ALASKA STATE LEGISLATURE  JT. SENATE AND HOUSE HEALTH, EDUCATION AND SOCIAL SERVICES  STANDING COMMITTEES  HOUSE SPECIAL COMMITTEE ON EDUCATION  February 19, 2003 1:30 p.m. TAPE(S) 03-2, 3, 4  SENATE HESS MEMBERS PRESENT Senator Fred Dyson, Chair Senator Lyda Green Senator Gary Wilken Senator Bettye Davis Senator Gretchen Guess SENATE HESS MEMBERS ABSENT  All members present HOUSE HESS MEMBERS PRESENT Representative Peggy Wilson, Chair Representative Carl Gatto Representative Paul Seaton Representative Kelly Wolf Representative Sharon Cissna Representative Mary Kapsner HOUSE HESS MEMBERS ABSENT Representative Cheryll Heinze HOUSE SPECIAL COMMITTEE ON EDUCATION MEMBERS PRESENT Representative Carl Gatto, Chair Representative Paul Seaton Representative John Coghill Representative Peggy Wilson Representative Kelly Wolf Representative Mary Kapsner   HOUSE SPECIAL COMMITTEE ON EDUCATION MEMBERS ABSENT Representative Les Gara OTHER LEGISLATIVE MEMBERS PRESENT Senator Con Bunde Senator Kim Elton Senator Robin Taylor Representative Mike Hawker Representative Carl Moses COMMITTEE CALENDAR ^OVERVIEW: Alaska School District Cost Study Mr. Jay Chambers, American Institute for Research   WITNESS REGISTER Mr. John Walsh Northwest Arctic School District Ms. Amy Lujan Nome Public Schools PO Box 131 Nome AK 99762 Mr. John Torgerson PO Box 1068 Kasilof AK 99610 Ms. Elizabeth Frances Southwest Region Schools PO Box 90 Dillingham AK 99567 SUMMARY OF INFORMATION  TAPE 03-2, SIDE A [SENATE HES TAPE] CHAIR FRED DYSON convened the meeting of the Jt. Senate and House Health, Education and Social Services and House Special Committee on Education at 1:30 p.m. and invited Senator Therriault to relate the events that led to the "School District Cost Study." SENATOR GENE THERRIAULT handed out a time-line of events that led to Legislative Budget and Audit contracting $350,000 for a study on school district cost factors. The only factor that was not included was the school lunch program, because it is federally funded. SENATOR THERRIAULT explained that there were two respondents to the request for proposal (RFP), the American Institute of Research (AIR) and Information Insights-Fairbanks. AIR was not the low bidder, but their proposal was much more comprehensive and it was chosen. Two groups were put together to assist the study, one of school district business officers that interacted with AIR and the other an oversight committee comprised of Representative Fate, himself (Senator Therriault), Eddy Jeans from the Department of Education, Pat Davidson from Legislative Audit and David Teal from Legislative Finance. They made sure the consultant was getting access to information he felt they needed. As chairman of Budget and Audit, Senator Therriault said he was very protective of the consultants so the school districts couldn't pressure them to use certain numbers or do the study any particular way. The first time he saw the information was a draft report in November when the oversight committee met with AIR for two days in Anchorage along with the other (ALASBO) committee members. There were some concerns over anomalies in the data and the consultant was asked to make sure the district information was reported consistently in the survey instruments that were used to gather the information. The report was brought to Budget and Audit in early January. CHAIR DYSON said he understood that the contractor had produced the study and it had been accepted, so the legislature had no role in accepting the document. SENATOR THERRIAULT replied that was right and it was up to the legislature to make a policy call on whether or how to implement the findings of the study. AS 14.74.60(b) has a full page of the cost factors as they exist now. SENATOR CON BUNDE said that the methodology of the study was done on district cost factors and some people have suggested that district need should have been used instead. He asked Senator Therriault to explain why cost factors was chosen. SENATOR THERRIAULT replied that districts needed another way of saying adequacy and the appropriation language says to look at the cost factors. The consultant was not supposed to give the legislature a read on whether he thought the funding appropriated by them each year was adequate. He was to answer a specific question - if you only had a dollar to spend, how would you equalize the purchasing power around the geographic areas in the state. REPRESENTATIVE MARY KAPSNER said one of the criticisms she heard was that certain school districts benefited under the study. SENATOR THERRIAULT responded that at the November meeting, he realized the two Alaska Association of School Business Officials (ALASBO) presidents he invited represented two districts that were suggested to get an increase. This was not knowledge he had before they were invited and the reason he invited the one, is because he was the outgoing president. I don't believe they had any ability to steer the consultant beforehand nor did they have any ability to steer the consultant afterwards. It's just when we all got in a room and we opened up the report, we all looked at the numbers at the same time, basically, they were suggested for an increase. 1:50 p.m. MR. JAY CHAMBERS, American Institute of Research consultant, joined the committee and said that he had two team members on line, Lori Taylor and Joe Robinson, to help answer questions. He said that this undertaking was complex, especially because it was in Alaska where there are dramatic differences in diversity that do not exist in the same way in the Lower 48. His institute was here 20 years ago and did extensive traveling at that time and that benefited their ability to conduct this study. The study required a collaborative approach to get a better understanding of the factors that are faced by the diverse school districts around the state. The technical working group (TWG) was made up of eight school business officers who were intended to be representative of the type of districts that exist in the state. It consisted of Kerry Jarrell of Bering Strait SD, Michael Fisher of Fairbanks North Star Borough School District (SD), Melody Douglas of Kenai Peninsula SD, Dave Jones of Kodiak Island Borough SD, Dennis Niedermeyer of Lake and Peninsula Borough SD, Lucienne Harger of North Slope Borough SD, Barbara Stocker of the Sitka Borough DS, and Karen Goodwin of Southeast Island SD. Melody Douglas and Dave Jones provided help to the AIR as well. MR. CHAMBERS said that no one study could ever completely capture 100 percent of the variations in costs. "It's just too complex...." They worked at getting the big picture across the diversity of school districts within the state. The technical oversight committee helped review the data collection instruments and AIR tried to design instruments that would get the information that they needed to conduct their study. There are three major factors that affect the cost of education - pupil needs, scale of operations and the price of comparable resources. This study focuses on the third - the geographic cost of education index (GCEI). He said that the study that AIR did early in 1984 dealt with all three. MR. CHAMBERS stated the basic question they are asking is how much more or less does it cost to recruit and employ comparable school inputs across geographic locations within the state of Alaska. School inputs are teachers, administrators, various categories of noncertificated personnel, energy services, instructional materials and travel. He noted that AIR subcontracted with a group of engineers who helped them with the energy component of the study. The [GCEI] Index is on page 4 of the Summary and ranges from a low of 0.99 in Mat-Su to a high of 1.58 in the North Slope Borough. This means that the highest cost school district requires about 58% more resources than Anchorage to provide comparable school inputs. The high-cost areas are the far North and the southwest districts and the lower-cost areas are the interior and the southeast islands. There is still quite a bit of variation within the regions due to geographic factors. He explained that the next slide showed a comparison of the current Alaska cost index done previously by the McDowell Group and the new AIR geographic cost index. The McDowell Group focused heavily on existing patterns of spending and AIR's approach tried to identify differences in prices for comparable resources. This method is comparable to the one used by the Bureau of Labor Statistics in trying to identify differences in the cost of living over time in the United States or the Consumer Price Index (CPI). 2:07 p.m. Personnel expenditures are the biggest portion of school district budgets. In the Lower 48, this ranges from the high 70 - mid 80%. In the State of Alaska, there are many other factors - energy, transportation, and travel costs - that reduce that to some degree in some of the remote areas. There are some districts that spend less than 50% on personnel, but those are exceptions. Most of the districts spend 65 - 80% of their budgets on personnel. AIR's approach was to use econometric models. They started by understanding the full range of factors that affect the variation in wages and those include two basic collections of factors, discretionary factors and cost factors. The discretionary factors are where there is a choice of experience, education levels, the colleges personnel come from, etc. The cost factors are all the other things that impact variations in wages - the factors that underlie differences in the cost of living in different geographic regions or labor market conditions, amenities, access to shopping and cultural facilities. There is also the ability and willingness for people to work in different locations throughout the state. MR. CHAMBERS said to some degree the same factors that affect the labor markets for teachers and school administrators affect the labor markets for all other occupations within a region. "So this is not unique to education." The analysis focused on three categories of personnel, full-time teachers, other certificated personnel, and classified workers. His slide indicated the various categories the study used. They tried to predict the salary that would have existed in various districts using a simulated person who captured all the factors that affect wages in the statistical analysis. That number became a predicted salary for a certain area and that became the personnel cost index. The next slide illustrated patterns in variation of the index with the far north and southwest being the highest in average cost. He also provided a table that showed the variations in the mean index for personnel by distance from the nearest center of commerce. In the Lower 48, the vast majority of students and school districts are within 50 - 80 miles of city centers, but that's not true of many school districts in Alaska. Their numbers show that higher wages will have to be paid to attract teachers to those remote areas of the state. 2:17 p.m. MR. CHAMBERS said the energy budget takes up less than 5% of budgets in the Lower 48 and sometimes less than 1 - 2%. That is not the case in Alaska where up to 23% of the budget has been allocated to energy costs. The methodology they used includes an engineering component and building a set of prototypes to try to determine energy consumption levels and the price of various sources of energy services. They tried to remove the local choice from the analysis as much as possible. This was very difficult to do because there just weren't enough sample observations. Therefore, their approach was to take a prototypical building or buildings that functions similar to the type of buildings that exist in Alaska and simulated moving those buildings to the different climate zones within the state. TAPE 03-2, SIDE B    The engineers worked with the TWG and staff to come up with the prototype. Variations in energy costs are associated with climatic differences, the price of fuels transported to the locations where they are going to be used and the prototype differences associated with the efficiency of alternative fuel sources. The kinds of things they would not be able to control easily in an analysis are the operating hours, the differences in efficiencies in the different buildings, configuration of heating and thermostatic settings and fuel source used. They used three prototypes - the moderate, cold, and very cold - and considered school buildings, district office buildings, and out buildings. Data was collected through the district surveys on the price of energy services from each school site including transportation and storage costs. Those prices were converted into constant units so they could be applied to the engineering models. The far north and the southwest were the highest on average, but all the index numbers for providing energy services in Alaska were high relative to Anchorage. MR. CHAMBERS said the scope of using the cost of all goods would have been too big for this project's budget and the TWG helped them come up with a couple of goods they felt were representative - paper and a specific sized window pane. The issue was to get some idea of how much more money different districts pay for those kinds of items and to identify their transportation costs. They also included the alternative prices if the most efficient method wasn't available. Those were costed at each sight and the numbers were aggregated to the district level to calculate a final index, which shows that the smaller districts tend to have somewhat higher costs associated with their inability to take advantage of large scale purchasing and with the fact that they tend to be located in more remote areas. MR. CHAMBERS said they included contracted services associated with maintenance and operations, professional development and travel between the schools and district offices associated with administrative oversight, some professional development and itinerant services. Our goal was not so much to try to determine how much those expenditures should be, but rather to try to reflect the relative differences in the prices of different kinds of travel between these different locations faced by school districts. MR. CHAMBERS emphasized that AIR worked closely with the TWG to develop a reasonable approach to measuring travel costs from each school sight to Anchorage. Ultimately to create the overall geographic cost of education index, we had to take the personnel components, the goods components, the travel components and the energy components and aggregate them into a single index in much the same way that the Bureau of Labor Statistics does to try to aggregate the consumer price index (CPI) into a single index that reflects over time what the differences are in the price of living. He explained that the methodology they used is called the superlative index, which is different than the market basket approach. Many years ago the CPI used a fixed market basket index, which took a basket of goods like shelter, food and transportation, etc., and fixed their costs and then followed what happened to those costs as the price of the goods changed. They could just as well have reversed the process in time to see how prices have changed. The problem is in the real world as the relative prices of various goods change, people make substitutions. The cost of coffee goes up relative to tea, I start drinking more tea....The same thing happens with respect to school districts... The two indices will surround the true index, one too low and one too high. There is no way to get at the actual number, but the superlative index takes an average of the two numbers between two geographic differences, in this instance. In Alaska the price differences are so dramatic that they needed to understand the potential impact of substitutions on the way they employ the inputs. MR. CHAMBERS said that implementing the new index couldn't be done overnight without causing major disruption to the school districts. He counseled the lawmakers to phase it in over a period of five years since it didn't need to be updated more than every five years. He also thought that although Alaska had a good personnel database, the linkage between the certification files and the current personnel data system could be improved. While they are currently linked, only one is available electronically. He also recommended that they adopt a database for non-certified personnel like the one used for certified personnel. The state could adopt regular data collection on some of the non-personnel components like energy prices and cost of goods, as well. Once these systems were in place, studies could be made much more cheaply and accurately. Finally, he recommended that they have professional economists do the studies. 2:43 p.m. CHAIR DYSON turned the gavel over to the Chair of House HESS Committee, Peggy Wilson. CHAIR WILSON asked if there were any questions from the committee. SENATOR DYSON asked if the most current data used in all areas, since it seemed that enrollment numbers were used from FY 2000 and the cost numbers were from FY 2002. MR. CHAMBERS replied that they used the most recent data available whenever possible. Personnel data was used up through the 2001-2002 school year. MR. CHAMBERS explained that the enrollment data was used from schools and aggregated to the district. He said that Eddy Jeans had the ability to put in new data so legislators could use different numbers. He felt that the impact of changing one enrollment figure from 2000 or 2002 would have miniscule impact on the index. MR. JOE ROBINSON, AIR on-line consultant, added that their data was provided by the Department of Education. CHAIR WILSON asked Mr. Chambers to respond to the legislators' two pages of printed questions and began by asking if he would analyze the cases of considerable cost discrepancies between school districts in the same geographic regions. MR. CHAMBERS replied that would require him looking at each of the districts, which would go beyond their time to address. Those are some of the kinds of things they tried to address in the draft report at the November meeting. AIR was assured that the data they received from the districts were accurate at that time. CHAIR WILSON asked when they saw the discrepancies between close geographic sights, did they address those at the time. MR. CHAMBERS replied, yes, and that is when they double-checked the data from the school districts and census that was used for the personnel data, but found only minor adjustments. They decided to implement an improved methodology for the energy component by looking at a more continuous variation in the impact of climatic differences on energy costs than they had in the original study. This is because of the situation she just described where a couple of districts that were adjacent to one another, Haines and Skagway, had indices that were quite different from one another and logic suggested that they shouldn't be. MR. ROBINSON added, "Everything was checked and doubled checked again - not only by us, but by the officials at the ALASBO Conference themselves...." SENATOR GUESS asked if they changed the methodology on any other part of the cost indice at the November meeting. MR. CHAMBERS replied they just changed the energy methodology. SENATOR GUESS asked if there was any written explanation of why or how they changed it. MR. CHAMBERS clarified that it would be better to say that the methodology was refined. Originally, three different building prototypes were used - mild, cold and very cold. They were able to calculate three points of energy requirements based on that. In the earlier model, they assigned one of those three points to the districts based on the climate. In the new one what distinguishes a mild from a cold to a very cold are the number of heating degree days. They fall within ranges - lots of heating degree days versus small, relatively. The refined methodology tried to provide interpolations of the data between those two points. We drew a line between prototype one and prototype two and the districts that fell in terms of heating degree days, between those two points were placed at a point between them rather than just selecting one or the other depending on the climate zone that they were in...Every district had its own climate zone. REPRESENTATIVE GATTO said he disagreed with Mr. Chambers' statement that the real world is much more complex than any econometric model. Some of their data indicated a huge disparity in numbers from districts that were very close and that made him suspicious of the model. All of the questions on data, which might refer to how it was collected, become irrelevant if the model is faulty. MR. CHAMBERS responded: The model is intended to reflect patterns, if you're talking about personnel, for example, because we did not use an econometric model to estimate the price of energy. We simply asked what do you pay per unit per kilowatt-hour, per btu or some unit that we could translate into that. The only place we used econometric models was the personnel analysis and we didn't use econometric models for energy primarily because we didn't think we could really control for variations in the choices made by school districts. We did try to use econometric models for personnel because we felt that was the only systematic way that we had of factoring out the discretionary components for salaries and to isolate the cost factor. TAPE 03-3, SIDE A    SENATOR DYSON said that "slavishly going down the list [of questions] is not the best strategy" and that they wouldn't get all their answers today. Dr. Chambers wasn't prepared to answer some of the detailed ones. He asked if getting answers to questions in writing was outside the contract. MR. CHAMBERS said it was outside the contract, but arrangements could be make to provide them with answers. SENATOR GREEN said she was disappointed that this wasn't a two or three day meeting since so many people were visiting from other districts with questions. SENATOR DYSON explained that the committee had asked for questions three weeks ago, but they were very slow coming in. 3:10 p.m. SENATOR DAVIS said since the legislature was the body that would have to implement anything in the study, she wanted to know if the Department of Education and school districts could get a full report from Mr. Chambers staff. She asked if he had talked to any school district people. MR. CHAMBERS replied that he met with the technical working group and oversight committee, but he wasn't asked to speak with anyone else. SENATOR DAVIS said she wanted legislators to have access to AIR to better understand the study. MR. CHAMBERS said he would have to operate within the contract they currently have, but if the state wanted to amend the contract, he would provide whatever explanation and support is necessary. SENATOR TAYLOR said he couldn't understand why the City of Sitka isn't buying all of their supplies through Wrangell because they would save 100%. He also said the study indicates it would be 19% cheaper to travel in and out of Petersburg than it is to travel in and out of Juneau. That would be great news to his people because they pay a lot more than that right now. MR. CHAMBERS said he could answer that question in detail, but he would need the time to dig into the data. He thought that Senator Taylor misrepresented that part of the study by simplifying it the way he did. He asked Senator Taylor to give him a list of districts he is concerned with to see what drives that particular index and he would provide information on them. The reason the study used windowpane and paper as proxies to get at travel costs is because they worked with the school business officers to try and select some items they felt were representative. Supplies and materials represent a relatively small percentage of school district budgets. You could spend a lot of money on a project like this focusing on things that take up a very small percentage of the total school district budget and get a more precise estimate or you can focus your attention on the areas where there are big differences - and the big differences are in personnel and energy - yes transportation and travel are important components. I think we represented those numbers and I feel fairly confident that the approach that we took is a reasonable one given the relative percentages of expenditures that go to these components... SENATOR GUESS said she didn't have any real difficulty with the methodology, but there were some things that puzzled her. One is the use of just the general fund expenditures as weights, given the amount of federal funds that impact certain districts and other types of grants that districts get. I think if we're trying to get at the purpose of the entire costs differences, then I would think that we need to look at the entire budget versus just a part of the budget even though it is a significant part. MR. CHAMBERS said that was a good question and they had pondered it. There is no reason why one could not expand the scope of the budget weights to include, if the state chose to do so, to include those other components, because they are not trivial....We discussed this with the oversight committee... MR. CHAMBERS explained that the data has two matrices. One is labeled as the operating budget, but the legislature calls it the general fund budget and the other includes some of the federal funds. "There is no reason why, if the state chose to do so, that those weights could not be adjusted to reflect all of the components." SENATOR GUESS asked if the data is available. MR. CHAMBERS replied, "Those are data that come right from the Alaska Department of Early Education Development." REPRESENTATIVE KAPSNER asked Mr. Chambers what forum would be better for him to go into some of the details. One concern was use of a prototype for the energy portion because many of the schools in rural Alaska do not even come close to the prototype. "A lot of them were built by the Bureau of Indian Affairs in the 1950s and it's not an adequate reflection of a lot of Alaska." MR. CHAMBERS said he is willing to provide additional information and answer additional questions, but that would be beyond the scope of the contract. They would have to develop an addendum to continue. He informed them that their engineers looked at and talked to folks in the state of Alaska about using different prototypes for buildings that were built prior to 1985. MR. ROBINSON elaborated that the engineers determined that 1984 was the breaking point for building efficient school buildings. Due to the willingness of Anchorage and another advantage was that the size of the school district provided us with a large enough in to find whether there was a statistically significant difference in energy utilization of buildings before 1984 and those after 1984. Our analysis found that there was no significant difference - not at any measurable level - of significance between building that were built prior to that break point date and those that were built after it... REPRESENTATIVE KAPSNER said a lot of her schools and school districts would disagree. CHAIR WILSON elaborated that Representative Kapsner's school buildings actually have holes in the floors and missing insulation. There are huge differences across the state because buildings have not been maintained the way they should. MR. CHAMBERS said he was very sympathetic with what she described and that the legislature could decide to make capital investments so that buildings could be more energy efficient. A study of the type she was talking about would cost $1 million, not $350,000. As an alternative, he said, the legislature could say that as a state they accept whatever the school districts are spending on energy and improve the recording system in such a way to come up with that number. He said that they worked very hard on this study and he cared about his integrity and the quality of work that he does. I do this because I like what I do and think it's interesting and it provides good public policy. I'm not saying we're always right. We do the best we can. I take a bit of exception when - and not everybody has treated us this way - but when we just get treated off hand and kind of dismissed. I think what we've done is important and all we have to do is ask the teachers, would you be willing to go anywhere in the state for the same wage. I think you'll get your answer if you ever do that...We've got to devise systematic ways of providing and addressing the very important issues, but you have to address all of them. You've got to address the adequacy issues, the needs, the scale of operations, the geographic cost differences and you've got to do what most states don't do and that is separate or begin to think together about capital facilities and educational services. We fund them differently like they're separable...when, in fact, it affects very deeply the way we serve children with disabilities. SENATOR DYSON pointed out that some districts had spent money on creating energy efficient systems and would be penalized in the future. He said that some of the anomalies like Senator Taylor pointed out on the transportation costs "jump off the pages at us and make us wonder, as Representative Gatto tried to say, what went wrong..." He said the highest wages in the state are paid in the North Slope Borough and asked if the studies' figures were based on people who made the money there or people who resided and worked there. MS. TAYLOR, AIR on-line consultant, answered that the analysis of comparable occupational wages outside of education were based on a database that describes compensation at the point of employment, not the residence. SENATOR GUESS asked if those statistics were available. MS. TAYLOR replied no, the data is collected from the employer. REPRESENTATIVE PUAL SEATON said since they had identified teachers' wages as the biggest component of the entire model and that tenured teachers' wages are outside of the control of the district because those teachers can't be fired, how was that captured in the model. He asked if he was missing something. MS. TAYLOR replied that the comparable wage index adjusted for occupational differences across different parts of Alaska. It will compare the wages of motorcycle mechanics to the wages of motorboat mechanics and note where they are higher for the kind of job. It does not correct for the demographics of the specific individual involved, such as the years of experience or the degree that is held. Education is a field in which you get a lot of range of skill among the individuals who are teaching...that influences their salaries. The analysis for the salary index corrects for those demographics. The analysis of the comparable wage index or the average wage cannot correct for those demographics... REPRESENTATIVE SEATON observed if the difference in wages beyond the control of the individual districts is not captured, the cost to the district might be higher than the cost differential would show. MS. TAYLOR replied that with the personnel indices they tried to predict the wages of a typical educator in each district by moving that same individual to different parts of the state. It does not correct variations in the individuals or the mix of skills a district has currently. REPRESENTATIVE SEATON said it seemed that the district could be committed to paying higher costs in wages than the model would indicate they would have funding for. MR. CHAMBERS replied that was true for the way they had applied the model, but in general, school districts have the ability over time to make adjustments in distribution of experience and qualifications of their staff, but sudden changes, like declining enrollment, make it difficult for districts to make those kinds of adjustments in the short term. Some states have included it as a component in the cost factors, but he thought it was better keeping those things as options. After three years, about half of the staff has turned over or about 14-15% per year. MS. TAYLOR agreed. SENATOR GREEN said if a district made every effort to trim costs and had done aggressive contracting for the high dollar items and was an actively growing district with commuters for the bulk of their population, that the model is so badly skewed that a teacher salary increase wouldn't show up. MR. CHAMBERS responded that he thought she was talking about wages outside of education and wanted to be clear about their analysis. The comparable wages became an explanatory factor - an independent variable, if you will - a cost factor in the analysis of teacher wages. Ultimately, the personnel indices that we estimated for you come from a wage analysis of variations in teachers, school administrators and school classified personnel. In turn, each of those were related to a set of cost factors among which was this comparable wage index. So the comparable wage index is only one factor as part of a list of cost factors that affect the wages of teachers. It's not the only factor. It is an important factor, but it is only one of a number of factors. There are other factors that are more closely aligned with the districts... MS. TAYLOR agreed. She added that the comparable wage index was across all occupations for full-time employees. SENATOR GREEN said her question still goes back to the district whose data places them very high in the comparable wage category because of contract negotiations that took place subsequent to this data collection. MR. CHAMBERS replied that those are the kinds of things that one would have to consider in future updates. The numbers reflect not just districts, but patterns that are part of the labor markets throughout the state of Alaska. This question was asked in Dade County, Florida and he artificially increased the wages by 5% over everybody else and reran the numbers. The impact of that was that the district would have lost money doing that relative to the cost index, because there was no way, even increasing the largest district - Dade County is one of the top four districts in the country in terms of size...the impact on the index was less than 1%. The impact on the index didn't happen is what I'm trying to tell you. There are a lot things that would happen as a result of the increase in wages in a given district over time. SENATOR GREEN reiterated that the way comparisons were done some people were having trouble believing the report. SENATOR BUNDE said it looked like if one district raises its wages and another doesn't, that wouldn't be reflected at all. MR. CHAMBERS explained, "The benefits to the districts would be far less than any relaxation or otherwise as a result of the negotiation process." SENATOR BUNDE assumed that districts would be impacted negatively where there barter, instead of money was used. MR. CHAMBERS replied that it wouldn't be reflected in anything they had done. MS. TAYLOR responded that she thought the issue comes back to the comparable wage index. TAPE 03-3, SIDE B    MS. TAYLOR said in general if a particular line of work leads to off-the-books transactions, it would probably do that in much of the state and, therefore, wouldn't be a problem. It would only be a problem where a community is prone to do that. SENATOR BUNDE said one of the problems is of some people working in one area and living in another and that it's difficult to report that data because of privacy issues. He wanted to know if they could access that data and how challenging that would be. MS. TAYLOR responded that this study was done with community averages for finely detailed levels of occupations. They didn't have information about specific individuals outside of the educator files they were provided for the study. It would take an act by the legislature to get them access to that kind of data and it would be a major analysis. SENATOR BUNDE said she used wages from people who work on the North Slope in the oil industry, but then she said they don't have access to that data. MS. TAYLOR explained everything is adjusted for occupations. Folks who don't work in the energy industry will get paid more for cutting hair in the North Slope Borough than in Anchorage, because it's a tough place to live. The differential is what is used to peg the comparable wage in North Slope, not the fact that energy workers are paid so very well there. "Energy workers are only compared to other energy workers." SENATOR BUNDE said it was his understanding that people in Wasilla can live more cheaply because they are not factoring in that folks work in Anchorage rather than Wasilla. He asked if that was accurate. MS. TAYLOR said that all of their data described location according to where folks worked. They have no information about where anybody lives. SENATOR BUNDE thought that would skew the study. MR. CHAMBERS added that Mat-Su has a personnel index of .98 and Anchorage is 1.0 and that difference is not statistically significant, because standard error is applied. CHAIR PEGGY WILSON said that 1 percent is a big difference to Alaskan school districts. MR. CHAMBERS agreed and he wasn't saying to ignore that, but he wanted them to know that the study was attempting to show the big picture. CHAIR WILSON said that she would turn the gavel over the Representative Gatto for the next portion of the meeting. 4:03 p.m. REPRESENTATIVE CARL GATTO, Chair, House Special Committee on Education, asked if anyone had questions. MS. AMY LUJAN, Nome Public Schools, said she felt there were problems with the energy component because of the differences statewide. She also observed that different wage scales can affect a district's ability to sell positions to provide adequate education for students. MS. TAYLOR explained that the model is based on the philosophy that districts that can pay well have tended to be able to pick and choose among the people they hire. The consequences of that situation affect the mix of skills amongst the people they hire. Having an index that controls all the observable characteristics of the personnel allows them to strip out that historical affluence. The variations in compensation are related to location and not to qualifications of individuals. MR. JOHN TORGERSON said he had technical questions that he wanted the committee to forward to Mr. Chambers and asked if the numbers in the study had been rounded off. MR. CHAMBERS replied that the numbers were rounded off to the nearest two digits. MR. TORGERSON asked if all the models were run on the methodology rather than actual figures. MR. CHAMBERS replied that they used their proposed methodology. MR. TORGERSON asked if a working model was available to the public so they could add their own variables to see what the results might be. MR. CHAMBERS replied that he could change the variables, but the model wasn't designed to handle all combinations of variables. SENATOR TORGERSON said he was considering using actuals for energy, for instance. MR. CHAMBERS said he could substitute that number into the model if he wanted to, but doing that is outside the current scope of the contract. MS. ELIZABETH FRANCES, Southwest Region Schools, said some districts use waste heat for energy and asked if the study took that into account. Also, the study indicated that Dillingham generates its own electricity, but it never has. MR. ROBINSON replied that Dillingham is the only district that reported using waste heat. There was no equalization for waste heat. They didn't have the lowest index value for energy either. They used the information that was reported to them on the district surveys. SENATOR WILKEN directed Mr. Chambers to V of the Summary where it says, "The largest differences are most likely attributable to methodological differences underlying the two studies' calculation..." and asked what the two or three of the differences. MR. CHAMBERS replied that they were not charged with providing a review of the McDowell study, which was an analysis of actual spending as opposed to looking at differences in the prices of individual inputs. 4:30 p.m. SENATOR GUESS asked if the superlative approach was better than a market-basket approach if a district has high energy costs, but prefers to spend more money on teacher salaries. MR. CHAMBERS said that would lead to the issue of adequacy and in the study 20 years ago, they tried to develop an appropriate delivery system by reflecting spending patterns that were reasonable for the district that they served. A superlative index is a compromise between basing the weights on the actual district or Anchorage. A fixed market basket approach would have Anchorage as a base. SENATOR WILKEN said that the 1983-84 study was an effort to try to establish adequacy, but there was no report. MR. CHAMBERS differed and said that a very detailed report was submitted to Larraine Glenn in the Department of Education. REPRESENTATIVE SEATON said he thought road access would be an important component. MS. TAYLOR responded that going into the study they thought road access would have a lot of power to explain variations in salary, but it didn't. The comparable wage index was probably picking up enough of the variations so that inclusion of a measure for road access only added noise to the model. REPRESENTATIVE SEATON said if they presume that the study is accurate and that it identifies districts with higher costs that haven't been included in the current budget, a phase-in of 20% per year means that those districts that are paying the higher cost are not going to receive compensation for what they are actually paying. He asked Mr. Chambers to comment on that. MR. CHAMBERS said that deserves to be clarified and one of the major reasons for phase-in is: ...I don't think you could adjust a district's budget down by 5 or 10% and expect them to be able to respond to it in a reasonable way. This would be a crisis for that district. So, I would never want to put a district in that position. That doesn't prevent you from, however, from saying we're going to hold harmless or phase in over time those districts that have lost index value and go ahead and implement the index value on the other side of that. SENATOR GUESS said she found it interesting that poverty was not on the list especially given the relationship they know between poverty and learning. She asked if they looked at that issue. MS. TAYLOR responded: ...While poverty tends to add to the needs of individual students, it also tends to be more common in low cost of living areas such that it's very difficult to disentangle that part of student need that might make the job more difficult for teachers that would get translated into an expectation of higher wages from the fact of using a statewide notion of poverty tends to mean you observe more poverty in low cost of living areas which would tend to imply lower wages for teachers. In some of the models it cancelled out and in other models the result was perverse. The notion that where there was high poverty that the cost of living component seemed to dominate and you got a prediction that salaries would be lower where there is greater poverty and I didn't find that plausible. SENATOR GUESS asked how the difference in shipping costs affected the study. TAPE 03-4, SIDE A  MR. CHAMBERS explained there were enormous differences in shipping costs within the same community. ...We asked the question - by the most primary method by which you got this item to the school, what price did you pay for that item and what was the transportation cost implicit in that item. Then we asked in those instances where you did have to make the judgment that you're describing and purchase something immediately, what was the price that you would have paid by an alternative method of transportation and what percentage of the time do you think you use that alternative method of transportation. So, for example, if you 90% of the time did it by barge and 10% of the time you had it flown in, we would have rated the two different prices 90/10 to reflect that over-all difference. We had to ask each district that question about each of the items. That made it a little more complex as you can imagine. CHAIR GATTO said he was startled to find that in Pelican, items come in by floatplane. MR. CHAMBERS elaborated that this was an area where the study could be improved over time. He thought one could build in a number of different kinds of items to address the question that Senator Guess just raised and also to try to get districts to answer those questions in a very compatible and better way over time to improve the quality of that component of the index. "I would definitely think that is something that is very doable." CHAIR GATTO said one last concern was motivation and if he knew what was going on in this study, he might try to skew the figures he reported. He asked Mr. Chambers if he relied on his numbers rather than someone's casual answer to that question. MR. CHAMBERS said that he just gave the best explanation of why they did the energy cost they way they did it. He would expect that people would try to do the best they could for their districts. He said it would be extremely difficult for people to manipulate facts. SENATOR GUESS asked him to discuss the process that determined what the center of commerce was. MR. CHAMBERS said he would give her a list, but that basically they used distance as the determining factor. REPRESENTATIVE WILSON said in a lot of areas the kind of window that is being transported makes a big difference, because sometimes it weighs twice as much. She asked if districts were asked those specifics. MR. ROBINSON replied that the districts were asked to provide information on the typical window that they would get. The study group didn't want to burden the respondents with specific times. REPRESENTATIVE WILSON asked if one of the questions used was what it would cost to get a gallon of bleach. MR. ROBINSON replied yes. REPRESENTATIVE WILSON said some people would consider that a hazardous product and would send it a certain way and others would just go to the grocery store and buy it. She asked if they quantified that. MR. ROBINSON replied that they put hazardous materials on the survey and some clearly said they could just go to the store and buy some bleach. He pointed out that that question was not actually used anywhere in calculating any index values. MR. CHAMBERS added that heating oil could be a hazardous material. REPRESENTATIVE WILSON said that the report says the highest cost districts pay on the average about 29% more than Anchorage while low cost districts pay about 7% less. This sounds as if the personnel costs are based on what districts are actually currently spending adjusted to show everyone hiring comparable personnel. She asked if that was right. MS. TAYLOR replied that was correct for the most part. The report used what they were currently spending adjusted for the personnel mix, but it's the part of what they're currently spending that is systematically related to the school and environmental factors presented in exhibit 3-2 on page 18. MR. CHAMBERS clarified that those are cost differences as opposed to just pure average wage differences. REPRESENTATIVE WILSON asked if the energy cost differences were for a model building. MR. CHAMBERS said that was correct. REPRESENTATIVE WILSON asked why personnel costs were based on what school districts are paying, but energy costs use a model building and adjust for climatic zones. MR. CHAMBERS explained: The personnel component tried to use the statistical modeling to examine the labor market variations. In other words, we used a statistical model to do the analysis for us. The energy model - there is no comparable personnel alternative to the energy model. The energy model I could have used in an econometric model where the dependent variable instead of being teacher wages was expenditures per pupil for energy. And on the right hand side of the equation, the cost factors, if you will, would have been heating degree days or factors reflected in the climate. I could have done that. I chose not to do that because higher heating degree days means higher relative costs. Those higher relative costs may cause me to spend less on energy through other means....I didn't feel we had adequate controls. The problem is in an economic model or a statistic model the cost factor variables might be capturing choices that the districts are making and not just purely cost factors and I feel in the personnel case, we have a reasonable way to control for those...whereas in the energy analysis I didn't see any reasonable way of controlling for those factors. REPRESENTATIVE WILSON said she heard him say that they wanted to use statistical information, not reality. MR. CHAMBERS clarified: Statistical information is a reflection of the reality of the labor markets. For energy we were trying to create a reasonable representation of the school buildings recognizing the comments that we made earlier about the limitations of that. I wanted to pick up the gross differences in energy costs associated with climate and with the prices they were paying for energy services and try to take out of that the choices that districts are making to spend more or less because of operating hours or the way they set their thermostats or issues like that. CHAIR DYSON recapped what he thought there was consensus for the committee to do and since folks did not do a good job of getting their questions in by the deadline, he respectfully asked them to get their questions to Representative Gatto who would get them collated and back to Mr. Chambers. MR. CHAMBERS said he would like to respond to very specific questions. CHAIR DYSON adjourned the meeting at 5:05 p.m.