Loading...
HomeMy WebLinkAbout851239.tiff Table 46 Descriptive Statistics of the Alternative Pooled Water Demand Model Standard Variable!/ Mean? Minimum Maximum Deviation Total water use per household per day (gal) 491.2 229.6 968.2 148.2 Median household income (1,000's) it lot size 7.08 1.49 28.15 4.95 Service sector employees per household 0.165 0.00 0.66 0.149 Non-service sector employees per household 0.998 0.00 4.61 0.759 Percent single family household a lot size 22.5 0.0 91.7 16.1 (lot size)2 0.099 0.02 0.84 0.149 Number of days of measurable precipitation 44.0 31.0 57.0 7.8 Three day, three hour watering restriction 0.071 0.00 1.00 0.257 Marginal price (1982 dollars) 1.12 0.00 2.34 0.44 1/ See Chapter 3 for detailed definition of variables. ?/ Number of observations = 254. Appendix 2 130 851239 elasticities for the complex model are similar to the elasticities for these variables in the linear model as shown in table 47. The coef- ficient of determination (R2) is 0.63, which approximates the explana- tory power (R2 of 0.64) of the linear model. Statistical tests for multicolinearity, heteroscedasticity and serial correlation reveal approximately the same characteristics as the linear model. Correlation among the interaction from variables exceeds the highest correlation found among variables in the linear model. Heteroscedasticity is likely to be present for several independent variables. As with the linear model, tests of serial correlation are inconclusive. The results of this model are similar to the simple linear model presented earlier. The explanatory power of the complex model is com- parable to that of the linear model. Because the more complex model form fails to improve upon the results of the simple linear model, and because variable interaction exceeds that seen in the linear model, the simple linear model is considered preferable to the more complex form. Appendix 2 131 Table 47 T-Statistics and Elasticities for the Alternative Pooled Water Demand Model Standard T- Elas- Variable Coefficient Error Statistics ticity Median household income (1,000's) * single family lot size 28.8 3.6 8.07 1/ Percent single famly * single family lot size 1.21 1 .29 0.93 1/ Lot size?/ (393.6) 125.7 (3.13) 1/ Number of days of measurable precipitation (3.51) 0.77 (4.59) (0.31) Presence of third day, three hour restrictions (33.6) 22.8 (1.47) (0.01) Marginal price (57.7) 14.0 (4.12) (0.13) Persons per household 41.2 16.8 2.45 0.25 Service sector employment per household 213.0 54.1 3.93 0.07 Non-service sector employment per household 44.9 10.0 4.50 0.09 1/ In these forms, elasticities cannot be meaningfully described. Income elasticity is 50.39. Percent single family elasticity is 0.05. Lot size elasticity is 0.38 Appendix 2 132 CHAPTER 5 USE FACTOR WATER DEMAND FORECASTING !MODEL • CHAPTER 5 USE FACTOR WATER DEMAND FORECASTING MODEL INTRODUCTION The pooled water demand model described in Chapter 4 was prepared in aggregate form, that is, without segregation by customer class. A thorough conservation analysis is believed to require a water demand forecast and model which disaggregates by customer class. GKY and Associates developed a forecasting model to meet this requirement. The logic of the use factor demand forecasting model is two-tiered. The first tier is represented by use factors for different classes of water consumers. These use factors incorporate average or typical water demand and can be used to make respectable first approximations of demand forecasts. Total water demand for a water supplier is, there- ...-. Appendix 2 133 fore, primarily driven by demographics; the number of households, the number of employees, and total population. The second tier of logic recognizes that there are differences in the socioeconomic characteristics (for example, lot size and family income) of water suppliers which affect water demand. This observation allows fine-tuning of the first approximation estimates. This tier is added to the model by regression analysis of household level data. The use factor model was developed using several data sources (described in Chapter 3) both at the water supplier level and the house- hold level covering the period 1974 through 1982. The method: . Isolates SF metered, SF flat, MF, commercial and industrial, and public use sectors as well as unaccounted for water (losses). . Is applicable to individual water districts with differing socioeconomic characteristics. . Utilizes DRCOG independent variable projections. . Is consistent and compatible with the conservation analyses. MODEL DEVELOPMENT Each of the two parts of the use factor water demand forecasting model were developed separately. These analyses are outlined below. Appendix 2 134 USE FACTORS Use factors for each of the five water consuming sectors were developed as described in Chapters 2 and 3. These use factors represent typical average use for the particular sector in the absence of any con- servation programs. They are reviewed in table 48. SF households (metered) typically demand water at a rate of 478 gallons per household per day and taken together account for approxi- mately 42 percent of demand in the EIS area. Forty-six percent of this is estimated to be used indoors and 54 percent outdoors, primarily dur- ing the summer months. Similarly, SF households (flat-rate) demand water at a higher rate of 630 gallons per household per day and account for 23 percent of total demand. However, only 37 percent is indoor use. The final residential use sector, MF households, demands water at a rate of 217 g.h.d and represents 16 percent of total demand. Eighty-three percent of this is used indoors. The two nonresidential sectors account for smaller portions of total EIS demand. Commercial and industrial users demand 12 percent of the total at a rate of 45 gal/employee/day. Similarly, public use accounts for seven percent at a rate of 14 gal/person/day. Neither of these factors were divided into indoor and outdoor usage. SOCIOECONOMIC ANALYSIS The second portion of the model development was to incorporate individual district characteristics (lot size, family income, marginal price, household size) which would contribute to variations in water demand from district to district. However, this analysis was only per- formed for the largest sector: SF households (metered). It is reason- Appendix 2 135 Table 48 Review of Use Factors Use Factor (gal/unit/day) Sector Indoor]/ Outdoor?/ Total Single family households (metered) 220 258 478 Single family households (flat rate) 230 400 630 Multifamily households 180 37 P17 Commercial and industrial (employment) NA NA 45 Public (population) NA NA 14 if Indoor use is estimated from winter use assuming no outdoor consumption. 2/ Outdoor use is the difference between total use and indoor use. Although this use primarily occurs in the summer at greater magnitudes, these num- bers represent that use evenly spread throughout the year. NA a not available. Appendix 2 136 able to suspect that analogous adjustments might be made to the other four use factors: SFF, MF, E, and P. However, it is less certain that the improvements in model sophistication would justify additional analy- sis. Such analysis on the SFM use factor was justified because this class of consumers demanded more water (42 percent) than any other class. SFF customers are the next largest group at 23 percent of total demand. However, since decline in the number of SFF households will occur in the future, this class of customer will become decreasingly important. The other three classes of consumers currently represent 35 percent of demand with the MF class being the largest of the three at 16 percent. The relative absence of data in the literature for these cus- tomer classes support the claim that these classes are difficult to ana- lyze. This combined with their smaller contributions to current SEIA area demand led to the decision not to go beyond simple demographics in the model for these customer classes. IDENTIFICATION OF SOCIOECONOMIC VARIABLES Several factors influence water demand at the household level; some are important, others are not. To identify the most relevant variables, a correlation analysis was performed, using the merged household level data base of EDF and Entercom data as described in Chapter 3. The cor- relation between seven variables was performed: SFM water demand, lot size, marginal price, income, household size, impervious area ratio (impervious area/total area) and assessed valuation. Results of the analysis on the disaggregated (household) and aggregated (household by district) data bases are shown in tables 49 and 50. For both data bases, the highest correlation occurred with demand and income followed by assessed lot and house value, which itself is highly correlated with income. For the disaggregated data base, the Appendix 2 137 Table 49 Socioeconomic Variables - Water Use Correlations 1982 EDF and Entercom Database (Disaggregated) Correlation Coefficients/Number of Observations DEMAND LSIZE MARPR INCOME HHSIZE Cl/ ALHV?/ DEMAND 1.00 .268 -.132 .396 .249 -.0457 .371 592 592 592 592 592 530 561 LSIZE 1.00 .118 .215 .0337 -.542 .290 592 592 592 592 530 561 MARPR 1.00 .0793 .0283 -.00480 592 592 592 530 561 INCOME 1.00 .0947 .102 .357 592 592 530 561 HHSIZE 1.00 .0294 .139 592 530 561 C!/ 1.00 .0411 530 526 ALHV?/ 1.00 561 ?/ Impervious area ratio (impervious area/total area). V ALHV Assessed lot and household values. Appendix 2 138 Table 50 Socioeconomic Variables - Water Use Correlations 1982 EDF and Entercom Database (Aggregated) Correlation Coefficients/Number of Observations DEMAND LSIZE MARPR INCOME HHSIZE Cil ALHV?/ DEMAND 1.00 .0537 -.309 .777 .209 .459 .689 15 15 15 15 15 15 15 LSIZE 1.00 .0696 .0952 -.306 -.554 .397 15 15 15 15 15 15 MARPR 1.00 .141 .0145 .151 .153 15 15 15 15 15 INCOME 1.00 .177 .611 .826 15 15 15 15 HHSIZE 1.00 .169 .0411 15 15 15 C 1.00 .470 15 15 ALHV 1.00 15 1/ Impervious area ratio (impervious area/total area). ?/ ALHV Assessed lot and household values. Note: Aggregation is to water district. Appendix 2 139 remaining variable correlation coefficients ranked as followed: lot size, household size, marginal price (inversely) and impervious area ratio (inversely). The remaining variable correlation coefficients for the aggregated data base differed in their ranking: impervious area ratio, marginal price, household size and lot size. The correlation coefficients for lot size and impervious area ratio showed the greatest deviation from the correlation coefficients computed using the disaggregated data base. An additional correlation analysis was performed using the 1982 demands from the EDF and Entercom aggregated data base and 1982 socio- economic variable values from the historical supplier level data base. Results are shown in table 51. Tables 50 and 51 represent aggregations to the same sample size using different data sources; they do not agree particularly well. The highest data correlation was found between demand and lot size followed by income, marginal price (inverse) and household size. Income and lot size were also highly correlated. These analyses do not provide clear indication of the superiority of any of the variables over another. However, because of the rela- tively high correlations between ALHV and Income, ALHV will be dropped from further analysis to avoid later multicollinearity problems. Simi- larly, C will be dropped in deference to LSIZE. This leaves four socio- economic variables for further consideration: LSIZE, MARPR, INCOME, and HHSIZE. OLS !MODELS Once selection of the meaningful variables had taken place, several alternative OLS mdoels were reviewed and still others were tried using a variety of data bases. Appendix 2 140 Table 51 Socioeconomic Variables - Water Use Correlations 1982 EDF•and Entercom and 1982 Historical Supplier Level Database Correlation Coefficients/Number of Observations DEMAND LSIZE MARPR INCOME HHSIZE DEMAND 1.00 .583 -.309 .540 .258 15 15 15 15 15 LSIZE 1.00 -.084 .551 .164 15 15 15 15 MARPR 1 .00 .232 .190 15 15 15 INCOME 1.00 .756 15 15 HHSIZE 1.00 Note: Aggregation is to water district. Appendix 2 141 Prior Denver analyses. Two econometric models, on based on aggre- gated data (Model I) and the other on disaggregated data (Model II), were developed by Morris and Jones (1980). Both models were estimated by OLS and specified the linear relationships between household water consumption and a number of socioeconomic variables. The analyzed data base included 384 households in 21 water districts. In Model I, an OLS multiple regression procedure was applied to explain the average annual household consumption within each of the 21 water suppliers as a linear function of income, lot size and marginal price. Model II also estimated annual demand by ordinary least squares. Individual household data within the sample of SFM households was used in specifying the model. The model included 11 socioeconomic variables, eight more than Model I. Of the two models, Model I had less varince within the error terms since aggregation by district tends to reduce variation found at the household level. Each model assumed the linear specification: D = a0 + at dP + a2W + a3 Z1 + .... + an+2 Zn + e. D = Water Demand. dP = Marginal Price. W = Wealth Zi = Other Influences on Water Demand. Appendix 2 142 e = Error Term. Both models confirmed the significance of family income and mar- ginal price (as found in literature reviews) on water consumption in the Denver metropolitan area, and identified lot size as an equally impor- tant variable. Except for the income variable, the estimated parameters of Model II on specific household data are similar to the parameters estimated in Model I when the same variables are included in both speci- fications. The Model I resultant equation is as follows: D (g.h.d.) = 188 - 263 dP($1/1000 gal) + 11.8 W($1000) + 697 A (2.1) (-2.3) (3.7) (2.8) (acres) R2 = .65 F statistic = 10.55 A = Average lot size The numbers in parentheses are the t-statistics of the estimated parameters. Current analyses. 0LS analyses of the water use and socioeconomic descriptors in the various data bases were performed in order to deter- mine the sensitivity of the demand to the socioeconomic variables. Each model estimated residuals and assumed the linear specification: D - = a1 (A - A) + a2 (dP - dP) + a3 (I - I) + aq (HH - HH) ± SE A = lot size; I = income; HH = household size. The models forced the intercept to zero; however, an intercept is implicit in determining residuals. Table 52 summarizes various models tested, their resultant coeffi- cient (ai) and t-statistics, and the coefficient of determination (R2) for the equation as a whole. Also shown are the model developed by Appendix 2 143 T CO N a II a e I a • S. a .y • a " • O O o.1 O L pa " ' Cl...O .i.i po Sp 001. G O I N a S a > a LOG S O OOO3 • 6 a 2• AZIta h i ='s0 • • E• H 31 fla 8 MI IN ' P in r1 a a a in N� IT CO ' N N t a N p y , •►-I N en ^1 • NI N a a MI O a •p . • ai a U1 I N I 10 a 9 e CO ! T m m Co M m C Z .m.. .m.. .m. a, .m. .T.. w y c a ' I� .O If1 ��pO N O�• • • �p• S II 0S �T m N f'f T N O a C q _ _ $ O.ai NI N 0 N N - a y O. •, N P IA T a S 8 a Al CO 0 ;, In 5 w • i 8 •m m • v m v • N ate. I. N .ai G 'H w e• �. N • h Y a G Q -I ern CO N• N, in O. CO p In a • O y �. G 0C.11r CJ r M O — 76 O t I pp� • ' O vo NI N NIVI O 1;.1 O 1.1 b H e a • • y 1I N I I N nI N N CO. a n. it .. a a • C m 5 y ...0 .- N T O Cr. L.i .. • S. y tl i ■( I O N P CO CO T P 16 1I O. N 01 .O a a N M tI I 1 1 I I I1 1 a N5 p •+ NI O T in N O CO 0 ..4i4 .... 5 • O a 1 N in N a, ■ O tr a ® • > O .1 • N N N 01 N N N in N ~ y p a to G g .. .. .. .. .. .. .. .. Cr. Q a• r. 3 a'lar N in .0N0 T b 01 P Y Y 3a a 2 a aa ■ 1I1�'f CI nil !. I I pa CC a FC f m le f l V a 9 a e .4 ■a t O yp8] a a y ■ N pQ• ..1 11 • ..4i O T ON I. q ■m O y I. i s 6 .5P ® • a ® i is p 0 ^I NI .. 1l Appendix 2 144 Morris and Jones-EDF (1980), and the residuals model using the histori- cal supplier-level data base. The R2 value is the percent of the vari- ation of the annual demand as explained by the given regression. In a two tailed test at a five percent level of significance, a t-statistic greater in magnitude than 1.96 implies significance for large n. In all of the disaggregated models, lot size, income and household size were highly significant at the five percent level. Marginal price was negatively correlated with demand in all of the tested models and was highly significant except when the Entercom data base was analyzed separately. The R2 values ranged from .154(Entercom) to .326(EDF). Results using the aggregated models of the same data bases deviated greatly from the results of the disaggregated models. While the R2 values improved, the t-statistics indicated less predictive success and, in some cases, the sign of the coefficients were revised probably due to the small sample size. The aggregated EDF (1976) model determined lot sizes, marginal price (inverse) and income to be significant and resulted in a R2 value of .65. Analysis of residuals in the historical supplier level data base resulted in lot size and marginal price (inverse) being significant. Income and household size were not sig- nificant. Income and lot size were highly correlated in the supplier data base. An R2 value of .290 resulted from using this data base. The combined EDF and Entercom disaggregated model is the preferred model based on the following: . The disaggregated models are all consistent in that the coeffi- cients for the four independent variables are all significant. The com- bined model (EDF & Entercom) coefficients are between the coefficients found using EDF and Entercom data sets alone, as one would expect. The Appendix 2 145 disaggregated models incorporate lot size, marginal price, income and household size effects on water use. . The combined EDF and Entercom disaggregated model implies elas- ticities that are consistent with values found in the literature. . The combined EDF and Entercom model utilizes all the available data, has a large sample size, and offers coefficients that are con- sistent with values found in the literature. YODEL SELECTION The use factor water demand forecasting model is a combination of the use factor and OLS analyses. However, several adjustments were made to account for differences in the data bases used in the analyses and to ensure accurate forecasts of the entire EIS area demand. INITIAL FORMULATION Applying the results of the OLS analysis directly to the use factor base of the model yields the following equation for predicting demand of a water district. Let: SFM = number of single family households (metered); SFF = number of single family households (flat-rate); Appendix 2 146 MF = number of multifamily households; E = number of employees; P = population f = single family household (metered) use factor (g.h.d.); A = mean lot size (A = 0.241 acres) dP = mean marginal price (dP = $1.17/1000 gallons); I = median household income (I = $20.6 thousand); HH = mean household size (HH = 2.71 people) ; and D = district demand (gallons/day) Then, D = f * SFM + 630 * SFF + 217 * MF + 45 * E + 14'P(l ) Where, f = 478 + 350 (A-A) - 118.7 (dP-dP) + 3.59 (I-I) + 31.2 (HH-HH)(2) The model is applied by first calculating the SFM use factor (f) using equation (2). This adjusts the factor for the particular socio- economic characteristics of a district. Then this factor along with those for the other customer classes are applied to the demographic pro- jections of households, employment, and population to arrive at a fore- cast district demand using equation (1). These are then summed for all districts arriving at the projected water demand at a given time for the entire EIS area. HEFIN@ffiiTS The use of multiple data sources is an asset toward building the use factor model, but it also requires that some refinements be made to the initial formulation. Four are described here: Appendix 2 147 . adjustment household size and income coefficients to reflect only metered SF households rather than all households; . adjustment of all OLS coefficients to 1974 through 82 averages rather than 1962 data; . addition of an intercept to the OLS to force closure on the EIS area demand; and . adjustment of the SFM use factor for those districts without outdoor watering. The first refinement is required because data in the historical supplier level data base for household size and income reflects all households. Table 53 and 54 show calculation of the adjustment to only metered SF households. For income and household size, respectively, the ratios are 1.67 and 1.25. The second refinement is required because the OLS was performed on 1982 data while the model itself was built on 1974 through 82 averages (1977 excepted). The adjustments were based on the following assump- tions: kA afoutdoor A kdP afoutdoor 1/dP kI aftotal I kHH afindoor HH, where k = OLS coefficient. Therefore, gg kavg. = k'82 • (outdoor + Aavg. A A (outdoor A Appendix 2 148 Table 53 Household Income Adjustment for SFM Homes EDF & Entercom Metered Historical Supplier HH Data (1982) Level Data (1982) Median Income Geometric Arithmetic for Number Water Supplier Mean Median Mean all HR!/ of SRC/ Arvada 35.9 33.0 36.3 23.9 24,447 - Aurora 36.9 33.0 37.2 22.8 35,640 Bancroft-Clover 39.4 33.0 39.5 25.9 6,856 Cherry Creek Valley 46.1 437 46.8 21.5 752 Cherry Creek Village 59.8 61.8 60.5 42.4 483 Consolidated Mutual 44.9 43.7 45.2 21.3 17,171 Crestview 25.3 25.6 25.4 19.5 3,650 Denver City & County 37.7 33.8 38.0 14.1 27,987 Denver Read & Bill 47.2 43.7 47.7 30.1 24,820 Denver Total Serve 39.7 33.0 40.0 26.7 20,932 Green Mountain Park 35.7 33.0 36.1 27.3 6,154 Kelton heights 16.8 18.1 17.0 17.7 220 Northglenn 37.4 33.0 37.8 23.0 7,221 South Adams County 26.2 27.4 26.4 17.3 5,580 Westminster 30.4 25.6 30.8 21.8 13,082 (EDF & Entercom Geometric Mean) (Number SFM Households) RATIO = Districts (Supplier Level Median) (Number SFM Households) Districts = 1.67 1/ HE = household. ?/ SFM = single family metered. Note: 648 observations were used in the analysis. Appendix 2 149 Table 54 Household Size Adjustment for Single Family Metered Homes Historical Supplier Level Data (1982) EDF & Entercom Mean Data (1982) Mean HH Size Number Water District Metered HH1/ for all HEW of SFM?/ Arvada 3.42 2.97 24,447 Aurora 3.30 2.55 35,640 Bancroft-Clover 4.00 2.70 6,856 Cherry Creek Valley 3.38 2.08 752 Cherry Creek Village 3.71 3.37 483 Consolidated Mutual 3.04 2.41 17, 171 Crestview 3.17 2.58 3,650 Denver City & County 3.24 2.26 27,987 Denver Read & Bill 3.42 3.03 24,820 Denver Total Serve 3.04 2.78 20,932 Green Mountain Park 3.88 2.91 6, 154 Kelton Heights 2.83 2.19 220 Northglenn 3.83 2.90 7,221 South Adams County 4.00 2.86 5,580 Westminster 3.16 2.68 13,082 e (Mean EDF & Entercom) (Number of SFM Households) RATIO = Districts e (BBC Supplier Level) (Number of SFM Households) Districts = 1.67 1/ HH = household. 2/ SFM = single family metered. Note: 703 observations were used in the analysis. Appendix 2 150 with analogous results for the other three variables. The final ratio for marginal price is inverted. The calculations are summarized in table 55. Table 55 OLS Coefficient Adjustments Ratio of 1982 1974-82 1982 Applicable Variable k average Value Demand kavg. A 350 0.241 0.241 1.22 427 dP 118.7 1.17 1.14 1.22 141 I 3.59 20.64 22.92 1.11 3.59 HN 31.2 2.71 2.68 1.00 31.5 Note: 1974- through 1982 (1977 excepted) average water use was 478 g.h.d. with 220 estimated as indoor use. Use for 1982 was 431 g.h.d. with 220 estimated as indoor use. An intercept was added to the socioeconomic adjustment to the SFM use factor because the above modifications resulted in an error in total EIS area demand. The subtraction of 17 gallons per household per day brought the OLS equation back to zero for closure on the SEIS area demand. Finally, an adjustment was made for those districts which have zero lot size in the data base because there is no irrigation. In these cases, f was set to 230 gallons: 220 gallons for estimated indoor uses plus 10 gallons for miscellaneous outdoor uses such as car washing. r Appendix 2 151 The revised model formulation is as follows: D = f * SFM + 630 * SFF + 217 * MF + 45 * E+14 * 'P f = 478 + 427 (A - A) - 141 (dP - dP) + 3.59 (I - I)rI + 31.5 (HH - HH) ry, Where: rI = 1.67 rag = 1.25 MODEL PERFORMANCE AND LIMITATIONS MODEL RELIABILITY The motivation for the model presented in the previous section was to produce a reliable water demand forecasting model. The reliability of the refined model is shown by the following: . The model is developed from a diverse data base utilizing sound engineering judgment and is consistent with previous analyses. . The coefficient of determination (R2) is 0.998, indicating that it explains 99.8 percent of the variation in water demand from district to district (R2 = R2D + R2SE (1-R20); where R20 = 0.997, R28E = 0.271). . The model predicts the total EIS area demand within 0.04 percent of the actual demand over the 1974 through 1982 period. Appendix 2 152 . The model predicts demand for DC&C which represents over 45 per- cent of the total demand, within 6.2 percent of the actual demand. The predicted demand for the 12 largest districts, representing 86 percent of the total demand, is within 15.4 percent. The average departure from the actual district demand for all districts is 23.5 percent. Results from applying the model to historical data are shown in table 56. The number of districts with overestimated demands does not equal the number of districts with underestimated demands. This is due to the fact that three of the largest districts are underestimated and the model is designed to predict overall EIS area demand. A district by district adjustment would increase the accuracy on the historical data, but would not add to forecasting accuracy. Such an adjustment would represent the addition of unknown factors which may change an unknown extent over time. Only those factors which were well understood were incorporated into the demand model so that a high level of confidence could be placed in the models forecast. IMPLIED ELASTICITIES The elasticities of the independent socioeconomic variables [mar- ginal price (dP), income (I), lot size (A) and household size (HH)] for SFM households are defined as: x = ax x X f where, ex = elasticity for variable x; e X = the mean value of variable x; f = the SFM use factor; ax = OLS coefficient for variable x. Appendix 2 153 Table 56 Water Supplier Model Predictions --- Compared with Actual Demand Water Supplier SPM Use Factor Projected Demand Actual Demand Percent Error (g.h.d.) (g.p.d.) ii — Arvada 497 14,833,180 12,099,927 -22.6 Aurora 454 24,746,468 24,386,327 1.5 Bancroft-Clover 497 4,370,311 3,700,029 -18.1 Bow-Mar 836 257,315 243,979 -5.5 Brighton 971 2,606,988 2,325,225 -12.1 Broomfield BSA 481 1,968,950 1,801,601 -9.3 Broomfield DSA 513 1,751,629 1,418,190 -23.5 Castle Rock 542 793,294 699,285 -13.4 Cherry Creek Valley 503 978,892 1,237,316 20.0 Cherry Creek Village 659 373,405 365,947 -2.0 Con. Mutual 445 10,664,120 9,334,950 -14.2 Crestview 529 2,799,663 2,120,211 -32.0 DWD MUM 461 112,918,451 120,287,474 6.1 DWD Read 8 Bill 602 15,252,941 18,420,208 17.2 DWD Total Serve 517 12,418,738 13,355,856 7.0 East Cherry Creek Valley 486 345,101 329,177 -4.8 Edgewater 362 745,572 529,370 -40.8 Englewood PAM 430 8,028,046 7,710,179 -4.1 Erie 328 170,781 107,905 -58.3 Florence Garden 320 8,949 7,327 -22.1 Glendale 410 789,860 613,604 -28.7 Golden 490 3,019,683 2,394,348 -26.1 Green Mountain 519 3,873,580 3,137,290 -23.5 Hazeltine Height 660 46,494 32,844 -41.6 Hi-Land Acres \ 602 64,548 47,467 -36.0 Holly Mutual 1,041 27,454 26,280 -4.5 Kelton Heights 407 110,928 79,442 -39.6 Ken Caryl 482 681,004 797,720 14.6 Lakehurst 500 1,413,401 1,165,740 -21.2 Lakewood 387 490,638 436,970 -12.3 Louisville 472 1,515,852 922,117 -64.4 Meadowbrook 560 206,615 139,157 -48.5 Northglenn 289 3,292,132 4,326,960 23.9 Northaide 434 801,375 822,810 2.6 North Table Mountain 532 1,183,610 1,110,773 -6.6 North Washington 492 1,797,899 2,231,431 19.4 Panorama Park • 501 16,535 9,364 -76.6 Parker 718 220,149 176,115 -25.0 Sable 439 2,065,964 1,300,034 -58.9 Sedalia 635 44,063 24,633 -78.9 Silver Heights 743 90,398 63,390 -42.6 South Adams County 515 4,253,073 3,384,480 -25.7 Thornton 367 7,039,762 6,693,898 -5.2 Valley 373 850,192 912,496 6.8 Westminster 437 8,036,446 6,737,454 -19.3 Wheatridge 396 2,181,801 2,100,559 -3.9 Willowbrook 418 232,014 184,125 -26.0 Willows 564 1,337,143 1,231,106 -8.6 Appendix 2 154 The combined EDF and Entercom disaggregated model implies elastici- ties of: e dP = -141 1.17 = -.345 E I = 3.59 34.4 = .258 478 e A = 427 .241 = .215 and 478 c Hi = 31.5 3.39 = .223 478 The elasticities define the sensitivity of SFM water demand to the respective independent variables. Defining a value of one as moderate sensitivity, the calculated elasticities should be relatively insensi- tive to the independent variable. The implied elasticities for marginal price and income are consis- tent with literature values and in agreement with the COE Institute for Water Resources Report (1984). Literature values do not exist for lot size and household size. ADJUSTMENTS FOR CONSERVATION MEASURES IN PLACE The model is intended to forecast unrestrained water demand, that is, in the absence of water conservation measures. Three principal pro- grams were in effect over the period 1974 through 1982 (table 57). The model has been implicitly adjusted to remove the effects of the manda- tory 3 hour-3 day watering restrictions by omitting the year 1977 from the data base. Although effects of this program may have carried over to later years, 1977 was the only year omitted. Appendix 2 155 Table 57 Conservation Programs Mandatory 3-hr/3-day 3-day Voluntary Watering Watering Evapotranspiration Year Restriction Restrictions Program 1974 1975 1976 1977 X 1978 X 1979 X 1980 X 1981 X X 1982 X X The effects of the voluntary 3-day watering restrictions and the evapotranspiration (ET) program have been estimated by the DWD. Their analysis showed that the voluntary three day watering restrictions reduced peak demands, but did not affect average annual demand. There- fore, no adjustment to the model was necessary for this program. Analy- sis of the ET program in 1981 showed that, at most, the program reduced average annual demand by twelve percent. Assuming the same effective- ness in 1982, the greatest influence this might result in is an under- prediction of three percent. This is because the use factors are based on eight years of data, only two of which were affected by the ET pro- gram. Considering the uncertainties and size of this influence, it was decided that an adjustment was not necessary. LIMITATIONS The developed model is reliable and accurate in forecasting water demand for the SEIS demand area. However, several limitations should be noted: . The accuracy of the forecasting is dependent upon the accuracy of the input data. Appendix 2 156 ^ . Model application to historical data indicate that forecasts for individual water districts may err by 23.5 percent on average. District managers should consider their special circumstances for planning pur- poses. . The values used in the model are based on current conditions and should be reevaluated should significant changes occur. For example, long term changes in marginal price or the occurrence of more water sav- ing devices in households (natural retrofit) would require adjustments in the model. Appendix 2 157 CHAPTER 6 ECONOMIC AND DEMOGRAPHIC PROJECTIONS FOR THE DENVER METROPOLITAN AREA CHAPTER 6 ECONOMIC AND DEMOGRAPHIC PROJECTIONS FOR THE DENVER METROPOLITAN AREA INTRODUCTION Projections of population, employment and households are required for each water supplier in order to develop the EIS water demand fore- casts. Projections were generated for 1990, 2000, 2010 and 2035. These projections are the focus of this chapter. Forecasts developed by DRCOG are applied for 1990 and 2000. Because of the inherent uncertainties associated with long term economic and demographic forecasts, three alternative projections were developed for 2010 and 2035: Series 1, Series 2, and Series 3. This presentation Appendix 2 158 discusses the basis of the projections through 2000 and the three alternative series for 2000 through 2035. Figure 10 illustrates past and projected population trends according to these three series. DRCOG played an important role in the development of the economic and demographic projections by water supplier which drive the water demand model. DRCOG is the regional planning agency for the Denver metropolitan area which prepares forecasts utilized by Federal agencies including U.S. Department of Transportation, E.P.A. and R.U.D. Local governments and private industry also rely on DRCOG. DRCOG prepares forecasts for the Denver metropolitan area as defined by Adams, Arapahoe, Boulder, Denver, Douglas and Jefferson Counties. (The EIS demand area, however, includes all of Denver County but excludes portions of the other counties.) DRCOG prepared forecasts of population, employment and households for the Denver metropolitan area for the year 2000. These projections have been adopted by DRCOG as their "policy forecasts" which have been generally recognized as the appropriate forecasts for use in metro- politan-wide planning activities. DRCOG also developed forecasts for 1990. The EIS project team thoroughly reviewed the DRCOG policy forecasts and they were accepted as appropriate for purposes of the EIS. This evaluation of the DRCOG forecasts for 1980 to 2000 is summarized in this chapter. The three alternative projection series for the period beyond 2000 were developed with the assistance of DRCOG, but do not represent DRCOG policy forecasts. As is the nature of any long-term forecasts, the relative accuracy associated with each projection declines the greater the time frame. The premise for the development of each of these Appendix 2 159 CO, W M O4 M M _ O W • Vf O _ M O _ N O N O _ 4-4 O 04 O O O Q W C Q 2 ra J O O _ o+ C. O1 O .r H La ICI O $ .+ O C N W W O C >O _ CO C W l- C't- W 0 r-- C O ' t W M W Ct I- f C O O n O ti H I- Q J O C. O C. O C, O a O v a N C = la O M JO VI 1 1 I 1 H W O O O O Appendix 2 projection series was the assumption that the Denver metropolitan area would gradually lose its historic competitive advantage over the nation as a whole. As a result, annual in-migration to the metropolitan area would decline. The three projection series are differentiated by the rate of decline assumed for in-migration for the years beyond 2000. The following discussion presents the Denver metropolitan area forecasts and the projections by water supplier. The methodologies applied in developing the forecasts are then detailed. The final sec- tion is a summary of the evaluation conducted for each component of the forecasts. POPULATION, EMPLOYMENT AND HOUSEHOLD FORECASTS FOR THE EIS DEMAND AREA DENVER METROPOLITAN AREA FORECASTS Historical population growth in the Denver metropolitan area (Adams, Arapahoe, Boulder, Denver, Douglas and Jefferson Counties) from 1930 to 1980 with forecasts through 2035 is illustrated in Figure 10. Population is forecast by DRCOG to increase from 1.6 million residents in 1980 to 2.5 million residents in the year 2000. Series 1 forecasts population in 2035 to increase to 3.5 million while Series 3 projections show population reaching 3.9 million by 2035. These forecasts are exhibited in table 58. The DRCOG population forecasts are largely based upon employment projections. As shown in table 59, employment is forecast to increase Appendix 2 161 Table 58 Population Forecasts for the Denver Metropolitan Area 1980-20351/ Year Source Series 1 Series 2 Series 3 1950 Census 615,600 615,600 615,600 1960 Census 934,200 934,200 934,200 1970 Census 1,238,300 1 ,238,300 1,238,300 1980 Census 1,618,500 1,618,500 1,618,500 1990 DRCOG Policy 2,019,600 2,019,600 2,109,600 2000 DRCOG Policy 2,513,000 2,513,000 2,513,000 2010 Series Forecasts 2,948,000 2,963,700 2,979,400 2020 Series Forecasts 3,265,500 3,332,300 3,399,100 2030 Series Forecasts 3,444,300 3,600,800 3,757,500 2035 Series Forecasts 3,471 ,200 3,687,800 3,904,700 1/ The Denver metropolitan area encompasses the Counties of Adams, Arapahoe, Boulder, Denver, Douglas and Jefferson. This is a larger area than the EIS demand area. Source: DRCOG, Series 1 , 2 and 3 population proejctions; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. r- Appendix 2 162 Table 59 Employment Forecasts for the Denver Metropolitan Area 1980-2035 Year Source Series 1 Series 2 Series 3 1950?/ Census 235,500 235,500 235,500 _ 1960?/ Census 354,700 354,700 354,700 1970 Census 542,800 542,800 542,800 1980 Census 869,500 869,500 869,500 1990 DRC0G Policy 1 ,208,600 1,208,600 1,208,600 2000 DRC0G Policy 1 ,525,000 1,525,000 1 ,525,000 2010 Series Forecasts 1,666,600 1 ,677,300 1 ,688,100 2020 Series Forecasts 1,750,500 1 ,794,400 1 ,838,500 2030 Series Forecasts 1,752,300 1 ,852,800 1 ,953,400 2035 Series Forecasts 1,752,300 1,877,000 2,014,700 1/ The 1950 and 1960 employment data are not exactly comparable with later data. Source: DRC0G, Series 1, 2 and 3 population projections; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. Appendix 2 163 from 869,500 jobs in 1980 to 1,524,700 jobs in 2000. Series 1 forecasts show employment to stabilize at 1,752,300 in 2030. Employment under the Series 3 forecasts exceeds 2.0 million by 2035. Projections of number of households are derived from population projections. Households in the Denver metropolitan area increase from 608,400 in 1980 to 1,074,300 by 2000. Forecasts for 2035 range from 1,496,200 for Series 1 to 1 ,683,100 for Series 3. These projections are set forth in table 60. Population, employment and households projections exhibit declining growth rates from 1980 to 2035. This is consistent with a gradual loss of regional competitive advantage and convergence of Denver metropolitan area growth rates with those for the nation. As demonstrated in table 61, absolute population growth for 1980 to 2000 is expected to increase from 1970 levels. After 2000, growth steadily declines for each forecast series. Employment growth is projected to remain between 316,100 and 339,100 per decade until 2000. As shown in table 62, employment growth is forecast to markedly slow after that point in time. The annual employment growth rates exceed the population growth rates for 1980 to 2000 because of aging of the population into the prime working years. After 2000, the growth rates decline below population growth rates because of movement of the population into retirement years. The projected growth of number of households given in table 63 parallels the population forecasts. Growth rates for the 1980 to 2000 period are higher than the population growth rates as average household sizes continue to decrease. With the stabilization in household size projected after 2010, growth in the number of households mirrors pro- jected population growth. Appendix 2 164 Table 60 Households Forecasts for the Denver Metropolitan Area 1980-2035 Year Source Series 1 Series 2 AttLtEJ. 1950 Census 187,200 187,200 187,200 1960 Census 287,900 287,900 287,900 1970 Census 394,700 394,700 394,700 1980 Census 608,400 608,400 608,400 1990 DRCOG Policy 841,800 841 ,800 841,800 2000 DRCOG Policy 1,074,300 1,074,300 1 ,074,300 2010 Series Forecasts 1,270,700 1 ,277,500 1,284,200 2020 Series Forecasts 1 ,407,500 1,436,300 1 ,465,100 2030 Series Forecasts 1 ,484,600 1,552,100 1 ,619,600 2035 Series Forecasts 1,496,200 1,589,600 1 ,683,100 Source: DRCOG, Series 1 , 2 and 3 population proejctions; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. Appendix 2 165 Table 61 Population Growth Forecasts for the Denver Metropolitan Area, 1980-2035 Average Annual Absolute Growth Growth Rate All Series All Series 1950-1960 318,600 4.3% 1960-1970 304,100 2.9 1970-1980 380,200 2.7 1980-1990 401,100 2.2 1990-2000 493,400 2.2 Series 1 Series 2 Itata_2 Series 1 Series 2 §tri. 3 2000-2010 435,000 450,700 466,400 1.6% 1.7% 1.7% 2010-2020 317,500 368,600 419,700 1.0 1 .2 1 .3 2020-2030 178,800 268,500 358,400 0.5 0.8 1.0 2030-2035 26,900 87,000 1 ,472,200 0.2 0.5 0.8 Source: DRCOG, Series 1 , 2 and 3 population proejctions; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. Appendix 2 166 Table 62 Employment Growth Forecasts for the Denver Metropolitan Area, 1980-2035 Average Annual Absolute Growth Growth Rate All Series All Series 1950-1960 119,200 4.2% 1960-1970 141,500 3.4 1970-1980 326,700 4.8 1980-1990 339,100 3.3 1990-2000 316,100 2.3 Series 1 Series 2 AtElla_a Series 1 Series 2 Series 2000-2010 141 ,600 152,300 163,100 0.9% 1.0% 1 .0% 2010-2020 83,900 117,100 150,400 0.5 0.7 0.9 2020-2030 1,800 58,400 114,900 0.0 0.3 0.6 2030-2035 0 24,200 61 ,300 0.0 0.3 0.6 Source: DRCOG, Series 1, 2 and 3 population proejctions; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. Appendix 2 167 Table 63 Households Growth Forecasts for the Denver Metropolitan Area, 1980-2035 Average Annual Absolute Growth Growth Rate All Series All Series 1950-1960 100,700 4.4% 1960-1970 106,800 3.2 1970-1980 213,700 4.4 1980-1990 235,400 3.6 1990-2000 232,500 2.5 Series 1 Series 2 IgAts_a Series 1 Series 2 Series "' 2000-2010 196,400 203,200 209,900 1.7% 1.7% 1.8% 2010-2020 136,800 158,800 180,900 1.0 1.1 1 .3 2020-2030 77,100 115,800 154,500 0.5 0.8 1 .0 2030-2035 11 ,600 37,500 63,500 0.2 0.5 0.8 Source: DRCOG, Series 1 , 2 and 3 population proejctions; Denver Regional Council of Governments, "Task 2.1 Report, Population and Employment Projections," Regional Water Study, February 28, 1983; U.S. Bureau of the Census, selected years. Appendix 2 168 Based upon the forecasts for the six county area, DRCOG projected population, households and employment (by place of work) for EIS demand area water suppliers. DRCOG forecasts for 1980 and 2000 are presented in tables 64 and 65 respectively. The three series of population, households and employment forecasts were developed for each water sup- plier for 2010 and 2035. Series 1, 2 and 3 projections for 2010 are shown in tables 66, 67 and 68, respectively. Alternative projections for 2035 are presented in tables 69, 70 and 71. In sum, tables 64 through 71 present the population, households and employment forecasts which form the basis of the EIS water demand forecasts. DESCRIPTION OF THE FORECASTING METHODOLOGY DRCOG FORECASTS FOR 1980-2000 The regional forecasting techniques which produced population, household and employment forecasts for the six county Denver metro- politan area through 2000 are described in a June 1982 DRCOG publication entitled Employment and Population Policy Forecasts for the Denver Region--1980-2000. The forecasting methodology employed by DRCOG relied upon the integration of employment forecasts (labor demand) with popula- tion forecasts (labor supply). Underlying this methodology was the assumption that labor demand will call forth its own supply through inter-regional migration (DRCOG, 1982). This methodology was fully evaluated and is further discussed in this chapter. Appendix 2 169 Table 64 Population, Households and Employment by Water Supplier Districts in the EIS Demand Area, 1980 Employment By Place Name of Supplier Population Households of Work Arapahoe W&S 140 50 910 Arvada 88,630 29,440 18,220 Aurora 141,090 52,220 39,400 Beverly Hills MWA 70 20 0 Brighton 13,190 4,340 4,610 Brook Forest WD 490 170 20 Broomfield 21,230 7,030 7,550 Castle Pines MD 1/ 1/ 0 Castle Rock 4,120 1,420 2,240 Chaparral W&S 0 0 0 Charlou Park WD 360 100 0 Consolidated Mutual 61,250 23,730 27,560 Cottonwood W&S 270 80 90 Crestview W&S 15,310 5,560 6, 110 Denver SE Sub. W&S 2,700 790 360 Denver Water Department Service Area 792,130 313,870 551 ,360 Dolly-O-Denver W&S 0 0 0 Eastlake W&S 130 50 0 East Cherry Creek Valley W&S 2,270 740 80 East Valley W&S 110 30 0 Englewood 31 ,400 13,230 24,240 Erie 1,250 440 190 Evergreen MD 10,520 3,190 4,030 Florence Gardens WD 50 20 50 Forest Hills MD 230 80 0 Genesse W&S 560 230 20 Glendale 2,950 2,020 6,470 Golden 16,820 6,500 7,630 Greenwood Plaza WD 0 0 2,080 Hazeltine Heights W&S 160 50 0 Hi-Land Acres WA 300 100 0 Holly Mutual WD 80 20 0 Idledale WD 120 50 10 Inverness W&S 0 0 6,060 Lafayette 9,050 3,460 1,300 Lincoln Park West MD 0 0 0 Louisville 5,810 2,250 7,710 Louviers MS Co. 370 110 170 ' Maple Grove WA 50 10 0 Mission Viejo (Highlands Ranch) 100 30 40 Morrison 650 200 220 �., Mt. Carbon W&S 0 0 0 North Table Mountain W&S 5,800 2,000 2,770 Northglenn 29,880 9,540 6,040 Orchard Hills WD 170 50 0 Parker W&S 570 170 590 Appendix 2 170 Table 64 (Continued) Employment By Place Name of Supplier Population Households of Work Roxborough Park MD 170 50 0 Sedalia WD 160 50 240 Silver Heights W&S 270 90 0 South Adams County W&S 25,260 8,280 11,580 Stonegate MD - 0 0 0 Thornton 53,090 17,770 10,760 Thunderbird W&S 310 100 0 View Ridge WA 210 70 0 Weisner Estates WS 80 30 0 Westminster 63,860 22,150 16,740 Willows WD 11,860 3,460 930 Water Supplier Totals 1,415,350 535,430 768,380 Adams County 2,020 670 1 ,730 Arapahoe County 3,290 1 ,020 800 Boulder County 780 240 770 Douglas County 9,380 2,900 1,810 Jefferson County 5,360 2,080 9,940 Balance of Study Area Totals 2/ 20,830 6,910 15,050 Total EIS Demand Area 1 ,436,180 542,340 783,430 1/ Less than 10. Z/ Hillcrest is a mobile home park, and Hilltop is a nine-home single family development. Water demand for these suppliers is accounted for on an individual household basis with other self-supplied consumers. Note: The following abbreviations are used in this table: WO = water and sanitation; MWA = mutual water association; WD = water district; MD = metropolitan district; WA = water association. Source: Denver Regional Council of Governments, Denver Water Department, 1983. Appendix 2 171 Table 65 Year 2000 Population, Households and Employment by Water Supplier in the EIS Demand Area Employment By Place Name of Supplier Population Households of Work Arapahoe W&S 7,950 3,340 26,360 Arvada 123,770 46,050 29,340 Aurora 280,460 116,040 100,450 Beverly Hills MWA 70 20 0 Brighton 32,050 11 ,810 12,340 Brook Forest WD 490 190 20 Broomfield 45,100 16,730 32, 100 Castle Pines MD 4,000 1 ,490 600 Castle Rock 36,400 14,050 13,500 Charlou Park WD 430 130 0 Chaparral W&S 2,390 900 0 Consolidated Mutual WC (-MG) 20,960 8,350 11,180 Consolidated Mutual WC (-DWD) 55,910 23,250 24,680 Cottonwood W&S 12,080 4,010 13,020 Crestview W&S 26,400 10,740 8,040 Denver SE Sub. W&S 13,000 4,260 1,400 Denver Water Department 949,860 417,210 834,060 Dolly-O-Denver W&S 200 70 0 Eastlake W&S 130 50 0 East Cherry Creek Valley W&S 42,840 17,450 1 ,460 East Valley WAS 340 100 0 Englewood 37,590 18,790 29,320 Erie 2,730 1, 120 340 Evergreen MD 17,580 5,870 5,950 Florence Gardens WD 430 180 340 Forest Hills MD 230 80 0 Genesse W&S 1 ,900 830 800 Glendale 4,400 2,890 27,000 Golden 24,410 10,000 20,630 Greenwood Plaza WD 0 0 4,500 Hazeltine Heights W&S 180 60 0 Hi-Land Acres WA 330 120 0 Holly Mutual WD 150 40 0 . Idledale WD 190 80 20 Inverness W&S 0 0 18,270 Lafayette 21,400 9,160 3,700 Lincoln Park West MD 0 0 15,100 Louisville 17,300 7,080 23,300 Louviers MS Co. 600 200 400 Maple Grove WA - 50 10 0 Mission Viejo (Highlands Ranch) 53,600 18,010 20,600 Morrison 1,000 340 1 ,020 Mt. Carbon W&S 750 250 2,500 North Table Mountain W&S 8,310 3,210 5,410 Northglenn 31,300 11 ,250 8,500 Orchard Hills WD 240 80 0 Parker W&S 9,400 3,140 4,700 Appendix 2 172 Table 65 (Continued) Employment By Place Name of Supplier Population Households of Work Roxborough Park MD 4,500 1,480 100 Sedalia WD 600 210 0 Silver Heights WAS 270 90 0 South Adams County WAS 41,030 15,060 15,430 Stonegate MD 5,490 2,390 6,530 _ Thornton 97,640 36,540 17,720 Thunderbird WAS 310 100 0 View Ridge WA 210 70 0 Weisner Estates WS 100 40 0 Westminster 109,150 42,400 24,720 Willows WD 15.110 4,940 10,570 Water Supplier Totals 2,163,310 892,350 1 ,376,420 Adams County 5,400 2,010 2,400 Arapahoe County 8,870 3,070 1,250 Boulder County 780 270 770 Douglas County 22,530 7,800 2,750 Jefferson County 8.060 3,320 12,280 Balance of Study Area Totals 11 45,640 16,470 19,450 Total SETS Demand Area 2,208,950 908,820 1 ,395,870 11 Hillcrest is a mobile home park, and Hilltop is a nine-home single family development. Water demand for these suppliers is accounted for on an individual household basis with other self-supplied consumers. Note: The following abbreviations are used in this table: WAS = water and sanitation; MWA = mutual water association; WD = water district; MD = metropolitan district; WA = water association. Source: Denver Regional Council of Governments and Denver Water Department, 1983. Appendix 2 173 ^ Table 66 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2010, Series 1 Projections Employment by Place Water Supplier Population Households' of Work Arapahoe 7,950 3,490 31,858 Arvada 139,738 55,895 31,742 Aurora 343,209 150,108 113,638 Beverly Hills 70 20 0 Brighton 41,430 16,150 14,010 Brook Forest 490 200 20 Broomfield 53,067 21,227 37,403 Castle Pines 5,485 2,180 730 Castle Rock 51,016 20,947 15,932 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 37,653 Cottonwood 11,870 4,110 15,813 Crestview 26,194 11,210 8,457 Denver SE Sub 17,201 5,993 1,625 DWD 1,054,335 483,640 883,620 Dolly-0-Denver 265 100 0 Eastlake 130 50 0 E. Cherry Creek 50,000 21,530 1,758 East Valley 340 100 0 Englewood 41,556 19,328 30,417 Erie 3,126 1,421 372 Evergreen 20,627 7,369 6,365 Florence Gardens 607 262 403 Forest Hills 230 80 0 �.., Genesee 2,493 1,140 968 Glendale 5,025 3,442 31,435 Golden 27,662 12,087 23,438 Greenwood Plaza 0 0 5,023 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual. 150 40 0 Idledale 230 90 22 Inverness 0 0 20,908 Lafayette 25,332 11,600 4,218 Lincoln Park 0 0 18,362 Louisville 22,896 9,557 26,668 Louviers 598 210 450 Maple Grove 50 10 0 Mission Viejo (Highlands Ranch) 74,815 26,600 25,041 Morrison 1,175 427 1,193 Mt. Carbon 936 370 3,040 W. Table Mtn. 9,655 3,919 5,980 Worthglenn 29,350 11,740 9,031 Orchard Hills 240 80 0 Parker 12,293 4,286 5,588 Rozborough Park 6,009 2,168 122 Sedalia 784 292 435 Silver Heights 270 90 0 S. Adams Co. 47,105 18,842 16,262 Stonegate 7,000 3,200 7,941 Thornton 122,625 49,050 19,223 Thunderbird 310 100 0 View Ridge 210 70 0 Weisner Estates 100 40 0 Westminster 128,785 52,841 26,444 Willows 15,110 5,150 12,652 Adams Co. 20,727 7,972 3,816 Arapahoe Co. 18,481 7,108 2,666 Boulder Co. 772 297 770 Douglas Co. 43,027 15,936 4,166 Jefferson Co. 9.150 3.850 12.785 Total 2,581,320 1,112,254 1,520,463 Appendix 2 174 Table 67 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2010, Series 2 Projections Employment by Place Water Supplier Population Households of Work Arapahoe 7,950 3,490 32,273 Arvada 140,417 56,167 31,924 Aurora 345,507 151,113 114,634 Beverly Hills 70 20 0 Brighton 41,430 16,150 14,136 Brook Forest 490 200 20 Broomfield 53,067 21,227 37,804 Castle Pines 54,85 2,180 739 Castle Rook 51,016 20,947 16,116 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 37,788 Cottonwood 11,870 4,110 16,024 Crestview 26,194 11,210 8,488 Denver SE Sub 17,355 6,046 1,642 DWD 1,054,335 483,640 887,365 Dolly-O-Denver 268 101 0 Eastlake 130 50 0 E. Cherry Creek 50,000 21,530 1,781 East Valley 340 100 0 Englewood 41,691 19,391 30,500 Erie 3,148 1,431 375 Evergreen 20,747 7,412 6,396 Florence Gardens 613 264 407 Forest Hills 230 80 0 --- Genesee 2,513 1,149 981 Glendale 5,046 3,456 31,770 Golden 27,794 12,144 23,650 Greenwood Plaza 0 0 5,062 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual 150 40 0 Idledale 230 90 22 Inverness 0 0 21,107 Lafayette 25,332 11,600 4,258 Lincoln Park 0 0 18,608 Louisville 23,076 9,632 26,922 Louviers 598 210 453 Maple Grove 50 10 0 Mission Viejo (Highlands Ranch) 74,815 26,600 25,377 Morrison 1,182 430 1,206 Mt. Carbon 945 373 3,081 M. Table Mtn. 9,703 3,939 6,023 Northglenn 29,350 11,740 9,072 Orchard Hills 240 80 0 Parker 12,293 4,286 5,655 Rozborough Park 6,070 2,190 123 Sedalia 791 294 437 Silver Heights 270 90 0 S. Adams Co. 47,376 18,951 16,325 Stonegate 7,000 3,200 8,047 Thornton 122,625 49,050 19,337 Thunderbird 310 100 0 View Ridge 210 70 0 Weisner Estates 100 40 0 Westminster 128,785 52,841 26,574 Willows 15,110 5,150 12,810 Adams Co. 21,258 8,176 3,923 Arapahoe Co. 24,118 9,276 2,773 Boulder Co. 773 297 770 Douglas Co. 42,141 16,208 4,273 Jefferson Co. 9,150 3.850 12.824 Total 2,590,776 1,116,661 1,529,875 Appendix 2 175 Table 68 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2010, Series 3 Projections Employment by Place Water Supplier Population Households of Mork Arapahoe 7,350 3,490 32,692 Arvada 141,086 56,434 32,107 Aurora 347,772 152,103 115,640 Beverly Hills 70 20 0 Brighton 41,430 16,150 14,263 _ Brock Forest 490 200 20 Broomfield 53,067 21,227 38,208 Castle Pines 5,485 2,180 749 Castle Rock 51,016 20,947 16,302 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 37,925 Cottonwood 11,870 4,110 16,237 Crestview 26,194 11,210 8,520 Denver SE Sub 17,506 6,099 1,659 DWD 1,054,335 483,640 891,145 Dolly-O-Denver 271 102 0 Eastlake 130 50 0 E. Cherry Creek 50,000 21,530 1,803 East Valley 340 100 0 Englewood 41,823 19,453 30,584 Erie 3,170 1,441 377 ---- Evergreen 20,865 7,455 6,428 Florence Gardens 619 267 412 Forest Hills 230 80 0 Genoese 2,533 1,158 994 Glendale 5,067 3,471 32,108 Golden 27,923 12,201 23,865 Greenwood Plaza 0 0 5,102 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual 150 40 0 Idledale 230 90 22 Inverness 0 0 21,308 ' Lafayette 25,332 11,600 4,297 Lincoln Park 0 0 18,857 Louisville 23,253 9,706 27,179 Louviers 598 210 457 Maple Grove 50 10 0 Mission Viejo (Highlands Ranch) 74,815 26,600 25,716 Morrison 1,188 432 1,219 Mt. Carbon 955 377 3,122 M. Table Mtn. 9,751 3,958 6,067 Northglenn 29,350 11,740 9,112 ' Orchard Hills 240 80 0 Parker 12,293 4,286 5,723 Rozborough Park 6,129 2,211 125 Sedalia 797 297 440 Silver Heights 270 90 0 S. Adams Co. 47,644 19,057 16,388 Stonegate 7,000 3,200 8,155 Thornton 122,625 49,050 19,452 Thunderbird 310 100 0 View Ridge 210 70 0 Weisner Estates 100 40 0 Westminster 128,785 52,841 26,706 Willows 15,110 5,150 12,969 Adams Co. 21,780 8,377 4,031 Arapahoe Co. 24,640 9,477 2,881 Boulder Co. 775 298 770 Douglas Co. 42,838 16,476 4,381 Jefferson Co. 9.150 3.850 12.862 Total 2,596,629 1,119,071 1,539,379 176 Table 69 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2035, Series 1 Projections Employment by Place Water Supplier Population Households of Work Arapahoe 7,950 3,490 35,185 Arvada 162,256 64,902 33,196 Aurora 419,413 183,437 121,620 Beverly Hills 70 20 0 Brighton 54,786 21,915 15,020 Brook Forest 490 200 20 - Broomfield 65,688 26,275 40,613 Castle Pines 5,485 2,180 808 Castle Rook 74,427 30,559 17,404 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 38,738 Cottonwood 11,870 4,110 17,504 Crestview 26,194 11,210 8,709 Denver SE Sub 19,030 6,630 1,761 DWD 1,107,155 513,640 909,330 Dolly-0-Denver 320 120 0 Eastlake 130 50 0 B. Cherry Creek 55,039 23,700 1,939 East Valley 340 100 0 Englewood 41,220 19,490 31,082 Erie 3,802 1,755 392 Evergreen 23,491 8,393 6,616 Florence Gardens 650 280 441 Forest Hills 230 80 0 Genoese 3,170 1,450 1,070 Glendale 5,250 3,640 34,119 Golden 28,860 12,610 25,138 Greenwood Plaza 0 0 5,339 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual 150 40 0 Idledale 230 90 23 Inverness 0 0 22,504 Lafayette 31,780 14,552 4,532 Linooln Park 0 0 20,336 Louisville 28,857 12,045 28,706 Louviers 598 210 480 Maple Grove 50 10 0 Mission Viejo (Highlands Ranoh) 84,096 29,900 27,729 Morrison 1,210 440 1,297 Mt. Carbon 1,190 470 3,367 N. Table Mtn. 11,264 4,573 6,325 Northglenn 29,350 11,740 9,353 Orchard Hills 240 80 0 Parker 12,293 4,286 6,125 Rozborough Park 7,040 2,540 135 Sedalia 860 320 455 Silver Heights 270 90 0 S. Adams Co. 49,000 19,600 16,765 Stonegate 7,000 3,200 8,794 Thornton 159,579 63,832 20,133 Thunderbird 310 100 0 View Ridge 210 70 0 Weisner Estates 100 40 0 Westminster 154,843 63,533 27,487 Willows 15,110 5,150 13,913 Adams Co. 58,205 23,757 5,530 Arapahoe Co. 50,484 20,638 4,380 Boulder Co. 822 316 770 Douglas Co. 89,320 35,728 5,880 Jefferson Co. 9,506 4,000 13,091 Total 3,000,302 177 1,295,826 1,594,154 Table 70 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2035, Series 2 Projections Employment by Place Water Supplier Population Households of Work Arapahoe 7,950 3,490 40,027 Arvada 171,583 68,633 35,311 Aurora 450,976 197,241 133,234 Beverly Hills 70 20 0 Brighton 60,321 24,128 16,491 Brook Forest 490 200 20 Broomfield 70,521 28,208 45,283 Castle Pines 5,485 2,180 922 Castle Rock 83,418 34,251 19,547 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 40,317 Cottonwood 11,870 4,110 19,963 Crestview 26,194 11,210 9,076 Denver SE Sub 19,030 6,630 1,958 DWD 1,107,155 513,640 947,275 Dolly-0-Denver 320 120 0 Eastlake 130 50 0 E. Cherry Creek 55,039 23,700 2,201 East Valley 340 100 0 Englewood 41,220 19,490 32,048 Erie 4,102 1,893 421 Evergreen 23,491 8,393 6,981 Florence Gardens 650 280 496 �-. Forest Hills 230 80 0 Genesee 3,451 1,578 1,219 Glendale 5,250 3,640 38,025 Golden 28,860 12,610 27,611 Greenwood Plaza 0 0 5,800 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual 150 40 0 Idledale 230 90 ' 25 Inverness 0 0 24,827 Lafayette 34,256 15,686 4,989 Lincoln Park 0 0 23,209 Louisville 31,325 13,075 31,672 Louviers 598 210 524 Maple Grove 50 10 0 Mission Viejo 84,096 29,900 31,641 Morrison 1,210 440 1,450 Mt. Carbon 1,190 470 3,842 W. Table Mtn. 11,931 4,843 6,828 Horthglenn 29,350 11,740 9,821 Orchard Hills 240 80 0 Parker 12,293 4,286 6,907 Rozborough Park 7,040 2,540 154 Sedalia 860 320 486 Silver Heights 270 90 0 S. Adams Co. 49,000 19,600 17,497 Stonegate 7,000 3,200 10,037 Thornton 173,771 69,508 21,457 Thunderbird 310 100 0 View Ridge 210 70 0 Weisner Estates 100 40 0 Westminster 164,851 67,639 29,005 Billows 15,110 5,150 15,747 • Adams Co. 81,203 33,144 7,917 Arapahoe Co. 83,898 34,244 6,767 Boulder Co. 843 324 770 Douglas Co. 114,350 48,880 8,267 Jefferson Co. 9.506 4,000 13,537 Total 3,172,406 1,369,834 1,701,602 178 Table 71 Population, Households and Employment by Water Supplier in the EIS Demand Area, 2035, Series 3 Projections Employment by Place Water Supplier Population Households of Work Arapahoe 7,950 3,490 45,373 Arvada 180,920 72,368 37,647 Aurora 482,573 211,061 146,058 Beverly Hills 70 20 0 Brighton 65,868 26,347 18,115 Brook Forest 490 200 20 - Broomfield 70,521 28,208 50,440 Castle Pines 5,485 2,180 1,048 Castle Rock 90,437 37,133 21,912 Chaparral 2,390 940 0 Charlou Park 430 140 0 Consolidated 75,689 32,970 42,061 Cottonwood 11,870 4,110 22,680 Crestview 26,194 11,210 9,482 . Denver SE Sub 19,030 6,630 2,177 DWD 1,107,155 513,640 989,125 Dolly-0-Denver 320 120 0 Eastlake 130 50 0 E. Cherry Creek 55,039 23,700 2,491 East Valley 340 100 0 Englewood 41,220 19,490 33,115 Erie 4,402 2,032 452 Evergreen 23,491 8,393 7,384 Florence Gardena 650 280 557 "-... Forest Hills 230 80 0 Genesee 3,630 1,660 1,383 Glendale 5,250 3,640 42,337 Golden 28,860 12,610 30,342 Greenwood Plaza 0 0 6,308 Hazeltine Heights 180 60 0 Hi-Land Acres 330 130 0 Holly Mutual 150 40 0 Idledale 230 90 27 Inverness 0 0 27,392 Lafayette 36,738 16,823 5,493 Lincoln Park 0 0 26,381 Louisville 33,797 14,107 34,94 Louviers 598 210 57 Maple Grove 50 10 Mission Viejo (Highlands Ranch) 84,096 29,900 35,96 Morrison 1,210 440 1,61 Mt. Carbon 1,190 470 4,36 N. Table Mtn. 12,598 5,114 7,38 Northglenn 29,350 11,740 10,33 Orchard Hills 240 80 Parker 12,293 4,286 7,77 Rozborough Park 7,040 2,540 17 Sedalia 86o 320 52 Silver Heights 270 90 S. Adams Co. 49,000 19,600 18,30 Stonegate 7,000 3,200 11,40 Thornton 187,995 75,198 22,92 Thunderbird 310 100 View Ridge 210 70 Weisner Estates 100 40 Westminster 170,605 70,000 30,68 Willows 15,110 5,150 17,77 ---.. Adams Co. 98,708 40,289 10,56 Arapahoe Co. 101,403 41,389 9,41 Boulder Co. 864 332 77 Douglas Co. 152,446 66,578 10,91 Jefferson Co. 9.506 4.000 14.02 Total 3,325,111 179 1,435,198 1,820,225 r EMPLOYMENT FORECASTS Employment projections by sector for the Denver metropolitan area were formulated by applying a top-down approach. National employment forecasts formed the basis for State of Colorado forecasts, which in turn formed the basis for Denver metropolitan area forecasts. National employment forecasts were projected for economic sectors by Chase Econometrics through a comprehensive series of 150 econometric equations. The number of new jobs was defined or constrained by national population growth and labor force participation rates. The forecasts utilized were last updated in July 1981. The Chase Econometric forecasts are well-respected and widely used throughout the United States. The Governor's Blue Ribbon Panel projected Colorado employment based upon the Chase Econometric national forecasts. Colorado's share of national employment in each sector was estimated from regression models which relate past Colorado employment to historical U.S. employ- ment for each sector. Employment for some Colorado sectors, such as mining, are independently forecast based upon the knowledge of local industries. In the third stage, DRCOG used regression equations to relate Denver employment to Colorado employment. The regression statistics from this analysis are provided in table 72. This regression analysis was based on 1970 to 1980 data and since the Denver metropolitan area has accounted for such a large share of Colorado employment, the tech- nique produced high correlation coefficients. Taking Denver metropolitan area mining employment as an example, the level of Colorado mining activity is one influence which explains Appendix 2 180 0 O C .-1 0 a CO •- N 0 M S 0 7 0 ' M 0^ in a0 .o O 0` CO 0) 0 S 0 11110 40 .0 .0 1!1 %a N O U\ 0. L 4. 0 0 O O 0 0 0 0 0 • O 4- 0 .... OD 4-4 0 0) I . cc o 0 O O N 1 O 0' 0. 0 C 0 O v ,-1 120 i.) m Vp al aU) al 1 C. a 4 C a N C 0 0 in Opp ) 4 .el a 0 +i C .0 C a 07 C Cs. O .2i O i.) a'.1 a i-0 N U 3 a1 0 4.3 .. U yE B> g. 0p 0L O U U 0) 0 0 a /0 6 O .i C 0 0) 0 N C !0 0 el 0. L 60. C 4013 vOi Dal -.A 0 4 O. L CO MI 0. W .0 "I 1 8 a) 7 00 > C a) .i 0 6 a7 7 >, 0 0 C .C C 0 C C C .0 Co al ,i ,C0 > 0 V a. O O) .i a 0) 0 C >. 0 X � [r] b] F -.-I 3 C07 D 0 0 O. C e 0 0 0 0 0 0 0 O .7 0 0 D 0 0 O O M -.1 >3 0 >. 'C V .i 'C V *Ord V 'C V V 'O 1 L V C 0 H Cd N ca td al a) 0 N Ia U of al al C. 00 >. al 0 el L L 0 4 L C. a L L +l L 0 L L 0 C L C. .i 0, 0 0 0 0 0 0 0 O .i O V 0 0 > 1 RI0 .0 8 .i .i 'O Mel .i .i 'O el .i .0 .-I O .-1 .i C C .i .i a) W 00C C -HO 0 OC 0 00 OF. 0 0 0 0 03 0 .i Coed FU U OM U U0., 06. oon 2 V) 0 0 N8 0 t-- U D a 0 0 .i U a Cd a° en U M in 1011` N N S 0 N a. 0 L 10 .0 CV M N O N L[1 N ri i01 I O N N m I a0.. 4.3 G .01 I-I N a) 9 0 C a) 4 V) al O vv) al a as co P 00' 0• T 00`. 00`. 01 n. Cr, CO ON ON 0 d L el b0 el 0 a) 0 CC C 0 0 0 C U '0 .i 0 m v-i L CO .l aa ' C C .0 F n-I O 000iO .01 of °' ''11 a CO 0 0 �" U U V +> 0 9 L iL1 _I C td a O 0 S B C m 00 0 0 C '0 C. 0 0 0 7 • 0 LC N .0 al aN 00 0 i.043 C 0 C 0 i0. a1 a) a -W 0 4 a .-I C CC)) yg, O C m U 0 0 0) 0 M N g C 0 C 0 C al la 0 .i U Cu.i V) C 0 .i N v L W CM M 0 .l 0 L 'O 0. U +i N 7 0 0 > C a .i 41 .i 8 4 "� C C C C a L ai 7 0 > m 0) 0 ,i 0 O .l 0 L .C 0 0 C 0 E U E CU V) F 3 0 a D V) Appendix 2 181 local area mining employment. These two variables together accounted for 99.9 percent of the historical variation in Denver metropolitan area miing employment. Colorado nonagricultural wage and salary employment in part explains the level of mining employment in the Denver region as an administrative center for the mining industry. DRCOG has developed several alternative forecasts for each employment sector based upon different Governor's Blue Ribbon Panel employment assumptions reflecting Colorado's regional competitive advantage and oil shale development. POPULATION FORECASTS Population forecasts begin with a projection of labor supply based upon natural population growth calculated by a cohort-survival model. Migration in or out of the metropolitan area is related to the differ- ence between labor supply and projected jobs: . The base year population, in five year age groups, is multiplied by the assumed death rates for each age group. Those deaths which would occur during year are then subtracted from the initial population. . The number of women of childbearing age (the 15-44 age groups) is multiplied by a birth rate forecast for each age group to produce the expected number of births during the time interval. These births are then added to the surviving population to obtain the total population at the end of the year. This new population becomes the base population for the next time interval. . The labor force available under the cohort-survival population forecasts constitute the natural labor supply in the absence of migra- tion. Appendix 2 182 . Labor force demand is projected from the employment forecasts described previously. Unemployed persons in the labor force are also considered. The difference between labor demand and labor supply rep- resents jobs to be filled by persons migrating into the region. . Indicated migration into the region is calculated by adding estimates of migrating workers and their nonworking family members and other persons not in the labor force migrating into the region. . The in-migrants are then added to the population base. Births and deaths for this group are calculated in subsequent years. This pro- cedure is reiterated for each year in the forecast period. . A constraint on the DRCOG population forecasting approach is the adjustment of net in-migration. DRCOG and an outside panel of economic experts, the Economic Planning Task Force, concluded that the comparison of labor demand to labor supply created an excessive level of future in- migration resulting in excessive employment and population forecasts, and a limitation or cap was established on in-migration population of 30,000 per year. The Economic Planning Task Force, which met several times in early 1982, consisted of several knowledgeable individuals from universities, banks, a private firm and state government. HOUSEHOLD FORECASTS Household forecasts for the Denver metropolitan area were based upon the population forecasts by age and sex. The forecasts are detailed in the July 1982 DRCOG publication entitled Household and Household Size Forecasts for the Denver Region--1980-2000. Rates of household formation were estimated for each age group for men and for women. Household forecasts are then derived by multiplying household formation rates by population in each age group. The household size Appendix 2 183 estimates follow the same trend as the D.S. Census Bureau's Series D projections. Population in group quarters such as dormitories, repre- senting about two percent of the total population, are excluded from the household size assumptions for both the national and local household forecasts. WATER SUPPLIER ALLOCATIONS DRCOG allocated population, households and employment to each water supplier in the EIS demand area. The initial step was the determination of water supplier service area boundaries, primarily based upon the information received from the DRCOG Regional Water Study questionnaire. The latter survey, conducted by DRCOG in coordination with the EIS research team, was completed in April 1983. The 1980 and 2000 forecasts of population, households, and employ- ment were developed through DRCOG's small area allocation procedures. DRCOG utilized two small area measures to allocate the economic and demographic variables: urban service areas and traffic analysis zones. Urban service areas are defined as areas where residential or economic development will take place and urban services such as water will be provided. The urban service area boundaries are determined in conjunc- tion with local governments and generally coincide with jurisdictional boundaries. The allocation of Denver metropolitan area forecasts to urban service area is accomplished through interaction with local pro- jections. The smaller traffic analysis zones take into account road transportation plans which reflect specific areas where future develop- ment might occur. The allocation process or splitting of these small areas between water districts is based upon geographic distributions adjusted to reflect concentrations of residential and employment activity. Aerial Appendix 2 184 photographs and a knowledge of the area were applied in this refinement process. Where projected water provider boundaries overlap or were inconsistent, interviews were conducted with local officials and modi- fications were made. The 1990 projections are a simple interpolation between the 1980 and 2000 forecasts. FORECASTING METHODOLOGY FOR 2000-2035 PROJECTIONS BACKGROUND In support of this study, DRCOG originally developed population, households and employment forecasts for the Denver metropolitan area and for water suppliers for 2010. Based upon the 2010 forecast, DRCOG then generated three alternative population projections for 2035 assuming different rates of decline in annual migration. The original 2010 and 2035 forecasts were evaluated in light of widely held views of the Denver metropolitan area's future growth pros- pects. Eventual convergence of regional growth trends with national growth trends is generally accepted by forecasters as the basis for developing Colorado and Denver metropolitan area projections. DRCOG in Employment/Population Policy Forecasts 1980-2000 states that growth in manufacturing, a key basic sector, will approach that of the nation. The Governor's Blue Ribbon Panel forecasts assume that by the late 1980's, both population and jobs growth in Colorado will slow. Accord- ing to these studies, the future will bring a weakening in the Colorado competitive advantage due to higher relative wages, rising land costs and possibly deterioration of the quality of life. The U.S. Bureau of Economic Analysis (BEA) states that eventually state economies will' tend over the long run to reduce disparities in growth rates. The BEA OBERS forecasts for the Denver metropolitan area are based upon this premise. Appendix 2 185 Examination of the original 2010 Denver metropolitan area forecasts found that, by several measures, the 2000 to 2010 growth patterns did not reflect the expected trends toward convergence with the nation. The growth projections assumed net migration to the Denver Region of about 30,000 per year for 2000 to 2010. This is the same level as projected for the 1990 to 2000 time period as part of the DRCOG Policy Forecasts. The net migration of 30,000 per year was higher than average migration for any decade since 1940. Anticipated migration appeared excessive in light of the declining pool of persons in the U.S. most likely to migrate (persons ages 20 to 34). Unlike the year 2000 projections, the 2010 projections had not been adopted by DRCOG as policy forecasts and were developed solely for the purposes of the EIS and the DRCOG Regional Water Study. It was determined that these projections did not follow the generally accepted concept of convergence with national trends and that they should not be applied in the EIS. Thus, the original projections beyond year 2000 were rejected for the EIS. POPULATION FORECASTS Modified projections for 2010 and 2035 were developed with the assistance of DRCOG by altering one demographic assumption: the year at which the migration decline commences. As in the previous forecasts, DRCOG applied a cohort survival model to generate population projections for males and females by age group. The forecasts applied projections of fertility rates and mortality rates previously generated by DRCOG. Modified projections for 2000 to 2035 were developed by DRCOG. The new forecasts are: . Series 1, net annual migration declining from 30,000 in 2000 to 0 by 2035. (Straightline interpolation assumed for interim years.) r Appendix 2 186 . Series 2, net annual migration declining from 30,000 in 2000 to 10,000 by 2035. . Series 3, net annual migration declining from 30,000 in 2000 to 20,000 by 2035. EMPLOYMENT AHD HOUSEHOLDS FORECASTS Employment and households projections were developed from the modified 2000 to 2035 population projections generated by DRCOG. The approaches were designed to be consistent with the methods applied to the forecasts through 2000. Employment. Denver metropolitan area employment was forecast by first projecting civilian labor force and then adjusting for military employment, self-employment and unemployment. The U.S. Bureau of Labor Statistics has projected labor force participation rates for men and women for each age group for the United States which were applied to generate civilian labor force for each series. The rates by cohort were held constant; any differences in overall labor force participation rates among forecasts over time result because of differing age distributions. The civilian labor force was adjusted to approximate total employment by applying the adjustment factors utilized by DRCOG for 2000 as reported in DRCOG, Hnployment/Population Policy Forecasts 1980-2000. Households. Household forecasts for the region were developed utilizing the household size projections developed for 2010 by DRCOG, Households and Household Size Forecasts for the Denver Region 1980-2000 and Task 2.1 Report: Population and Employment Projections. The DRCOG forecast of 2.32 total persons per household is utilized for 2010. Appendix 2 187 r The DRCOG 2010 forecast of household size is also applied for 2035. Although an aging population tends to reduce average household size and increase the number of households, it is not evident that this trend would continue from 2010 to 2035. No national projections of household formation rates were identified for this period. For these reasons, average household size for the region was assumed to remain stable from 2010 to 2035. - WATER SUPPLIER FORECASTS Year 2010 and 2035 population, households and employment forecasts were developed for each water supplier area within the EIS demand area for the Series 1, 2 and 3 projections. The water supply area projec- tions were developed in light of the urban growth and development pro- cesses at work in the Denver Region. In the next fifty years, develop- able land will become more scarce in the central areas of the metropoli- tan area. At some point in the future, growth in these areas will result from redevelopment of existing lower density residential and commercial districts. For example, DRCOG foresees very limited growth beyond 2010 for Denver and Jefferson Counties. Several suburban water supply areas are also expected to be fully developed by the year 2000 or 2010. These water districts include Eastlake, Orchard Hills and Weisner Estates. In some cases, buildout has already occured. Beverly Hills, Forest Hills, Maple Grove, Silver Heights, Thunderbird and View Ridge water districts are examples of these areas. Development potential for other parts of the Denver metropolitan area are high, however. DRCOG estimates that the corridor from Cherry Creek Reservoir to Franktown in Douglas County could develop into a residential community of 700,000 persons. The area to the north and Appendix 2 188 east of Thornton has a capacity of 260,000 residents (Henningson, Durham and Richardson, Inc., 1984). The land around Brighton has similar development potential. Aurora is also likely to continue to grow to the east. Broomfield, Louisville and Lafayette are additional growth centers. It is expected that areas within the six county region but outside the EIS demand area will capture a larger share of Denver metropolitan area growth beyond the year 2000. Longmont and areas south of Castle Rock are two such areas with high growth potential. The City of Boulder, the largest population center outside the EIS demand area, has a policy of limiting growth, however. Methodology. DRCOG has developed population, households and employment projections for each water supply area for 1980, 2000 and 2010. The 1980 and 2000 projections are adopted for use in the System- wide EIS. These forecasts with the 2010 projections also form the basis for the water supply area forecasts applied beyond 2000. Other key inputs into the forecasting process are the projections reported by individual water districts in the 1982 DRCOG Regional Water Study Questionnaire. Preliminary results from the 1984 Denver Water Depart- ment Land Use Survey are applied. Finally, DRCOG assisted in reviewing the 2010 and 2035 projections. Unconstrained and constrained household estimates were generated. Unconstrained household growth represents household growth in the absence of constraints such as land availability or local ability to provide services. Unconstrained projections were calculated by assuming each district captured the same share of regional 2000 to 2010 growth and 2010 to 2035 growth as DRCOG projected for the 1980 to 2000 period. Appendix 2 189 Based upon a review of the projections with DRCOG, the growth rates were increased for the following water supply areas: Brighton, Castle Pines, Castle Rock, Cottonwood, Parker, Thornton and the unincorporated areas of Adams, Arapahoe and Douglas Counties. Maximum household potentials were also generated for each water supply area for 2010 and 2035. These estimates represent reasonable development capacity for that point in time. DRCOG's original projec- tions of households in 2010 formed one estimate of development capacity. Projections reported in the DRCOG Regional Water Study questionnaire were also examined. The lower of these two forecasts was assumed to be the development potential in 2010. Long term projections were available from the DRCOG questionnaire for the following areas: r . Broomfield, Castle Pines, Castle Rock, Consolidated Mutual, Crestview, Denver Water Department (from Land Use Survey) , East Cherry Creek Valley, Englewood, Erie, Evergreen, Highlands Ranch, Lafayette, Lincoln Park, Louviers, North Table Mountain, Northglenn, Parker, Thornton (HDR report), and Westminster. These forecasts were applied as the 2035 maximum household potentials. For several other districts DRCOG has projected buildout in the following districts. They include: . Arapahoe, Beverly Hills, Brook Forest, Chaparral, Charlou Park, Cottonwood, East Valley, Eastlake, Forest Hills, Hazeltine Heights, Hi- Land Acres, Holly Mutual, Maple Grove, Orchard Hills, Silver Heights, Thunderbird, View Ridge, Weisner Estates, and Willows. i-. Appendix 2 190 The original DRCOG 2010 household forecasts were assumed to repre- sent maximum 2035 development for the following districts: . Denver Southeast Suburban, Glendale, Golden, Greenwood Plaza, Mt. Carbon, North Table Mountain, Roxborough Park, and Sedalia. Based upon available information and DRCOG review, the following districts were forecast to grow at 1980-2000 rates without any cap: . Arvada, Aurora, Brighton, Louisville, and South Adams County. The population per household ratios developed by DRCOG for 2010 were utilized to calculate 2010 and 2035 population for each water sup- ply area for Series 1, 2 and 3. Slight decreases in household sizes were applied for water suppliers in the northern portions of the Denver Region and for the areas outside any water district. These modifica- tions were necessary to maintain the overall population to households ratio of 2.32 in 2010 and 2035. The 1980 to 2000 share of employment growth projected for each water supply area by DRCOG was applied for the 2000 to 2010 and 2010 to 2035 time periods. Slight decreases in the DWD growth rate and increases in the balance of Adams, Arapahoe and Douglas Counties growth rates were the only modifications made. Appendix 2 191 EVALUATION OF THE POPULATION, EMPLOYMENT AND HOUSEHOLDS FORECAST COMPONENTS This section of the chapter examines each important underlying com- ponent of the projections. Population, employment and households fore- casts are discussed in turn. POPULATION FORECASTS Three key components of the 1980 to 2035 population projections are examined in this section: net migration, fertility rates and mortality rates. An overview of Denver's projected population growth relative to the nation initiates the discussion. Projected age distributions for the Denver metropolitan area are also presented. OVERVIEW Each of the three population projection series assume convergence of the Denver metropolitan area growth trends with national trends. Absolute population growth and population growth rates for the Denver metropolitan area are forecast to decline from historic levels under each projection series. As shown in table 73, the Denver metropolitan area has been steadily increasing in share of U.S. population. Appendix 2 192 Table 73 Denver Metropolitan Area as a Share of U.S. Population Year Series 1 Series 2 Series 1950 0.41 0.41 0.41 1960 0.52 0.52 0.52 1970 0.61 0.61 0.61 1980 0.71 0.71 0.71 • 1990 0.81 0.81 0.81 2000 0.94 0.94 0.94 2010 1.04 1.05 1.05 2020 1.10 1.12 1.15 2030 1.13 1.18 1.23 2035 1.13 1.20 1.27 Source: U.S. Bureau of the Census, Projections of the Population of the U.S.: 1983-2050. Under Series 1, the Denver metropolitan area's share stabilizes at 1.13 _ percent of the U.S. population. At this point, there would be an approximate equilibrium of Denver's relative attractiveness relative to the nation. Under the Series 2 projections, Denver metropolitan area's share of U.S. population increases to 1.20 percent. While the rate of increase in share slows, it does not stabilize by 2035. Denver's share of national population increases to 1.27 percent under Series 3. No equilibrium is evident for Series 2 and 3. Another measure of convergence with national growth trends is to express projected absolute population growth by decade for the metro- politan area as a share of national population growth. As shown in table 74, the Denver metropolitan area captured 1.68 percent of national growth in the 1970's. The 1980 to 2000 projections reflect significant increases in Denver's share of national population growth. Appendix 2 193 Table 74 Denver Metropolitan Area Growth as a Share of U.S. Population Growth 1950-2035 Denver Metropolitan Area Share of U.S. Population Growth Period Series 1 Series 2 Series 3 1950-1960 1.13% 1.13% 1.13% 1960-1970 1.24 1.24 1.24 1970-1980 1.68 1.68 1.68 1980-1990 1.72 1.72 1.72 1990-2000 2.70 2.70 2.70 2000-2010 2.87 2.97 3.08 2010-2020 2.41 2.79 3.18 2020-2030 2.24 3.36 4.49 2030-2035 1.21 3.90 6.60 Source: U.S. Bureau of the Census, Projections of the Population of the U.S.: 1982-2050. Denver's share of population growth peaks in the 2000 to 2010 period for Series 1. The growth share under this series declines 1.21 percent for 2030 to 2035. Series 2 and Series 3 projections exhibit increasing shares of national population growth from 2000 to 2035. In the 2030 to 2035 time period, the Denver metropolitan area would capture 6.6 percent of U.S. population growth during this period under Series 3. This brings into question the consistency of the Series 3 projections with the common assumption of convergence. MIGRATION Projected Migration Levels. Over one-half of Denver's population growth since 1940 is attributed to in-migration. The forecasts are very sensitive to the migration rates. For example, the DRCOG net migration cap of 30,000 persons per year begins to constrain growth by 1990. Appendix 2 194 Without the migration cap, projected in-migration would average 42,000 persons per year during the 1990's. By year 2000, the migration cap causes a five percent reduction in the Denver region's population and a seven percent reduction in employment. Table 75 presents past net migration levels into the Denver metro- politan area. Peak net in-migration was experienced from 1971 to 1974, wiht net in-migration exceeding 51,000 persons in 1972 to 1973. Net migration averaged 21, 100 for the last two years for which data are available. Table 75 Net In-Migration Rates Into the Denver Metropolitan Area, 1940-1982 Annual Period Migration 1940-50 10,600 1950-60 18,800 1960-70 16,300 1970-80 25,200 1970-71 23,500 1971-72 36,700 1972-73 51,200 1973-74 40,400 1974-75 24,900 1975-76 18,200 1976-77 20,600 1977-78 14,000 1978-79 12,700 1979-80 9,600 1980-81 27,400 1981-82 14,800 Source: Denver Regional Council of Governments, Employment and Population Estimates for the Denver Region, 1980-2000, p. C-1, January 1982. Appendix 2 195 Projected annual net migration is presented in table 76 for each pro- jections series. DRCOG 1980 to 1990 forecasts shows net migration averaging 21,900 per year, roughly on par with the most recent experi- ence. DRCOG projects annual net migration to increase to 29,800 for 1990 to 2000. The DRCOG cap of 30,000 person net migration per year limits the migration during this period. For the Series 1 projectins, it is assumed that net migration will decrease from 30,000 for 2000 to zero by 2035. In other words, the number of persons moving to Denver metropolitan area in 2035 will equal the number of individuals moving out of the metropolitan area. Table 76 Average Annual Net Migration to the Denver Metropolitan Area, 1940-2035 Period Series 1 Series 2 Series 1980-1990 21,900 21,900 21,900 1990-2000 29,800 29,800 29,800 2000-2010 26,000 27,400 28,700 2010-2020 17,400 21,600 25,700 2020-2030 8,900 15,900 22,900 2030-2035 2,600 11,700 20,800 Source: DRCOG, Population/Employment Policy Forecasts, 1980-2000. The assumption of 10,000 persons annual net migration to Denver in 2035 is made for Series 2, while 20,000 persons net migration in 2035 is implicit in the Series 3 forecasts. A key method which DRCOG has applied to evaluate migration projec- tions is to calculate the Denver metropolitan area's "capture" of the Appendix 2 196 nation's pouplation most likely to migrate. The pool of most likely migrants consists of persons ages 20 to 34. From 1970 to 1980, Denver captured 2.9 net migrants between the ages of 20 and 34 for every 10,000 person in this age group in the nation. Table 77 shows the historic and projected capture ratios. Table 77 Net Migrants to the Denver Metropolitan Area Per 10,000 U.S. Population, Ages 20-34, 1940-2035 Period Series 1 Series 2 Series 1940-1950 1.9 1.9 1 .9 1950-1960 3.2 3.2 3.2 1960-1970 2.6 2.6 2.6 1970-1980 2.9 2.9 2.9 1980-1990 2.1 2.1 2.1 1990-2000 3.1 3.1 3.1 2000-2010 2.9 3.0 3.2 2010-2020 1.7 2.1 2.6 2020-2030 0.9 1.7 2.4 2030-2035 0.2 1.2 2.2 Source: DRCOG, Population/Employment Policy Forecasts 1980- 2000; U.S. Bureau of the Census, Projections of the Population of the U.S.: 1982 to 2050. Under Series 1 projections, the capture rate declines to 2.9 for 2000 to 2010 reaching 0.2 for the 2030-2035 period. Relative net migra- tion also declines after 2000 for Series 2. Under Series 3, however, the capture rate increases to 3.2 net migrants per 10,000 population. This is equivalent to the highest historical capture rate experienced for the Denver metropolitan area and appears to be in conflict with the concept of convergence. Appendix 2 197 Caparison with Recent Trends. Long-term migration assumptions can be cautiously compared with recent short-term trends. Population fore- casts and recent estimates for 1983 are presented in table 78. Table 78 Comparison of DRCOG Forecasts and Estimates of 1983 Population for the Denver Metropolitan Area 1983 Population DRCOG Long-Term Forecast i/ 1,703,600 1983 Estimate 1,717,850 Difference (14,250) Percent Difference (0.8)% 1/ Population forecast interpolated from 1980 Census figures and 1985 DRCOG forecasts. Assumes constant natural population increase during this period and net migration as reported in DRCOG, Employment/Popu- lation Policy Forecasts 1980-2000, 1982. Source: DRCOG, Employment/Population Policy Forecasts 1980- 2000, 1982. DRCOG has estimated Denver metropolitan area population for January 1983, based upon the methodology explained in DRCOG, County Population and Household Estimates, 1980-82, July 1982. DRCOG 1985 population forecasts were interpolated to January 1983. The population forecast underpredicts 1983 population by 14,250 people or 0.8 percent. The DRCOG forecast of 1980 to 1983 natural population growth, about 18,000 persons per year, is similar to estimated births and deaths over this period. Net migration was underpredicted, however. Estimated net migration was 27,400 persons for 1980 to 1981 and 14,800 persons from 1981 to 1982. Migration from 1982 to 1983 is not available. DRCOG forecast net migration to average 16,700 persons per year from 1980 to 1983. Appendix 2 198 FERTILITY RATES DRCOG used the total fertility rate; births per woman in seven dif- ferent age categories between the ages of 15 and 49 are summed to obtain this measure. DRCOG assumes that the fertility rate will increase to 2.1 by the year 2000. This rate is uniformly applied for each projec- tion series for 2000 to 2035. A fertility rate of 2.1 children per women is known as the replacement rate. Fertility rates in the Denver metropolitan area fell from 2.15 in 1970 to 1.67 in 1980. This decline is attributed to several factors: urbanization of counties surrounding the core city, unpromising economic conditions or prospects, increased participation of women in the labor force, and delayed marriage of women in child bearing ages. U.S. Census Bureau projections assume fertility rates for the nation will increase from 1 .83 births per woman in 1980 to 1.96 in 2000. Fertility rates are then projected to decrease to 1.9 by 2050. Chase Econometrics also assumes that fertility rates will stabilize at 1.9 by the year 2000. The Colorado Division of Local Governments forecasts project a 1.8 fertility rate in 1998. National and local forecasts are consistent with DRCOG in assuming an increase in fertility rates from the present levels. DRCOG projected higher fertility rates than other forecasts, reasoning that economic growth and increasing incomes in the Denver metropolitan area would lead to increasing expectations of those in the child bearing years about their future prosperity and quality of life. Because birth rates have been demonstrated to vary directly with adult's perceptions of the future, birth rates in the Denver metropolitan area are forecast to increase. However, even if the Denver metropolitan area fertility rate stabilized at the U.S. Bureau of the Census projection of 1.96 rather Appendix 2 199 than the higher rate of 2.1, DRCOG population forecasts for the year 2000 would decrease by only one percent. While the DRCOG rates are not entirely in concurrence with other projections, the reasoning behind the projections appears reasonable and the forecasts are relatively insensitive to these differences. MORTALITY RATES DRCOG projections of mortality rates are based upon Census projec- tions for males and females by age group. Differences in mortality rates among various state and national forecasts are minimal. Large changes in mortality rates are necessary to significantly alter fore- casts. AGE DISTRIBUTION Differences in the age distributions account for many of the dif- ferences among DRCOG demographic assumptions and Bureau of the Census or other assumptions for the nation. The projected age distributions for the Denver metropolitan area are presented here to provide the necessary information to evaluate other forecasting assumptions. Table 79 compares the 1980 and projected year 2000 age distribu- tions for the Denver metropolitan area and the United States. Appendix 2 200 Table 79 Age Distribution of Denver Metropolitan Area and United States Population 1980 and 2000 (percent) 1980 2000 Age Denver Denver Group Region U.S. Difference Region U.S. Difference 0-9 14.5 14.6 (0.1) 14.0 13.6 0.4 10-19 16.7 17.4 (0.7) 13.1 14.4 (1.3) 20-29 21.1 18.0 3.1 13.2 12.9 0.3 30-39 16.7 13.9 2.8 17.5 15.2 2.3 40-49 10.5 10.0 0.5 18.1 15.6 2.5 50-59 9.1 10.3 (1.2) 11.5 11 .4 0.1 60-69 6.1 7.9 (1.8) 6.2 7.3 (1 .1) 70-79 3.5 5.4 (1 .9) 4.1 5.9 (1.8) 80+ 1.7 2.5 (0.8) 2.3 3.8 (1.5) Total 99.9 100.0 100.0 100.1 Source: Denver Regional Council of Governments Employment/Population -� Policy Forecasts 1980-2000, June 1982. U.S. Department of Com- merce, Bureau of the Census, Projections of the Population of the United States: 1982 to 2050, Series P-25, No. 922, October 1982. U.S. Department of Commerce, Bureau of the Census, Statistical Abstract of the United States, 1984. Denver metropolitan area population in 1980 is heavily concentrated in the 20 to 39 year old age groups. Almost 38 percent of Denver's 1980 population was in this age group compared with 32 percent nationwide. Only 20.4 percent of the Denver metropolitan area population was 50 years of age or older compared with 26.1 percent of the nation's popula- tion in this age group. Differences in the age distribution are expected to persist through the year 2000. Over 35 percent of the Denver population will be between the ages of 30 and 49, compared with 30 percent for the United States. Twenty-four percent of the Denver metropolitan area population will be Appendix 2 201 ,.— over 50 years old, while this age group will account for 28.4 percent of the national population. The reason for the young ages of Denver resi- dents is the tendancy for migrants to be 20 to 40 years of age. The aging of the Denver metropolitan area population into the 60 years and over age groups after the year 2000 is demonstrated in table 80. In 2000, less than 13 percent of the population will be 60 or over. Over 25 percent of the population is projected to be in this age group by 2035 under Series 1 . Because of the greater influx of young migrants assumed in Series 3, the 60 and over age group increases to about 23 percent by 2035. The aging of the population during this period has a profound effect on several of the components of the economic and demo- graphic measures which are incorporated into the water demand forecasts. Table 80 Percent of Age Distribution of Denver Region Population, 1980, 2000 and 2035 Age 2035 Group 2000 2010 Series 1 Series 0-9 14.0 13.1 12.6 12.8 10-19 13.1 13.0 12.6 12.6 20-29 13.2 13.0 12.1 12.4 30-39 17.5 14.6 12.5 13.8 40-49 18.1 15.7 12.9 13.5 50-59 11.5 15.0 12.2 12.0 60-69 6.2 8.9 11.1 10.2 70-79 4.1 4.1 9.0 8.1 80+ 2.3 2.5 5.1 4.6 Total 100.0 99.9 100.1 100.0 Appendix 2 202 EMPLOYMENT FORECASTS LABOR FORCE PARTICIPATION RATES The labor force participation rate is defined as the ratio of per- sons in the civilian labor force divided by the population 16 years of age or older, whereas the Denver metropolitan area labor force in 1980 was 70.5 percent of the population 16 years of age and older. The U.S. Bureau of Labor Statistics (BLS) rate for the nation was 63.8 percent. As shown in table 81 , BLS and DRCOG forecast continued growth in labor force participation rates. Table 81 United States and Denver Metropolitan Area Percentage of Population 16 Years and Over In the Labor Force, 1980-1995 1/ U.S. Department of Labor (BLS) Medium Growth Scenario DRCOG Projections Year Male Female Total Male Female Total 1980 actual -- -- 63.8 80.3 61.1 70.5 1985 77.7 56.5 66.5 82.0 65.8 73.8 1990 77.2 59.6 67.9 82.2 69.9 76.0 1995 76.8 61.2 68.6 82.0 71.4 76.7 1/ BLS and DRCOG rates are not strictly comparable since BLS measures participation of the civilian labor force whereas DRCOG includes the military population. Source: U.S. Department of Labor, Bureau of Labor Statistics, Eco- nomic Projection to 1990, Bulletin 212, March 1983. DRCOG projects Denver metropolitan area rates to remain well above national rates. One main reason for this is the concentration in the Denver metropolitan area of young adults in the prime working years (see table 30). Also, Denver has a high proportion of in-migrants which tend to have very high labor force participation rates. Appendix 2 203 • C'� C 00I a^ U> .O N MN? .O 0 N CO el 0 0 d d �O mf.IN N M MMN MO C N .N id C N L 4 d \, 4)0 L 0 a I K -. a d 0 d a> It 0I O .-CO M 0 —.OM 01.0 CO U >' O .0 0 O �a ^ ui •Cr) ? in .• O.O ? OM ? C el d m L o 0 •-• O°• a CO 0 0 4°. N 0 0 CI at.. CO d I L K .. N a..O 0 C'7 CO OD N N CO N C.-.OMOl a '0 a la 01 '-I .DO UN NN N S ? UN O . C• M C L 0 O•- 7 a vz 2 a d 40, a 0 C L U L Of O L 0i a �,9-I a � w y I. a .. C DOI CO? •" in C1 O..ON U1 a .C co >+ - C ^0,01 O1 N .O la N U1? U1N • • S d col 0 d .0i d 43 �2 2 10 0 O el U .'I c C) G. d e▪ l 0 000 0 0 000000 O O It .C am 5 C C) 0. 0 0 0 0 0 0 0 0 0 0 0 0 4) 0. 0 0 d L 0 .O S ? S .0 OC0NNOO &- O id. d ate... +2 0 Cr)U1 S N C- .7- N N S l- N ? d L 00 rat d d E N MCON O. M OMN '- ON CV y4. 0 .0 .-1 7 4a> 0) 4' 7 C L O V O d d 0 C al P. 0 0 C 00 C O d C Ca 000 0 0 000 00 O 0 0 a> 0 as 0 0 0 0 0 0 0 0 0 0 0 0 C CO _N M S U1 01 CO Ul CO ON IllC ,0 a-1 4. ▪+ CO 10 7' U1 t- M N N In .- In 01 d a N + 0 C O a N L. .0- C in 01 UI.O ^ . U1 .O .di W 0 w B 7 9 i-.. d W N 0 C p d -. .01 C 000 0 0 000000 0 a� U 0 0 0 02 44 8P° F W d 000 0 0 000000� O 4) 0 L. G d 0 0. OU1 S 0 CO? UN UN a CO 04. .0 C' L CI) C O d .Oa a ? CO U1 .O CO 01 CO O CO N ON CO 4. a 0 4 NCO M .- NCO ON la ? S 01 d d .0 W a.2 bO rd r U1 ..l L 0 'd0 C N d 8 t a NO - 0 0 0 0 0 0 0 0 0 0 0 a CO d N 0 0 C CC N 0 0 0 0 O O O O O 0 O d a.1 g B > f] O M 00 0% 0 Co I� O M C' CO C- .0 4 0 0 d 'O .0 mSC� O U1 OU1.ON ^▪ S C Net N C O. NCO M Na0 N 2 m G d L B O O. yN O d = B d d gt 0 0 0 O O 0 0 0 0 O O a,In 0 el tor— Off. Y 0 7 0 NI M.O0 .O0 0 N NS .00 ? N inn O y a 0 ^J 8 U01 r 0. M N 1`0/1 la IS f1 Ill ON ▪ in Z • d D\ t la O U O 4 .. y 0 fr7 N CO W •'1 p N It C '0 .0,1 N •.Uj .01 O -.01 0 C V 0 a.3 00)) > 3 m rn 0 0 C it d-I CC 0 .0 u ID 4) d 6.. 1/4O a. � ••• x � W 4 '0 d Oat C at T, d m W L C .l 4 0 C .i VOL a) gdg .-I0. O 'O CO CD O a01 yp C C el 'rl 4 0 07f m A O. 4 d • d C 0 ..-I d O >. e s. .C >,.C d d d O L 07 0 +i .i .> 4 .i 07 W d E O .0 O CC CO 7 -0 a. L d Oral el 0 d 0 .0 >. B C ID B 0 0 0 +l d .-I d 4 d .i 4 is] W a. O CC CO CO O..-I O CI O 0 C 7 CO 4 01 4 4) C 4 4-. 0 .0 0 .0 C C .-1 L 0 4) 4 .a 0 .-1 4. 0 .-I a> 7 C 7 .Il d 0 C > d +1 •.1 d it a m L ,�. C C C d O. O 4 C d L > L .i ,C a1 U CO 7 EUEF 3 tr, Ca0CM O 0 ^I NI m ml U Ld fj Appendix 2 204 DRCOG projections also assume higher participation rates for elderly persons than BLS forecasts. Based upon the most current BLS information, DRCOG expects decreases in mortality rates, improved health and removal of employment barriers to persons beyond the age of 65. This also accounts for some of the differences between the BLS and DRCOG projections. Beyond the year 2000, the BLS forecasts for men and women by age group are directly applied to projected population by age group. The rates during this period by sex and by age group are held constant during this period. This assumes convergence of Denver metropolitan area labor force participation rates and rates for the nation. If Denver Region labor force participation rates declined more rapidly toward national rates, fewer residents would be in the labor force resulting in a substantial increase in job-related in-migration. In the absence of a migration cap, the effect of reducing DRCOG labor force participation rate assumptions would be to increase the Denver metropolitan area population forecasts, because more people would be necessary to support the same number of jobs. EMPLOYMENT GROWTH BY SECTOR Sectoral projections. Historical employment levels and the DRCOG forecasts for the Denver metropolitan area are presented in table 82. The annual growth rate during the three decades from 1950 to 1980 exceeded four percent. Hence, the 2.8 percent annual growth rate pro- jected for 1980 to 2000 is conservative compared with past trends. Projected employment growth is balanced among most economic sectors. Mining employment, largely headquarters or administrative staff, will experience slow growth, construction growth will be moderate Appendix 2 205 and manufacturing employment is expected to exhibit rapid early growth. Later erosion of the region's competitive advantage is anticipated. Manufacturing will account for only 13.4 percent of regional employment in the year 2000. Trade, finance and real estate, and services employment growth is forecast to be strong; as Denver continues to benefit from its role as the commercial and finanical center of the rapidly growing Rocky Mountain Region. Jobs in these sectors will represent over one-half of employment in the Denver Region in 2000. Moderate growth in government employment is expected, compared with past years. Comparison of DRCOG Employment Forecast with Recent Trends. The Colorado Division of Employment and Training estimates nonagricultural wage and salary employment on a monthly basis for the Denver-Boulder Labor Market Area. This region includes Gilpin and Clear Creek Counties as well as the six county Denver Region. The 1980 and 1983 employment growth rates are examined for each economic sector in table 83. Com- parison of the annual 1980 to 1983 growth rates with DRCOG projected growth rates reveals significant differences in two areas. Manufactur- ing employment actually declined from 1980 to 1983 as this sector con- tinued to exhibit the effects of the national recession. Government employment also declined from 1980 to 1983 reflecting cutbacks in Federal, state and local government jobs. In absolute terms, actual total employment levels were less than those projected by DRCOG during this three-year period. However, the adverse economic conditions prevalent during the early 1980's led to short term declines that should not contradict long-term DRCOG employment projections. Appendix 2 206 Table 83 Estimated 1980 to 1983 Employment Growth Rates for the Denver-Boulder LMA 1/ DRCOG Actual Forecast Annual Annual Growth Growth • Sector Rate Rate?/ Mining 12.3 8.6 Construction 1.5 6.3 Manufacturing (0.5) 4.1 Transportation, Communications and Public Utilities 3.1 2.2 Wholesale and Retail Trade 2.4 3.5 Finance, Insurance and Real Estate 4.5 4.7 Services 4.5 3.4 Government (0.4) 2.2 Total Non-agricultural 2.3% 3.6% Wage and Salary Employment 1/ The Denver-Boulder Labor Market Area includes Gilpin and Clear Creek Counties. ?/ Annual growth rates as projected for 1980-1985. Source: Colorado Division of Employment and Training, unpublished data, March 1984, Colorado Division of Employment Training, Colorado Manpower Review, Vol. 18 No. 1, January 1981. DRCOG, Employment/Population Policy Forecasts 1980-2000, June 1982. State of Colorado Assumptions. Assumptions related to economic conditions in the State of Colorado have an important influence on the DRCOG forecasts because of the interrelationship of the Colorado and Denver metropolitan area employment forecasts. The State of Colorado Office of Management and Budget has developed the following assumptions associated with its Governor's Blue Ribbon Panel forecasts: Appendix 2 207 . Colorado will experience increased coal production and a decline in conventional oil and gas output. Moderate oil shale development is assumed for Colorado's Western Slope. . Irrigated agriculture will gradually decline as groundwater diminishes state-wide and the cost of pumping and fertilizer increase. . Colorado's manufacturing sector will continue to expand faster than the U.S. as a whole. The difference between the growth rates will gradually narrow. It is assumed that there will be considerable expansion of the high technology component. HOUSEHOLD FORECASTS HOUSEHOLD SIZE DRCOG's projected household size for the Denver Region by applying household formation rates to population forecasts for each age group. These rates follow the Census Bureau's Series D household formation rates. Compared with Census Bureau Series A, B and C forecasts, Series D forecasts project the smallest decline in household size. The U.S. Series D and Series A forecasts are presented in table 84. Appendix 2 208 Table 84 Persons Per Household for the Denver Metropolitan Area and the U.S., 1940-20101/ Denver Metropolitan Area U.S.2/ Year Series D Series A Series D Series A 1940 -- -- 3.67 3.67 1950 -- -- 3.37 3.37 1960 -- -- 3.33 3.33 1970 3.05 3.05 3.14 3.14 1980 2.61 2.61 2.75 2.75 1990 2.35 2.21 2.58 2.41 1995 2.32 2.11 2.55 2.31 2000 1/ 2.29 2.02 -- -- 2000 3/ 2.27 -- -- -- 2035 2/ 2.27 -- -- -- 1/ Does not include persons in the group quarters which account for 1.9 percent of the population. ?/ Assumes U.S. Bureau of the Census Series II population forecasts. ^ 1/ DRCOG extrapolation of trend. 4/ Assumed constant household size. Source: U.S. Department of Commerce, Bureau of the Census, Projec- tions of the Number of Households and Families: 1979 to 1995, Current Population Reports Series P-25, No. 805, May 1979; U.S. Department of Commerce, Bureau of the Census, 1980 Census; Denver Regional Council of Governments, Household and Household Size Forecasts for the Denver Region, 1980-2000, July 1982. The 1970's exhibited sharp increases in age and sex specific rates of household formation for both Denver and the nation. DRCOG projec- tions, based upon U.S. Census Bureau Series D estimates, reflect a gradual stabilizing of the rates. Even so, the number of Denver house- holds are forecast to increase at a higher rate than population, because of the aging of the baby boom generation into the prime household form- ing age groups. As a result, average household size for the Denver Appendix 2 209 Region is expected to decrease sharply between 1980 and 1990, and gradually thereafter. Denver metropolitan area Series D based household size projections remain below the Series D forecast for the nation because the effect of migration will increase the portion of Denver metropolitan area popula- tion in the prime household forming age groups. Denver Region household size under the Series D forecasts reaches 2.29 persons per household in the year 2000 compared with 2.02 persons per household for the Series A based forecast. The Series D based forecast produces 147,500 fewer households or 12.1 percent less than the year 2000 Series A based fore- casts. COMPARISON WITH RECENT TRENDS DRCOG estimated that Denver metropolitan area households total 655,050 in 1983. The DRCOG long-term households projections inter- polated for that year is 671,800. This exceeds the actual 1983 figure by 16,750 households or 2.6 percent. The substantial reduction in household size expected between 1980 and 1985 has not fully materialized. One explanation is poor economic conditions in the early 1980's which would tend to depress the rate of household formation. OTHER FORECASTING ASSUMPTIONS The assumptions or caveats which DRCOG has applied during the fore- casting period are provided in Table 85. • Appendix 2 210 Table 85 Other DRCOG Forecast Assumptions for the Denver metropolitan area, 1980-2000 Factor Assumption Taxation/Regulation No major change in the region. Power/Water No long term shortages. Productivity 2.0% annual growth, 1985-2000. Inflation 6.1% annual increase 1985-2000. DRCOG also reiterates other caveats of the national forecasts, including a positive foreign trade balance, restrictive fiscal policies, and temporarily high Federal deficits. These assumptions are very tenuous, given recent trends national trends and conditions. Incorrect assumptions of this nature would affect the Denver economic and demo- graphic conditions, but the local projections are expected to be rela- tively insensitive to these trends. Other considerations specific to the Denver Metropolitan Area have not been considered by DRCOG in their long-term forecasts. Air pollu- tion, transportation restraints, crime rates and other quality of life factors are simply not addressed. Air quality is a prominent public issue in the Denver area as local officials and the U.S. Environmental Protection Agency have been in almost continual debate over Denver's air quality and what can be done to improve it. AN EPA monitored State Implementation Plan including inspection and transit options has been developed to achieve appropriate attainment levels by 1987. It is highly uncertain whether recent gains will be maintained or attainment levels achieved by 1987 or later. Air pollution is only one of many quality of life (QOL) considerations which make up the business and human environment as Appendix 2 211 perceived by present and prospective employers and residents have not been addressed by DRCOG. QOL measures include: . climate . air quality . economic opportunity . demographic characteristics . housing availability and costs . transportation systems . shopping opportunities . health and other public service levels . educational services . tax rates . crime rates . cultural opportunities . leisure time opportunities . cost of living . general aesthetics Denver has traditionally been perceived to have a positive quality of life (QOL) when compared with other major metropolitan areas. During the 1970,s, a number of QOL measures became less favorable, including air quality, housing costs, crime rates and transportation, while others, such as shopping opportunities, cultural amenties and leisure time opportunities, improved. The effects, if any, of the changing QOL levels upon the Denver economic and growth prospects are highly uncer- tain. It is noted, however, that Denver continued to experience rapid growth despite QOL measures during the early 1970s. Appendix 2 212 COMPARISON WITH OTHER FORECASTS The DRCOG Denver metropolitan area forecasts are based in part upon the Governor's Blue Ribbon Panel Colorado projections and the Chase Econometrics U.S. forecasts. In order to fully analyze the Denver metropolitan area forecasts, these state and national forecasts are com- pared with alternative state and national projections. Finally, alter- native economic and demographic forecasts for the Denver metropolitan area identified, described and evaluated. NATIONAL FORECASTS The Chase Econometrics population and employment forecasts were developed in July 1981. U.S. Bureau of Economic Analysis Office of Business and Economic Research Services (OBERS) forecasts were developed based upon 1977 Census Bureau Series II population forecasts. Table 40 compares these national population forecasts. The Census projections were formulated in 1982. Table 86 compares these forecasts. Appendix 2 213 Table 86 Population Forecasts for the United States, 1980-2030 Chase Year Census OBERS Econometrics 1980 (actual) 221,000,000 221,000,000 221,000,000 1990 249,731,000 242,979,000 240,000,000 2000 267,990,000 259,845,000 253,000,000 2010 283,141,000 274,803,000 N.A. 2020 296,339,000 289,583,000 N.A. 2030 304,330,000 299,817,000 N.A. Source: U.S. Department of Commerce, Bureau of the Census, Projections of the Population of the United States: 1982 to 2050 (Advance Report), Series P-25, No. 922, October 1982; Governor's Blue Ribbon Panel, Colorado: Investing in the Future, Volume Two: Forecasts, July 1981; and U.S. Department of Commerce, Bureau of Eco- nomic Analysis, Regional and State Projections of Income, Deployment and Population to the Year 2000, Survey of Current Business, Volume 60, No. 11, Novem- ber 1980, pp. 44-70. COLORADO FORECASTS DRCOG Denver metropolitan forecasts are based upon the Governor's Blue Ribbon Panel projections for Colorado developed in 1981 by the State Office of Planning and Budgeting. These forecasts linked the Colorado economy to that of the nation. Colorado population forecasts were then derived from the state employment projections. The Blue Rib- bon Panel projected Colorado population to grow from 2,865,000 in 1980 to 3,592,000 in 1990 and 4,552,000 by the year 2000. The Governor's Blue Ribbon Panel forecasts were developed in conjunction with the State Office of Planning and Budgeting in 1981 . The State Office of Planning and Budgeting in turn worked with the State Demographer, and the State Department of Agriculture and Department of Appendix 2 214 National Resources. Forecasts were also developed through projecting Colorado's share of U.S. employment as forecast in 1981 by Chase Econometrics. The Blue Ribbon Panel applied the state employment and population forecasts to project state-wide public infrastructure needs, including public water system investment requirements. Colorado nonagricultural employment is projected to grow from 743,000 in 1980 to 2,529,000 in 2000. Compared with Chase Econometrics U.S. forecasts, Colorado employment increases from 1.4 percent of U.S. employment in 1980 to 1 .7 percent in 1990 and 2.0 percent in the year 2000. Growth rates are projected to decline to 3.8 percent annual growth from 1980 to 1990 and 3.3 percent growth for 1990 to 2000. Table 87 presents the 1981 Blue Ribbon Panel employment forecasts by economic sector. Trade and services sectors account for over one- half of 1980 to 2000 employment growth. Annual growth in manufacturing --• is projected to decline from 4.5 percent for 1980 to 1990 to 1.7 percent beyond 1990. Slow growth is anticipated for government employment. In table 88, the Governor's Blue Ribbon Panel population and employment projections are compared with three other forecasts prepared for Colorado. Figure 11 depicts these alternative Colorado forecasts. The Governor's Blue Ribbon Panel and OBERS population projection differences reflect the underlying employment growth assumptions. Blue Ribbon Panel forecasts of annual employment growth are 3.8 percent for 1980 to 1990 and 3.5 percent for 1990 to 2000. OBERS forecasts are 2.9 percent annual employment growth for 1978 to 1990 and 1.5 percent annual growth for 1990 to 2000. Appendix 2 215 ,---, O N • > N C.• a• T IA M aN N • ' - O '- 0 1 C 02 00 0 m 1 ..• 10• •w•• l•• a• co t—• 10• T O '� SNP, N en T T - N• 0 L Id TSQI O 0t- IO OJ O. T T T 0 R T en - N en T a N en 0e • C O Q y 0 N • O• •1 T N O1 O• 0 a Jr •O a N en a a .- T m t., a 0 T N T N T O1 m a a . O C O•._ P 0a a 10 10 .0 T In 0 Cam}C ~m 1 !. N O1 in aCO kg) CO • l` •O• N N • r en a m a T at a p ei T O S S O 00 N N 'J 0 0 O N a l- P at a N N In ' P ' O a N N 0 0 N N In T • N O 0 p p L 2O O O S SO OOi 00 m a N a On O9 M N m a u N •O N a a CO 2 b r 8 p S at O O a 0O- cc O P b m CI CI 0 0 O N 0 tl O I y CON N T CO T N IN IT ' l- O L• OJ OJ 0 y T N N T 1. �` t- 0 3 • m O i- t. 8 p d co Fo a a0 a S u. S S S S S S O S 8 . el O O ..- a r a IA a N N 0 0 O N P en - b - N m N "'i L A N O O N pp pp 88 8 a C S O O a S 0 0 ; > I. T •O •O a l� O1 •O N C in i- N 0 a .• Ie.1 N en C O • .4 O• la 8 8 S O O S o S O S S S O O O O ONE C aD Clm lT- lN— lP- CO a N NI ` U .y H ^I •I m O •a O 0 8 U8 8g § 9g - s - In ' a a 8R le bl lak8 § § § ll § 1O °� L. 1 m ▪ T m a ' in N O tn N • L • 0N N E SL g {� 405 0i N is b Y •II:: V Y '' a O ^ 1• 8 0 `0 .�a w 0001 0 U w° 5 O G Y C 3. 844 C • O DO N A f• • Z V i F N 28 03 0 1 m Appendix 2 216 4. C 6 4. I 0 0 0 43 +°-I CO 3 �..� C 4 M O C 0 y i-i cc a3 e CM CM 8 O. y C 0 O C O co co cm o 0- maE 0 ON.-I ^ -I a T G 0 In In o �+ a C .I O O O I ° al o c >sCG O 0 O 4-1 tc.. .-71 .0 3 O N In � ' c -II O_ > IO CC CO N In °` COO tn +-I a3 C O C7 1- ^ N 43 C0 O 40 N 0 �1 I t. 0 0 CO O a > 0 r` w � w L. CO 4. 0 m 0 0 I 00 00 00 C) ° L v] In O .- . ° c o aCa a ay .43 •'I M U COO 4 0 In 0 O COO N 4„4 en S S 0 +'] 0 a C O O ' L 0 o° 0 o a O .i N W aO CO .-I ..-Ie6 0 coco a) co y .C 0 0 0 I 0 d O 0 0 0 O0 I L '°O U F O C 60. ON In t- S. w 0 7 co in u1 0 U CO CO [ '° 43 C CO CM en S 0 0 9 0' 0 >, 0 „I 0 .mil a O a 0 a 0 'a 6 w m I- A tl 'O O C .+ G O O O O V — N O a1 °• i of 0 O NO ON 0 0 O C O +01 ..a L. O� V— N L. • al �1 O .-I > 0 CO IL 0' N 4+ p' min'74 CO MD en • CO O U CI O 6° N RI S N0 7 O , 0 m a m oo 0 a 0 ,°� M W O 0 C O rn a 0 > 0 .- 0 Ci2 L .. L 0 .-I 0 0 0 I o 0 0 OW 0 0 0 c 0) -�' 4 .--I .O a 0 ON N N O4 L N .-I Ill O W .,1 G. CO ON In O CO Wa > CC CO an In C L 0 0 0 N R1 ? 00 +-I J-1 ''.4 in ow co O+ = m CO .- 0 O all co° •0. 0 0 0^ 0 N ON ON N0 N0 \ 0 I-I Appendix 2 217 O Cl Otn o W - in co (7 o N O CO c !rot 0 - N 0 N M 0 0l— .W G • 0 cn I c `0 c -J - U v as 0 r 0 0 w O v- �CC o N_ O V CD m;,.40 i O s w 2 cn 0in 0 O N •� Lkj RC I r-1 i z m \ 0 cn u... rn O I m a 0 co 1 1 01 0 0 0 0 0 0 0 t0 to 4 M N .- O suoiflm ui uoilolndod Appendix 2 218 In sum, the Blue Ribbon Panel employment forecasts are based upon long term historical data, individual analysis for each economic sector, and flexible use of independent forecasts. Analysis of alternative population forecasts demonstrate that the Blue Ribbon Panel's expecta- tions for state economic growth are similar to the U.S. Census Bureau and Colorado Division of Local Government. However, economic growth forecasts are substantially higher than OBERS forecasts for the state. DENVER REGION FORECASTS Alternative population forecasts for the Denver metropolitan area are illustrated in Figure 12. The DRCOG population forecasts for the Denver metropolitan area resemble projections of the Colorado Division of Local Government. As presented in table 89, DRCOG population fore- casts are 1.6 percent below Division of Local Government projections for 1990 and 1.6 percent above for the year 2000. The Division of Local Government utilized a cohort-survival model with constant levels of migration evident during 1970 to 1982 to develop these forecasts. DRCOG population forecasts are substantially higher than the BEA OBERS forecasts for the Denver metropolitan area. DRCOG projections are 7.8 percent higher than OBERS forecasts for the year 2000. DRCOG's "medium" forecasts for 2000-2035 previously developed for the EIS are represented by the dashed line in figure 12. The year 2035 forecast is lower than Series 3 and higher than Series 2. Comparison of DRCOG projections with the Governor's Blue Ribbon Panel forecasts for Colorado shows that the Denver Region accounts for 55.2 percent of Colorado's year 2000 population. As indicated in table 89, the DRCOG projections of the Denver metropolitan area's share of Appendix 2 219 1 F 0 . Q My M � : - N W O _ In C f] fJ rr C1/41vi in 1 = N I , c,c N z O o N Ml O CO ' w N m o - Y V) Q %, c N "Si C - O V) ..r N w i N Cd 0 0'� - v U o o� o L Jo } w O 2 N C14 t. > - C ZO O O O ,,� - O 4..? N CIS - O a 01 . O a O I I I 01 O O O O O O U1 4 P) N .- p suollin u! uogoJndod Appendix 2 220 C J.Oi a O O • Y 0 .- P 0 a Et0. • 3 • N OW as a ga g a ; • 8 8 8 in 8 0 cc ll co▪ in co a 0O O N O •-• N ■ p• en O Pi • '= t a I0 ... u $• e L al 0 1 0 ■ LI D.06 •a rym� • O M O 3 V `•a 8 2I • 8 e .L°a° q 0. Sa co ; c• >' d b P N V • L ▪ .. N al 8 ■ C TCC .L p �Lor a p •a a Jr l. • o la .q k a•i C G L%.-I N N N >.4 •• O 2�, ol S. 0 N ON ma LO • ^ O. ₹co L. _ • 8 8 8 cc . ~Li >• C in N L • S' • � ^i O A co b N > 0 "4 •C C i • b P T & g •••. N •L. —•a .cL Y O fO • O D• Oy a O a y �. N in L aC of re q b Lam , N N in '0 ~"S " L a d�W Y � 0. aO b ••N • C y 8 gal 8 8 8 0a ga d V O O O O) a - § O • a i O C • ca T WI Uri t. Di• C • .• O D Y y O• 1 0. F •- N N • „•j C C f C 8 a " e • C S . i e v Q C J J N N S a ! a I pp IL� in N in O " O O G I. Oi al O3 g ' 8L III O l•44 ci a w c s s i Le pp 0 W .a al N v •U CiI O 9 8 3I S.' . 8 Appendix 2 221 Hello