President Pendergrass AMBAG Oct 7, 1997 The 1997 AMBAG Population Forecast Minority Report The AMBAG Population Forecast is a failure. Forecast Technical Advisory Committee members have repeatedly complained that the population forecast has been constrained by a lack of money to do an adequate job, needs an adequate jobs/housing balance analysis and needs to be constrained by infrastructure and resources. We agree - and then some. We are am sorry to report that the AMBAG Population Forecast is a seriously flawed waste of thousands of dollars of your money and hundreds of hours of your staff's time and ours. The process and the Draft Forecast have ignored: 1) Historic trends. (Forecasts are notably higher than trends) 2) Historic mistakes. (30% overestimate for Marina in 1994) 3) Alternate methodologies. (Genuine Trend extrapolation and Bottom-up forecasts) 4) Concerns and comments from Forecast Technical Advisory Committee members. (e.g. Constraints ignored) 5) All data which conflicts with pre-determined results. 6) Large discrepancies ( more than 10% ) between US and State data sources. 7) Making data meaningful by using graphs. 8) HUGE, additive, cumulative Margins of Error. (Variance exceeding 150,000 for life of forecast for Monterey County alone) 9) The Self-Fulfilling Prophesy principle of forecasts unconstrained by resources such as water and roads. 10) Cumulative Environmental impacts CAUSED and induced by the forecasts. 11) Analyzing the limits to population growth by EXISTING infrastructure! 12) Analysis of the Carrying Capacity of the Region, Counties Communities and cities. Here are a few problems that can still be fixed: I. TWELVE THOUSAND PEOPLE PER YEAR ??? Lets look at Monterey County. * Today in 1997 Monterey County population is around 340 thousand (US Census Bureau) to 362 thousand (California Dept of Finance). The proposed AMBAG Draft forecast tells us that THERE WILL BE 400 thousand people in Monterey County by the year 2000 (400,907 pg 3). At the very LEAST that means Monterey County's population will increase by 12 thousand people per year. At most that means Monterey County's population will increase by some 25 thousand people per year. TWENTY FIVE THOUSAND PEOPLE PER YEAR! There is zero information or rationale in the draft (or anywhere else) supporting this. No evidence of the amount of jobs or housing needed. In fact Monterey County may have a housing deficit which would make this impossible. Past history is revealing. Ten years ago (1987) AMBAG forecast the population in 1990 would be 367,590. The actual census came in 14,000 people lower at 353,495. Off by 14,000 people when forecasting only 3 years into the future! AMBAG's forecast was found to be overinflated by an extra 14 thousand people when trying to forecast ONLY 3 years into the future. A reality check by AMBAG's Board in 1987 could have prevented AMBAG from being held up for ridicule when this was discovered. The proposed forecast numbers are reminiscent of the Hot Fudge Sundae Diet - There ain't no such thing. Please fix this forecast so that it reasonably resembles something reasonable. II. POPULATION FORECASTS ARE UNCONSTRAINED BY WATER AND TRAFFIC The forecast population numbers are TOTALLY UNconstrained by large real limitations - yet are constrained by other smaller items. San Benito County complains about this in their letter of August 14. Only an ostrich would assert that our communities are not constrained by current lack of water and traffic capacity. Yet that is exactly what the forecast numbers represent EVEN THOUGH THE LIMITS TO POPULATION GROWTH FROM EXISTING INFRASTRUCTURE ARE DESCRIBED IN THIS SAME DOCUMENT! Constraints to population growth are deaths, out migration (moving away), lack of drinking water, lack of traffic capacity and so forth. AMBAG staff chose to use only deaths and out migration - even though the FTAC (Forecast Technical Advisory Committee) OFTEN requested that other constraints be used. * Please explain why proposed population Forecasts are not constrained by lack of drinking water, lack of traffic capacity and other resources such as endangered species. Three types of forecasts were discussed at the FTAC meetings. 1. Forecasts unconstrained by resources or infrastructure. 2. Forecasts constrained by water, traffic and other existing infrastructure limits. 3. Forecasts constrained by sustainable use of resources sometimes called "Carrying Capacity." Only the first was given ANY consideration by AMBAG staff. The other two were quicky dismissed at every point even though a majority of FTAC members repeatedly expressed concern about this. * An alternate Forecast constrained by existing infrastructure should be prepared. * An alternate Forecast constrained by the "Carrying Capacity" of each community should be identified. III. THE FORECAST DOCUMENT ADVOCATES GROWTH! A Population Forecast has no business advocating growth. It is intended to be as accurate a forecast of future populations as possible. This document advocates and encourages growth in a almost every place it possibly can. *** In every important area the methodology is selected to make the highest possible forecast numbers. Combined with the refusal to identify the other varying data is literally scientific misconduct. For example: 1. 1995 Baseline population numbers vary from state and federal data sources by 10 percent. The higher numbers were used, the lower numbers are not disclosed. 2. Births vary by 3 percent depending on whether state or federal data sources are used. The higher numbers were used, the lower numbers are not disclosed. 3. Deaths vary by 3 percent depending on whether state or federal data sources are used. The higher numbers were used, the lower numbers are not disclosed. 4. Migration data is calculated from birth data which varies by 3 percent depending on whether state or federal data sources are used. The higher numbers were used, the lower numbers are not disclosed. In addition: Births were DROPPING SIGNIFICANTLY in Santa Cruz and Monterey Counties for the most recent data chosen - (Santa Cruz: 4,317 - 4,125 - 4,037 - 3,888) (Monterey : 8,198 - 7,494) Yet the average was used - not the downward TREND. Using the Average number of Births gives a lower population forecast than using the trend. Since AMBAG claims to base this entire forecast on Trends, why does the methodology suddenly leap to another method? Especially when the chosen method gives SIGNIFICANTLY higher forecasts than if the method remained consistent? * Please explain why averages are used here when they make forecasts higher - when the use of a Trend would make the forecast lower? Is it a coincidence that these inconsistencies will make the forecast numbers HIGHER? This is just one example of where methods are selected to support overinflated population forecasts (growth). Remember the first example given above - where the use of the California Department of Finance data rather than the US Census data "Today in 1997 Monterey County population is around 340 thousand (US Census Bureau) to 362 thousand (California Dept of Finance)." In this case the data supporting 10% higher population figures is not merely selected and explained - the lower data numbers ARE TOTALLY IGNORED AND NOT EVEN DISCLOSED. * Please take out ALL the "growth-is-good" methodology choices * Please use neutral methodology, * Please DISCLOSE all alternate data sources, * Please DISCLOSE all alternate methods of using each data packet * Please DISCLOSE how much the forecast would change if the other data or method were used. IV. BAD ASSUMPTIONS: Baseline data choices vary widely! US Department of Census data (originator of all baseline population data) differs from California DOF data by 10 percent and more. Yet California DOF data is used, in my opinion, BECAUSE it gives HIGHER Forecasts - not more accurate forecasts. Yet no margin of potential error is disclosed. * The forecasts should be prepared using US Department of Census data AND CDoF data with the difference illustrated. * The forecasts should disclose the margin of error is AT LEAST as large as the difference between the two government sources of US Census and the Calif Dept of Finance. 1. The forecast assumes the birth rate is correct. 1a. The forecast assumes the birth rate is constant. a. It uses as little as 2 years of data to forecast 29 years into the future ("Methodology and Assumptions" by Fassinger) b. Births were DROPPING SIGNIFICANTLY in Santa Cruz and Monterey Counties for the most recent data chosen - (Santa Cruz: 4,317 - 4,125 - 4,037 - 3,888) (Monterey : 8,198 - 7,494) Yet the average was used - not TRENDS. * Please explain why averages are used here when they make forecasts higher. Please explain why - when the use of a Trend would make the forecast lower. Yet Trends are used when it will make forecasts higher and averages would make the forecast lower? In every case the methodology which makes the highest possible forecast is used. This is an example of where methods are used to support overinflated population forecasts which induce growth. * Please explain why the methods should not be consistent? c. Birth data varies by more than three percent depending on whether you get it from the US Census or the California Dept of Finance. (Herald article Apr 25 1997) Death data varies by more than three percent depending on whether you get it from the US Census or the California Dept of Finance. (Herald article Apr 25 1997) These two pieces of data are used as the basis of the forecasts, yet this margin of error IS NOT DISCLOSED. These two pieces of data when properly combined could produce an error of 0 (ZERO) or as much as 6 (SIX) percent in the final forecast numbers all by themselves - yet THIS LARGE POTENTIAL MARGIN OF ERROR IS NOT DISCLOSED. This information show the absolute necessity of providing a confidence level for the forecast numbers just like the California Dept of Finance does. * Please provide a confidence level for the base data and the calculated numbers forecast from them. 2. The forecast assumes the death rate is correct 2a. The forecast assumes the death rate is constant It uses only 5 years of data to forecast 25 years into the future ("Methodology and Assumptions" by Fassinger) 3. The forecast assumes the in migration rate is correct 3a. The forecast assumes the in migration rate is constant 4. The forecast assumes the out migration rate is correct 4a. The forecast assumes the out migration rate is constant a. This claims to use 10 years of data to forecast 23 years into the future ("Methodology and Assumptions" by Fassinger) b. Migration - it turns out the ONLY data is "Phone conversation with Mary Heim (presumably from the Calif Department of Finance)." No data table is disclosed, no chart, no trends - NO NUMBERS ON PAPER! Who is "Mary Heim?" How does she know what Migration data is? We need to know what the in-migration rates have been and what the out-migration rates have been for as many years as possible. c. During FTAC discussions it was revealed there is a WIDE range of uncertainty for populations of migrant farm workers in Monterey County. The amount of migrant farm workers in Monterey County, and their locations needs to be discussed and a margin of error attached to those numbers separately. 5. The forecast assumes that ALL TRENDS WILL CONTINUE. Yet the forecast IGNORES THE TREND with the most years of data (90 years of population forecasts) and pretends to be a meaningful exercise using only 2 years of data at a critical, key point. * An alternative forecast should be prepared using the 90 year county Census population trends. 6. The forecast finds trends where there aren't any. Two years of (birth rate) data is not a trend. Five years of data (Monterey County death rates) are not a trend when the first three years are climbing and the last two are descending. This missing data is highly significant (resembling swiss cheese) - yet the flimsy data is used as though it was as solid as steel. * An alternative forecast should be prepared using the 90 year county Census population trends. 7. "A CHAIN IS ONLY AS STRONG AS ITS WEAKEST LINK." The forecast is based on one set of critical data which is only 3 years (possibly only 2 years) in duration. This mere 3 years of data is expected to be valid some 10 TIMES THAT DURATION into the future. (Birth data 1989-1991 is used to forecast to 2020) * An alternative forecast should be prepared using the 90 year county Census population trends. V. NO MEANINGFUL GRAPHS! The Draft document is almost completely uncontaminated by graphs, but loaded with data tables. The single graph is tiny and meaningless. By contrast the AMBAG forecast of ten years ago had 17 meaningful graphs charts. Since 1987 AMBAG has spent thousands of dollars on powerful computers and software - and we get no meaningful graphs???? Is AMBAG going backwards? * To convey meaning graphs must be full page. The result is the meaning of the forecast numbers is almost completely hidden. Pages of tables of data are meaningless without graphs. Only graphs, sometimes called "visual statistics", can bring trends and the significance of numbers into understanding. "Why do we use charts? (From "Using Charts and Graphs" by Jan V White 1984) 1. To show and tell facts effectively. 2. To give the audience a context within which to perceive the important facts. *** 3. TO PLOT RELATIONSHIPS THAT COULD NOT OTHERWISE BE UNDERSTOOD. *** 4. TO DISPLAY STATISTICAL RELATIONSHIPS MORE CLEARLY THAN WORDS OR NUMBERS PERMIT." * The Basic County forecast numbers should be graphed. * All Data Tables should be graphed or the data supplied to FTAC members so they can examine the graphs themselves. Please include a graph showing this forecast and contrast it to the 1994 forecast. VI. MARGIN OF ERROR The FTAC often urged the use of a Margin of Error, a confidence level or high and low forecasts. The will of the FTAC was fiercely denied in almost every instance by AMBAG staff EVEN though Mr Papadakis agreed it could be included (Draft pg 59). Simply examining the raw data shows that the proposed forecast varies by 22,000 people in 1995 - BEFORE it even starts forecasting the future. My simple calculations of the cumulative margin of error using the varying data and margin of error described above in Section III shows that the proposed forecast could vary by more than 150,000 people by the end of the forecast's 20 year lifespan. * A Margin of error should be included for the County level forecasts. It would show that the margin of error increases the farther into the future the forecasts go. VII. SELF-FULFILLING PROPHESY AMBAG's Excessive Population Forecasts INDUCE GROWTH. AMBAG forecasts -> Excess Freeways -> Excess Housing -> AMBAG forecasts r---> 1. Cal-Trans uses AMBAG (inflated) forecasts to justify Freeways. | (Cal-Trans letter to D. Dilworth July 9 1993) Inflated | forecasts cause Excess Freeway capacity to be built. ^ Rational forecasts cause reasonable Freeway capacity to be | built. AMBAG's highly overinflated 1987 data was used to | justify the Hatton Canyon Freeway. | ^ 2. Excess Freeway capacity is used by developers and financially | strapped Counties to justify excess planning new housing. | L---< 3. *Planned* excess new housing is used to justify inflating AMBAG population forecasts. (At the August 4 meeting, Monterey County complained that the forecast numbers were 5000 people too high. So instead of reducing the numebrs - the 5000 "forecast people" were "MOVED" to Seaside and Marina. This is another example of scientific misconduct.) This leads us back to #1 - Cal-Trans uses AMBAG (inflated) forecasts to justify Freeways. * The goal of AMBAG forecasts should be accuracy - not the highest numbers that can be cooked up. * Please establish and disclose the amount of growth the AMBAG population forecasts induce. * Please prepare an EIR on the Forecasts. David Dilworth and Mike Weaver, AMBAG Forecast Technical Advisory Committee Members