt Pfeiler, Councilmember CITY OF LEMON GROVE Hon. Brian Cochran, Mayor (A) Hon. Jerome Legerton, Councilmember CITY OF NATIONAL CITY Hon. Rosalie Zarate, Councilmember (A) Hon. Michael Dalla, Vice Mayor CITY OF OCEANSIDE Hon. Dick Lyon, Mayor (A) Hon. Nancy York, Councilmember CITY OF IMPERIAL BEACH Hon. Mike Bixler, Mayor (A) Hon. Marti Goeth CITY OF POWAY Hon. Don Higginson, Mayor (A) Hon. Bob Emery, Deputy Mayor (A) Hon. Mickey Cafagna, Councilmember CITY OF SAN DIEGO Hon. Judy McCarty, Councilmember (A) Hon. Tom Behr, Deputy Mayor CITY OF SAN MARCOS Hon. Lee Thibadeau, Mayor (A) Hon. Mark Loscher, Councilmember CITY OF SANTEE Hon. Jack Dale, Mayor (A) Hon. Hal Ryan, Councilmember CITY OF SOLANA BEACH Hon. Marion Dodson, Councilmember (A) Hon. Paul Tompkins, Deputy Mayor (A) Hon. Joe Kellejian, Councilmember CITY OF VISTA Hon. Gloria E. McClellan, Mayor (A) Hon. Scott Packard, Councilmember COUNTY OF SAN DIEGO Hon. Brian Bilbray, Chairman (A) Hon. Pam Slater, Vice Chair (A) Hon. John MacDonald, Supervisor STATE DEPT. OF TRANSPORTATION (Advisory Member) James van Loben Sels, Director (A) Gary Gallegos, District 11 Director U.S. DEPARTMENT OF DEFENSE (Liaison Member) CAPT. Tom Gunn, CEC, USN Commanding Officer Southwest Division Naval Facilities Engineering Command SAN DIEGO UNIFIED PORT DISTRICT (Advisory Member) Jess Van Deventer, Commissioner TIJUANAIBAJA CALIFORNIA (Advisory Member) Hon. Hector G. Osuna Jaime Presidente Municipal de Tijuana Revised September 16, 1993 ABSTRACT TITLE: Demographic and Economic Forecasting Model (DEFM) Forecast 1993 To 2015, Volume 1, Model Overview AUTHOR: San Diego Association of Governments SUBJECT: Regional Growth Modeling and Forecasts DATE: July 1993 SOURCE OF COPIES: San Diego Association of Governments 401 B Street, Suite 800 San Diego, CA 921 01 PAGES: 25 ABSTRACT: This report is a general description of DEFM, including overall model structure, flow of information and general guidelines for its use for long-term forecasting, simulation and impact analyses. ACKNOWLEDGEMENTS The development of the Demographic and Economic Forecasting Model (DEFM) originally was the result of a cooperative effort by the San Diego Association of Governments (SANDAG), the San Diego Gas and Electric Company, and the County of San Diego. This update has relied upon the expertise of many individuals and agencies including: - SDG&E - California State Employment Development Department - County of San Diego Department of Planning and Land Use The following staff of SANDAG participated in the model update: - Kenneth E. Sulzer, Executive Director - Stuart R. Shaffer, Deputy Executive Director - Bob Parrott, Director of Research - Marney Cox, Special Services Director/SourcePoint - Terry Beckhelm, Senior Regional Planner, Project Manager - Jeff Tayman, Senior Regional Planner - Ray Major, Associate Research Analyst - Tony Vasquez, Senior Research Technician TABLE OF CONTENTS 1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . .1 2. GENERAL DESCRIPTION . . . . . . . . . . . . . . . . . . . . .4 DEFM Uses . . . . . . . . . . . . . . . . . . . . . . . . . .6 Flow of Information . . . . . . . . . . . . . . . . . . . . .6 3. DEMOGRAPHICS. . . . . . . . . . . . . . . . . . . . . . . . .9 Introduction. . . . . . . . . . . . . . . . . . . . . . . . 10 Components of Change. . . . . . . . . . . . . . . . . . . . 10 Labor Force . . . . . . . . . . . . . . . . . . . . . . . . 12 School Enrollment . . . . . . . . . . . . . . . . . . . . . 12 Group Quarters Population . . . . . . . . . . . . . . . . . 13 Household Population and Total Households . . . . . . . . . 13 Housing Demand by Structure Type. . . . . . . . . . . . . . 13 Demographic Variable Shifts . . . . . . . . . . . . . . . . 13 4. ECONOMIC SECTORS. . . . . . . . . . . . . . . . . . . . . . 14 Introduction. . . . . . . . . . . . . . . . . . . . . . . . 15 Employment. . . . . . . . . . . . . . . . . . . . . . . . . 15 Income. . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Construction. . . . . . . . . . . . . . . . . . . . . . . . 19 Prices. . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Public Finance. . . . . . . . . . . . . . . . . . . . . . . 21 LIST OF FIGURES Figure 1 DEFM Flow of Information . . . . . . . . . . . . . . .7 Figure 2 DEFM Sector Linkages . . . . . . . . . . . . . . . . 8 Figure 3 Components of Population Change . . . . . . . . . . . 11 Figure 4 Employment Sector Flow of Information . . . . . . . . 16 Figure 5 Public Finance, Revenue . . . . . . . . . . . . . . . 22 Figure 6 Public Finance, Expenditures. . . . . . . . . . . . . 24 1. INTRODUCTION INTRODUCTION Public planning agencies have increasingly recognized three needs: reliable long-range demographic and economic forecasts; quantification of the economic effects of new development proposals or public policy alternatives- and a broad historical database. The need for reliable forecasts is especially acute for agencies involved in planning capital facilities that require long lead times to develop, such as public utilities and transportation systems. Spending limitations and property tax limitations have forced planners to estimate more carefully and fully the economic and fiscal consequences of new development proposals. A reliable database sets the boundaries within which discussions can take place and informed decisions can be made. SANDAG developed the Demographic and Economic Forecasting Model (DEFM) to respond to these needs. The primary function of DEFM is to produce mid and long-range regional demographic and economic forecasts. Essential model input includes local economic and demographic characteristics and national economic forecasts which supply the overall trend for the local economy. Although DEFM's structure emphasizes long-range forecasts, it also produces annual forecasts for about 700 variables. Furthermore, SANDAG developed DEFM with an emphasis on testing alternative assumptions. The DEFM documentation includes five volumes oriented to different audiences'. Volume I, Model Overview. Volume I contains a general description of DEFM, including overall model structure, purposes and general guidelines for its use for long-range forecasting, simulation and impact analysis. Volume II, Technical Description. The Technical Description expands on the topics discussed in Volume 1. It contains a detailed flow of model information, specific sector details, and forecast parameters and equations. Volume III. Database Documentation. Volume III describes the DEFM database including data sources, data file structures and update procedures. Volume IV, Programmer's Reference. The programmer's reference is an informal document that outlines procedures to maintain the model source code, to update model equations and parameters, and to produce executable programs. Volume V. User's Instructions. Volume V serves two purposes. First, it documents the specific assumptions used to produce the Series 8 baseline forecasts, including shift variables and equation 2 residuals. Second, it provides instructions to run DEFM to evaluate alternate scenarios, or to generate impact analysis output. The Appendices to the DEFM documentation are bound separately. 3 2. GENERAL DESCRIPTION GENERAL DESCRIPTION This document describes DEFM in general terms, outlines its basic foundations, and illustrates its use. Volume II discusses technical aspects of DEFM development and operations. DEFM is a simultaneous, nonlinear econometric model designed to forecast population and economic variables for the San Diego region from a set of basic assumptions. The model produces medium to long-term forecasts. Historical data for the period 1950-1993 form the basis for DEFM equations and parameters. Regression analysis of the time series data determined the relationship between exogenous variables and the dependent variables forecast by DEFM. Additionally, DEFM made extensive use of cross-sectional data, including the U.S. decennial census of population and local industry input/output relationships. Forecasts include data for more than 700 economic and demographic variables for the period 1993-2015. In general, the model is a synthesis of two widely used techniques: the cohort-survival method for forecasting demographic variables and relationships, and time series correlation methods for forecasting economic relationships. Cohort-survival considers such factors as birth rates, death rates, and the age, sex and ethnic distribution of the resident population to arrive at forecasts of demographic variables. The econometric approach derives forecasts of employment, income and other economic variables from national, state and local growth patterns, and interindustry relationships. In DEFM, the linkages between the demographic and economic sectors are explicit. Job opportunities and relative employment levels in the region determine migration of the working age population and dependents. Changes in employment are based on regional growth in population and personal income, as well as projected state and national economic conditions. A simultaneous solution method reconciles the complex relationships among population, income and employment variables. Employment is forecast for each two-digit Standard Industrial Classification (SIC) industry. Most of the employment equations have a common form that specifies employment in each category as a linear function of a composite index developed from a set of explanatory variables. The index variables include population, employment and disposable income in the region, as well as state and national population and income levels. Direct interindustry gross flow fractions are used in combining explanatory variables into a single scale variable index for each employment category. Most employment equations use this single scale variable index as the key explanatory variable. 5 The exogenous assumptions in DEFM influence the forecasts and can be controlled by users. Users may alter input data, demographic parameters and economic coefficients, or may introduce other changes (shifts) to alter the model outcome. Each of the approximately 200 shifts serves one or more purpose in the model, depending on its location. DEFM USES DEFM was originally developed through a cooperative modeling effort that combined the capabilities of various independent forecasting tools. It has been expanded and modified extensively over the years. The current version of DEFM represents SANDAG's approximation of the complex economic and demographic relationships of the San Diego planning region. DEFM is a means to study long-term planning issues, such as capital facilities planning, transportation planning, and growth management. DEFM forecasts have a broad range of support and acceptance. They are reviewed and accepted by both local and state agencies as the official economic and demographic forecasts of the San Diego Region. Virtually all public and private planning agencies that require demographic and economic data rely on DEFM forecasts. One important application of DEFM as a planning tool is impact analysis. This is generally accomplished with population characteristics, employment or income shifts to alter economic or demographic forecasts, or to alter relationships among variables. The model is run with these modified input variables and the resulting forecast is compared to a baseline forecast. The differences can be attributed to the changes in the input variables. Thus, DEFM provides a method to quantify the economic effects of new development proposals, alternative development strategies or changes in the general level of economic activity. DEFM is a planning and analysis tool. It is designed to help evaluate the consequences of basic assumptions about the relationships among employment, population, housing, income and other socioeconomic variables. Those consequences can vary considerably with different basic assumptions. There is no single future. Rather, DEFM provides a means to evaluate possible futures according to different outlooks about influential economic and demographic variables and relationships. DEFM cannot produce reasonable forecasts without informed input and analyses. FLOW OF INFORMATION Although DEFM has several major submodels, the general flow of information and forecast solutions are illustrated in Figure 1. The linkages among with major sectors are illustrated in Figure 2. The resident survived population determines housing demand, the demand for public facilities and associated public finance, influences employment demand and provides the labor force. Employment demand is determined by local and national economic conditions and demographic variables. DEFM seeks equilibrium by reconciling the demand for and supply of labor, and the demand for and supply of housing. As unemployment rates rise or decline, the labor force is adjusted by changes in labor force participation and migration. As housing demand 6 exceeds housing supply, new construction occurs. Prices respond to increases in housing costs and wage rates. The forecast sectors and mechanisms are discussed in detail in the following sections. Click HERE for graphic. 7 Click HERE for graphic. 8 3. DEMOGRAPHICS DEMOGRAPHICS INTRODUCTION Population is forecast for five year age groups by sex and ethnicity for civilian and total population. (Uniformed military population and the associated age, sex, and ethnic group distributions are exogenous data.) The four ethnic groups are defined as: (1) Hispanic, (2) Non-hispanic White, (3) Non-hispanic Black, and (4) Non-hispanic Asian and Other. Summary data for each ethnic group include total and civilian population, births, deaths, total fertility rate, net migration, and population by age group aggregates (0-14, 15-64, 65+). Ethnic detail is also forecast for labor force and school enrollment by five levels (nursery school, kindergarten, elementary, high school, and college). Some aggregate forecasts, including group quarters population, household population and heads of household, are produced by age and sex, but are not delineated by ethnic groups. The demographic forecasts are based on age, sex and ethnic detail of the population in the decennial census year. Initially, population components are calibrated to census totals or proportions. The adjusted civilian population is determined by subtracting military dependent population from the civilian (e.g., excluding military in uniform) population. For each year between the census (I 990) and the forecast base year (I 992), detailed components of the population are calculated and controlled to totals contained in the input data base. COMPONENTS OF CHANGE Population is forecast using a modified cohort-survival method as illustrated in Figure 3. Each year's civilian population is the sum of the previous year's civilian population, the natural increase and net civilian migration. Total population is equal to civilian population plus uniformed military population. Most of the computations are performed for population subgroups by age, sex and ethnic group. Summing across ethnic groups yields total population by age and sex. Natural Increase Each forecast year's natural increase is equal to births minus deaths. Births are forecast from last year's female population aged 15 through 44, using exogenous age-specific birth rates for women in each ethnic group. The births are split into male and female components using historical ratios, Births to military dependents in each ethnic group are based on the military population and are not 10 Click HERE for graphic. 11 detailed by age of mother. Births by ethnic group are summed to obtain births by age of mother in the civilian population and total births to military dependents. Deaths are forecast from last year's adjusted civilian population using exogenous age- and sex- specific death rates for each ethnic group. Birth and death rates are specified at ten-year intervals and intermediate rates are interpolated for each forecast year. Deaths are subtracted from the previous year's adjusted civilian population. Survived adjusted civilian population is computed by moving one fifth of the resulting population in each age group into the next age group. Births to the adjusted civilian population are added to the first age group to complete the forecast of survived adjusted civilian population. Births to the adjusted civilian population plus births to military dependents, minus deaths equals natural increase. Migration Net economic migration is based on San Diego's unemployment rates and on recent changes in local employment levels. Net retirement-aged migration is based on age-specific migration rates for the U.S. population aged 65 and over living outside of San Diego. These rates are influenced by changes in relative housing costs in San Diego. Migration components are further divided by age-, sex- and ethnic- specific rates. First, the estimates of net economic migration and net retirement-aged migration are distributed to ethnic groups. Then age-, and sex-specific migration rates for each ethnic group are used to estimate detailed migration components for each ethnic group. Net migration of the adjusted civilian population is the sum of economic and retirement migration. The rates and factors are specified at ten-year intervals and intermediate rates are interpolated for each forecast year. Military The uniformed military population is an exogamous input. Military dependent population is estimated as a proportion of the uniformed military population. Uniformed military and military dependents are summed to obtain total military population by age, sex and ethnic group. Subtracting the previous year's military dependents and forecast military births from the forecast of military dependent population gives net change in military dependents, or net military dependent migration. Net change in uniformed military plus net change in military dependents equals net military migration. Age, sex and ethnic group distribution factors are specified at ten-year intervals and intermediate rates are interpolated for each forecast year. LABOR FORCE Civilian labor force is calculated from the civilian population by multiplying population in each age (I 5-74), sex and ethnic group by labor force participation rates. The labor force participation rates are specified at ten year intervals and the rates are interpolated for each forecast year. 12 SCHOOL ENROLLMENT School enrollment is forecast from civilian population by age, sex, and ethnic group for five levels: nursery school (ages 0-4), kindergarten (ages 5-9), elementary school (ages 5-19), high school (ages 10-24) and college (ages 15-54). School participation rates for each level are specified at ten year intervals and the rates are interpolated for each forecast year. GROUP QUARTERS POPULATION The population living in group quarters is forecast by age and sex for uniformed military living in barracks or onboard ship, college students living in dormitories, and for other persons in group quarters accommodations. Other group quarters population includes persons living in boarding houses, homes for the disabled, rest homes, jails and other group living situations. The age and sex specific rates for group quarters population are held constant throughout the forecast period. HOUSEHOLD POPULATION AND TOTAL HOUSEHOLDS Household population is calculated by subtracting the total group quarters population by age and sex from the total population. Age- and sex-specific household headship rates (fraction of population that is a household head) determine the number of heads of household. HOUSING DEMAND BY STRUCTURE TYPE Household headship, or demand for housing units, is further divided by structure type using housing preference factors by age and sex for single family structures, multifamily units and mobile homes. The household headship rates and housing preference factors are specified at ten year intervals and interpolated for each year during the forecast. Vacancy rates are computed for all housing units and for housing units by structure type. DEMOGRAPHIC VARIABLE SHIFTS The user-determined changes that can be introduced in the form of model shifts are varied. The following may be effected: - response of economic migration to job growth - direct multiplicative and additive shifts in economic migration - responsiveness of migration to relative unemployment conditions - direct multiplicative shifts on the net migration rates - sensitivity of retirement migration to relative housing costs - direct additive and multiplicative shifts on retirement migration. Additionally, by using combinations of the shifts, both economic and retirement migration can be assumed to continue at pre-selected target levels. These model shifts should be used in a manner that maintains internal consistency within the model. Shifts are described in detail in Volume V. 13 4. ECONOMIC SECTORS ECONOMIC SECTORS INTRODUCTION The economic sectors of DEFM provide the demand for labor services. DEFM maintains separate submodels for employment, income, construction, prices, public facilities demand and public finance. Each of these is described in the paragraphs that follow. EMPLOYMENT The flow of information for the Employment sectors is illustrated in Figure 4. Employment levels are forecast for two-digit SIC industries. With a few exceptions, including agriculture, construction and some miscellaneous categories, the employment forecasting equations have a common form. This form specifies employment in each industry 0) as a linear function of a composite index (C) of explanatory variables: Employment = a + b * C j j j j For most employment categories, DEFM is based on the premise that employment is determined by a combination of factors reflecting local, state and national economic conditions. These factors may include population and income, employment levels in closely related industries, cost conditions, and so on. Since it is difficult to attempt to sort out all the interrelations by statistical procedures alone, a composite index is constructed as a single scale variable index. Employment Indexes One source of explanatory variables for most employment equations is the set of final demand coefficients from the revised 1975 input/output Table developed by Copley International Corporation as a part of the San Diego County EIAS model. In particular, the direct gross flow fractions are used to reflect the relative importance of each industry in the composition of a single scale variable index. The estimated relationships for the employment equations use this single scale variable index as the key explanatory variable. The gross flows input/output table has the form of a 52 industry by 52 industry matrix of interindustry fractions augmented by a 52 industry by 7 sector matrix of final demand fractions. Together these form a 52 industry by 59 sector matrix of weights that reflect each sector's relationship to the remainder of the economy including all sources of final demand. The composite index (C) for each sector 0) used to estimate employment is the product of the various final demand factors: 15 Click HERE for graphic. 16 C = W1 * W2 * W3 * W4 * W5 * W6 * W7 j j j j j j j j Each component of the index is briefly described below. A more detailed description of the derivation of the employment index variables is presented in Volume II. Interindustry Weights. The first term, W1, is constructed from industry employment indices, or growth rates, and the inter-industry gross-flow fractions. from the I/O Table. This term is included to capture the effects of regional scale and the regional industrial mix on employment in an individual SIC group. Final Demand Weights. For manufacturing industries, the second term, W2, in the scale variable index is the geometric average of national employment growth and lagged growth in the industry as the key export- based explanatory variable. National employment is weighted to reflect the final demand from foreign countries, and the remainder of the U.S. If these final demand fractions are small, indicating an industry that primarily serves local demand, the term receives proportionately smaller weights in constructing the scale variable index. Conversely, if these final demand fractions are large, indicating an industry that services national demand, the term receives a proportionately larger weight in constructing the scale variable index. For each manufacturing industry, this provides the primary link between local employment in an industry and national conditions for that industry. For most San Diego manufacturing sectors, the directly comparable U.S. employment levels are used in the corresponding indices. For nonmanufacturing industries, the second term, W2, is the geometric average of U.S. income and population terms rather than the national employment components. Final demand fractions that population and income are used as indicators of the importance of the California economy. These fractions are used as weights in introducing the third term, W3. The next three terms, W4, W5 and W6, reflect the importance of sales to military bases, state and local government agencies, and other federal government, respectively. The final term, W7, introduces an index of regional income and population into the scale variable index. It assumes that both current and lagged real income and population growth are important determinants of local employment. Nearly all employment forecasting equation parameters are obtained from ordinary least squares (OLS) regressions. Construction employment is assumed to depend on current and lagged values of the real value of all construction authorized. Federal government employment is assumed to depend on national federal government employment (for federal nondefense), and uniformed military (for federal defense). State government employment is assumed to depend on state university enrollments (for state education) and population (for other state employment). Local government employment is assumed to be limited by local government revenues available for payroll (a revision motivated by 1978 California legislation limiting property taxes). 17 Employment Equation Shifts Shifts may be introduced into each of the employment equations in those cases where special information suggests a change in the historical relationships. Shifts are described in detail in Volume V. INCOME DEFM forecasts personal income from several sources, including payroll, other labor income, proprietors' income, dividends-interest-rent, and a variety of transfer payment sources. Civilian payroll is calculated as the sum of forecast payrolls in 57 employment categories. The payroll forecasts are based on the individual industry employment forecasts, industry specific payroll/employee coefficients and general wage rate trends. Use of the individual payroll/employee coefficients captures the effect of anticipated changes in the regional industrial mix on regional payroll levels. Social Security contributions are assumed to depend on the level of civilian payroll. Income from sources other than civilian payroll is forecast with regression equations. Forecasts of military payroll depend on the number of uniformed military and general wage trends in the private sector. Other labor income is assumed to depend on civilian payroll levels. Forecasts of proprietors' income depend on the levels of labor income. Dividends, interest and rental income depend upon regional population and national income from the same sources. Transfer income is forecast for several components. Again, the equation parameters are the result of a set of OLS regressions. Retirement transfers are linked to population over 65 years of age. Public assistance transfer income depends upon local government expenditures and resident population. Unemployment transfer income depends on the number of unemployed persons in the economy. The residual category (other transfer income) is assumed to depend on total population. The income components are combined to obtain regional personal income (after subtracting Social Security contributions.) A residency adjustment is made to account for the income of local residents who work outside the area and for the income of people who work locally but live outside the region. Disposable personal income is calculated by subtracting forecasts of personal tax payments from total personal income. Taxes paid out of personal income include federal income taxes, state income taxes and personal property taxes. The two income taxes are forecast using average tax rates applied to personal income less relevant deductions (tax payment deductions are incorporated at approximate marginal rather than average tax rates). 18 Income Equation Shifts Various shift forms are available for the income components to alter the forecasting equations. See Volume V for details. CONSTRUCTION Construction in DEFM includes residential and nonresidential activity. Housing Supply Housing units are forecast by structure type as a function of the previous year's stock plus the completion of last year's residential permit authorizations. For single and multifamily housing stocks, an historically estimated relationship translates new authorizations into stock changes. This relationship implicitly adjusts for less than 100% permit realization rates as well as for replacements for demolitions. The stock of mobile home units is forecast to increase at the average historic rate. Vacancy Rates Housing Demand, or the number of occupied units by structure type, is.computed in the Demographic submodel based on exogamous housing preference factors and household headship rates. Vacancy rates are computed from supply and demand for housing units by structure type and for all housing units. Housing Unit Authorizations The number of new housing units authorized by building permits in a forecast year is positively related to national housing market conditions, and inversely related to the housing stock vacancy rate. It is also dependent on two types of local housing demand -- the increase in the number of households and the replacement of old housing stock. No consideration is given to local mortgage market conditions since secondary mortgage markets rather than local deposit inflows now serve as the main source of mortgage funds. Given these secondary markets, the national housing market variables are assumed to account for national and therefore local mortgage market conditions. Permit activity is allocated between single and multifamily authorizations on the basis of changes in housing demand by housing type. The housing demand levels are forecast by applying 1990 census housing type preference factors to the forecasts of heads of household by age and sex. Construction Value Residential. The nominal value of authorizations for new residential units is influenced by the level of construction costs and the level of single family and multifamily permits (adjusted to reflect cost differentials between the two housing types). 19 Nonresidential. Valuation of new nonresidential and other (e.g. repairs and remodeling) construction is forecast by a regression equation which is based on the historical value of nonresidential construction investment nationally. Construction Sector Shifts The user-determined shifts in this sector are complicated and interrelated. Effective use of these shifts requires a thorough knowledge of the model and careful attention to the placement of shifts in each equation (one shift can appear in up to five equations and some shifts appear in several subsectors). Special attention to internal consistency within the model, the dynamics of the model, and the relations among subsectors is essential to effective use of these user- determined shifts. See Volume V for details concerning shifts. PRICES Several price variables appear in the model, including the consumer price index and its shelter component, the average price of a single family home, a construction cost index and wage rates. All real income and price data in DEFM are based on 1992 constant dollars. All price deflators are computed from the U.S. Department of Commerce Consumer Price Index, with 1982-1984 average = 100.00. Construction Cost Index Construction costs are forecast to remain tied to national construction levels. Housing Prices Housing price level forecasts are based on an historically estimated relationship in which the change in housing prices is assumed to depend on growth in construction costs. CPI Shelter Component The shelter component of the CPI is assumed to move with its national counterpart except for the influence of local housing prices. The forecasting parameters are determined by an ordinary least squares regression using historical data. Consumer Price Index The local CPI is assumed to follow its national counterpart up to the influence of differential changes in local and national shelter components. This latter differential is reflected through the current relative importance of the local shelter component in the local consumer price index. 20 Wage Rate The manufacturing wage rate is forecast to remain tied to national manufacturing wage rate levels. PUBLIC FINANCE DEFM contains data on revenues and expenditures of the County, the Cities, and School Districts. Revenues include property taxes, retail sales tax, federal and state grants, and other revenues (fines, service fees, assessments, etc.). Expenditures include educational expenditures and other local government expenditures. Revenue The flow of information for public revenue is illustrated in Figure 5. In DEFM, local government operations are assumed to be jointly constrained by revenue considerations and mandated Gann Appropriation limits. Before 1979, the scale of local government activity increased substantially relative to the economy as a whole. In the current model, the increase in certain categories of government expenditures is assumed to be limited to the rate of increase in specific economic indicators. The calculations are similar to those used at the state level. Funding for K-14 education is guaranteed at a minimum of 40% of the budget. The education spending limit is calculated as the prior years funding level adjusted for workload and adjusted for state per capita increase in either personal income or general fund revenues (which ever is less). Several categories of expenditures and revenues are assumed to remain at their real per capita 1978 relationships, reflecting the base year for the voter approved Gann spending limitation. Since this change in structure represents an extreme departure from historical relationships, it is important that the user have a thorough understanding of each equation in this sector. Assessor's Market Value. The tax base for property taxes is termed "assessor's market value" (AMV) and is computed separately for residential and nonresidential property. In each year following 1976, residential and nonresidential AMV are adjusted according to four criteria. First, for components of AMV that have not yet changed owners, the AMV is assumed to have increased by the smaller of two percent per year and the rate of inflation, as measured by the CPI. Second, for components of AMV that have changed hands, the, market value is assumed to have increased by the percentage change in housing prices during the most recent six years. Third, for new construction since 1976 that has not changed hands, the original value is inflated at the rate of the minimum of two percent per year and the rate of inflation. Finally, for new construction since 1976 that has changed hands, the value is inflated at the rate of change in housing prices since the year of construction. While these steps sound simplistic, they are designed to approximate legislation. The actual computations are complex. 21 Click HERE for graphic. 22 Property Tax Revenues. Property tax revenues of local government agencies are forecast from the assessor's market value tax base. Historically, the overall average rate for total property taxes collected (since Prop. 13) reflects the one percent full value tax and additional debt service in voter-approved tax rate areas. In DEFM, a real per capita property tax revenue limit of $832 (in 1992 dollars) is assumed throughout the forecast period. This represents the 1977 level of per capita property tax revenue. Starting in 1981, the property tax equation allows for tax rebates. Revenue collections in excess of the mandated appropriation limits must be returned to taxpayers under the Gann Initiative. Federal Grants. Federal revenue sharing has been discontinued. Other federal grants are assumed to continue at their real base year per capita levels. User-determined shifts are available to alter this forecast assumption. State Grants. These are divided into two categories Proposition 13 local relief funds and non-Proposition 13 grants. Proposition 13 relief funds are assumed to be at the base year per average daily attendance level. The permanent bailout, AB 8, primarily allocates relief funds between school districts and county health and welfare programs. In light of this allocation scheme, county relief funds are projected on the combined basis of population growth and ADA increases. For other 13 grants, it is assumed that per capita grants are based on California population modified to account for the share of state budget going to local assistance and Proposition 13 relief In addition, they are scaled over time to account for forecast changes in San Diego's share of state population. Total Retail Sales Tax. Retail sales tax revenue is the one percent of taxable sales (or 1/6th of the six percent sales tax), levied by the state, that is returned to local jurisdictions. After 1988 and continuing until 2008, the 1/2 percent local transportation sales tax, TRANSNET, is added. Other Revenues. Other revenues include fines, fees, service assessments, etc. These revenues represent most of the unconstrained local revenues not affected by the spending limitations. Since 1979, these revenues grew proportionately faster than the other sources of revenues and now make up the majority of non-education related local government revenue. Since these revenues are closely linked to local economic and demographic conditions, an historically estimated equation is used for forecasting purposes. This equation relates other revenues to absolute population levels and civilian migration. Expenditures The flow of information for public expenditures is illustrated in Figure 6. 23 Click HERE for graphic. 24 Educational Expenditures. Educational expenditures, are assumed to grow at the combined rate of average daily attendance (ADA) growth and the lesser of either the growth in per capita state income or the growth in per capita general fund revenues. Expenditures are projected on a ADA basis using an inflated ADA multiplier. Local Government Payroll. The level of local government payroll is assumed to maintain its historic relationship with expenditures for education and local government expenditures other than public assistance expenditures. 25 .
