Loading...
HomeMy WebLinkAboutAgenda - 04-21-1998 - 10bORANGE COUNTY BOARD OF COMMISSIONERS ACTION AGENDA ITEM ABSTRACT Meeting Date: April 21, 1998 Action Agenda Item SUBJECT: APARTMENT ASSOCIATION OF NORTH CAROLINA -- IMPACT FEES FOR MULTI- FAMILY DEVELOPMENT PROPERTIES DEPARTMENT: COUNTY MANAGER PUBLIC HEARING: Yes —x—No BUDGET AMENDMENT REQUIRED: Yes _x No ATTACHMENT(S): - 2/19/98 Letter to Commissioner Halkiotis - 1/23/98 Letter to County Manager - 5 -19 -97 Letter to Marvin Collins plus attachments - May 1997 Report of Dr. Emil Malizia (UNC - Chapel Hill - Dept. of City & Regional Planning) INFORMATION CONTACT: County Manager County Attorney TELEPHONE NUMBERS: Hillsborough - 732 -8181 Durham -688-7331 Mebane -227-2031 Chapel Hill -967-9251 PURPOSE: To receive a report from the Apartment Association of North Carolina that recommends lower school impact fees for apartment units. BACKGROUND: On June 7, 1993, Orange County adopted an ordinance establishing a system of impact fees to be used to finance a portion of the cost of public school space needed because of new residential growth. The ordinance became effective July 1, 1993 and impact fees have been collected in the Orange County Schools District and Chapel Hill - Carrboro City Schools District since that time. Fees collected in each school district are placed in a trust fund for the purpose of funding new public school space in that district. The current impact fee for the Orange County Schools District is $750.00 and $3,000.00 for the Chapel Hill - Carrboro City Schools District. The basis for the impact fee structure is documented in Technical Report - Calculation of Proportionate Share Impact Fees for Financing Public School Capital Needs, commonly referred to as the Technical Report. The Technical Report was first revised in May, 1995 to include new cost data for school construction. It was revised again in May, 1996 to 2 include updated cost information, the recommendations of the School Facility Construction Standards Work Group, and the findings of a graduate student study of student generation rates in new housing built between 1992 and 1994. The 1996 update of the Technical Report concluded that the maximum amounts that can reasonably be charged on impact fees are $3,404 for the Orange County Schools District and $11,593 for the Chapel Hill - Carrboro City Schools District. The Apartment Association of North Carolina has requested lower impact fees for apartment units based on a report prepared for them by Professor Emil Malizia of the UNC Department of City and Regional Planning. The report concludes that apartments generate fewer students per unit than traditional single - family dwellings. While interesting, the report may not have applicability to Orange County since it is based on statewide, rather than Orange county- specific data. Previous work on the subject performed by Orange County showed no relationship between housing size and number of students. The original Technical Report as well as updates to it were prepared by retired Planning Director, Marvin Collins. Plans for a new update are imminent and the UNC Department of City and Regional Planning has been contacted regarding preparation of a 1998 Technical Report update. The County Attorney's view is that an analysis of student generation rates by housing type in Orange County, considering separately single family housing and multi - family housing, can be undertaken to determine whether Orange county data supports a different impact fee for the two housing types. Such an analysis was done for Chatham County housing. The data resulted in a supportable school impact fee of $3,529 for new single family housing unit and $1,164 for a new multi- family housing unit. The researcher reached the following conclusion: "Multifamily fees are significantly lower than single family and mobile home fees due to the relatively low number of students per multifamily household in Chatham County." (Emphasis added.) Orange County's impact fee presently includes all student generation data per household. A reduction in the fee for multifamily housing will necessarily result in an increase in the fee for single family housing. A lower justified impact fee for multifamily housing could also tend to make the impact fee less regressive. RECOMMENDATION: The Administration recommends that the Board receive the report from the Apartment Association of North Carolina, but defer any response until such time as the update to the Technical Report is completed. 3 Estates, Inc. .1al Estate Development Corporation February 19, 1998 Orange County Board of Commissioners Steven Halkiotis, Vice Chair 1007 Panther Court Hillsborough, N.C. 27278 Dear Vice Chair Halkiotis, Estates, Inc. anticipates beginning construction of 240 apartments in Orange County this spring. When we began the planning process in 1995, for our project, county impact fees were S750.00 per unit. This fee was a blanket assessment that applied to single family dwellings and apartments. Since that time the fee has quadrupled to $3,000. Those of us that develop multi - family properties believe that an equal assessment of impact fees is unjust and punitive. We want to pay our fair share of costs. We know that new development impacts the cost of providing services hence impact fees. Proportionality, is a concept whereby the impact of costs to a governing body (such as Orange County) can be quantified to allocate true impact and recognizes that multi- family differs from single family. This is not an exact science but evidence indicates that the impact of single family to multi- family is about 3 to 1. Dr. Emil Malizia of the Department of City and Regional Planning, UNC at Chapel Hill, has done a great deal of work in this area Our contention that multi- family is being assessed disproportionately appears to be supported by his analysis. On February 10,1 met with Mr. Link along with several of my colleagues in multi- family development and Mr. Ken Szymanski, Executive Director of the Apartment Association of North Carolina to make our case that we felt impact fees in Orange County were disproportionately being applied It is my hope that this issue will be discussed at the Commissioner's meeting, March 17th. We are in a consensus that this issue need revisiting. As stated, we want to pay our fair share. We appeal to your sense of fairness. Thank you for your consideration. Sincerely, W144,*� Ned Pendarvis 1401 Main Street P.O. Box 1 1739 olumbia, South Carolina 29211 803.765.93 18 FAX 803.765.9427 AANC M Apartment Association of North Carolina Representing the Multi - Family Housing Industry in North Carolina January 23, 1998 Mr. John Link Orange County Manager P.O. Box 8181 Hillsborough, NC 27278 Re: Orange County School Impact Fees Dear Mr. Link PS/ �J "09- V Our Association continues to have a keen interest in having a conversation with yourself, Orange County Attorney Jeff Gledhill, and the Orange County Board of County Commissioners on the subject of the calculation of Orange County school impact fees. We contend that the imposition of these fees on multi-family rental housing, which is a readily identifiable class of properties that by its very nature houses fewer children than single - family housing and thus has demonstrably less of an impact on educational facility costs, is inequitably high and very possibly illegal. Please see the attached letter written to Mr. Gledhill on September 30, 1997. The North Carolina Court of Appeals, in a 1989 case, adopted the "rational nexus" test requiring exactions to be properly and fairly related to the need for new public facilities generated by a new development. In our view, apartment developers have, since 1993, paid and are continuing to pay school impact fees in your county which are substantially beyond that portion of the cost which is reasonably attributable to the needs created by new apartment construction. Our Association commissioned an independent University of North Carolina study on this subject last year. We have shared all of the findings with Mr. Gledhill, and Mr. Collins and Mr. Bell of the Planning Department. UNC's Dr. Emil Malizia wrote, "Charging impact fees on apartment units at half the fee charged to single - family or less would be more consistent with the actual impacts on public schools. Furthermore, the impact fee would be far less regressive as a result ". We note that the consultant for neighboring Chatham County states that the maximum supportable impact fees per housing unit are $3,529 for single - family and $1,164 for multi- family. Units of government have a duty to achieve the highest possible level of equity in attempting to recover actual costs, or in forecasting future cost recovery. I will phone you next week for your reaction to this letter and for your advice on how to best proceed. Thank you. Sincerely, Kenneth N. Szymanski, AICP Executive Director cc Mr. Jeff Gledhill 711 Fast Morehead St. • Suite 201 • Charlotte, North Carolina 28202.7043349511 • FAX 704 - 333 -4221 5 AANC , - Apartment Association of North Carolina Representing the Multi- Family Housing Industry in North Carolina May 19, 1997 Mr. Marvin Collins, AICP Orange County Planning Director 306 F Revere Road Hillsborough, NC 27278 Dear Mr. Collins: Thank you for visiting with me recently on the subject of the history and evolution of Orange County school impact fees. As you will recall, I left with you two copies of the recent report, "The Impacts of Recent Residential Development on Public Schools in North Carolina Urban Areas: Single - Family Housing Compared to Apartments ", prepared by Dr. Emil Malizia of the University of North Carolina. Further, you inquired whether this data might be available solely for the Raleigh - Durham- Chapel Hill area, with an eye toward implementing the results in setting Orange County impact fees. Dr. Malizia has responded to this request with the attached memo dated May 14, 1997. While he does not wish to analyze any geographic subset from our dataset, he states that data from the PUMS (Public Use Microdata Sample) can be used for specific urban areas. He suggests comparing "Single Family Detached" in Raleigh - Durham to "Apt (units >50)" in Raleigh- Durham. The data analysis (enclosed) indicates that, on average, 0.50 persons aged 5 through 17 reside in the "Single Family Detached "; while, on average, 0.02 persons aged 5 through 17 reside in the "Apt (units >50) ". This ratio is 25 to 1! The Chatham County report (referenced by Dr. Malizia) states that the maximum supportable impact fees per housing unit are $3,529 for single family and $1,164 for multifamily (a ratio of slightly more than 3 to 1). Mr. Collins, our Association has a strong conviction that units of government in fact have a duty to achieve the highest possible level of equity in attempting to recover actual costs, or in forecasting future cost recovery. While the allocation of these costs is an inexact science, and there will always be questions about equity, clearly greater equity is achieved by establishing differential fee amounts for single- family and multi- family residential types than by maintaining a flat, uniform fee. Dr. Malizia's independent research suggests that these two housing types generate, on average, substantially differential impacts on public schools. Tailoring the fees beyond the averages, based on the specific bedroom mix of proposed housing developments, can bring about even finer resolution of fairness, because of the observed variation of generation rates by bedroom size (for both single- family and multi-family). 711 East Morehead St. • Suite 201 • Charlotte, North Carolina 28202.7043349511 • FAX 704333 -4221 6 Letter to Mr. Marvin Collins tifay 19, 1997 p.2. You had alluded to an upcoming staff report/ recommendation to the Orange County Board of County Commissioners on this subject. It is our sincere rope that Dr. Malizia's new research can soon be translated into equitably structured impact fees in Orange County. We are very concerned about the disproportionate effects that Orange County impact fees have on multi- family housing costs. In other states, litigation has brought about the calculation and assessment of differential rates by housing type. By their very definition, impact fees charged must not exceed a proportionate share of the cost incurred or to be incurred in accommodating the development paying the fee. If our Association can assist your work in any way, please contact me. Thank you very much. Sincerely, Kenneth N. Szymanski, AICP Executive Director Attachment cc: Apartment Association of North Carolina Board of Directors Dr. Emil Malizia DEPARTMENT OF CITY AND REGIONAL PLANNING University of Notch Carolina at Chapel Hill MEMORANDUM TO: Ken Szymanski FROM: Emil Malizia DATE: May 14, 1997 SUBJECT: NC Urban Area Data on School Age Children by Housing Type I. Orange County Survey Data I have checked the dataset. Only 96 records (Z 1. 1% of the sample) were drawn in the Research Triangle Area. Furthermore, only 13 surveys were Orange County households. I will send a copy of the entire dataset with the Carolina Planning article as soon as the final version is typed. You can use the data in any way you wish. I do not want to analyze any geographic subset of the data since the survey was not designed to provide accurate area - specific statistics. II. PUMS Data On the other hand, the PUMS data can be used for specific urban areas. They are generally consistent with our survey results. The Carolina Planning article will contain an exhibit with PUMS data that are comparable to our survey data. It gives statistics for apartments and singiafantily households with 4 bedrooms or more. I have enclosed the table we compiled with PUMS data that shows households that send their children to public schools. You can compare "S.F. Detached" in Raleigh- Durham to "Apt (uniu >50) in Raleigh - Durham. Single - family households clearly generate more persons and many more public school - attending children. I have also enclosed p.3 from the report Tiisebler did for Chatham County. The generation rates for - Single Family" and "Multi- family" are PUMS data for Orange County amd Chatham County. Marvin Collins can get the entire report, dated September 30, 1996, from Keith Meggiason, Chatham County planning director (919 -542- 8204). M M. Conclusions These PUMS data indicate that singlo-family households generate significantly greater impacts on public schools, at all levels, compered to multi- family households. Charging impact fees on apartment units at half the fee charged to single - family bouses Qr few would be more consistent with the actual impacts on public schools. Furthermore, the impact fee would be far less regressive as a result. For example, a fee of $3.000 on a house priced at $240,040 is 1.25% of the price; the same fee on an apartment costing $60,000 is 5.00/9. Dropping the impact fee to S1,500 still generates a burden on the apartment of 2.5%. • Type l Val ( S+ocal 0 -4 3 -l0K 1 14 -17 lt• 9 .0tal Ashevtlle 2.39 0.16 0.18 0.09 0.23 1.65 Burlington 2.46 0.15 0.19 0.10 0.17 1.90 Charlotte 2.56 0.19 �• 0.21 0.10 0.14 1.92 Fayetteville 2.75 0.:7 0.:9 0.12 0.13 1.93 Crown abcro -ice -HP 2.47 0.17 0.20 0.09 0.13 1.99 Hickory - Morganton 2.56 0.17 0.20 0.10 0.15 1.94 Jacksonville 2.83 0.34 0.30 0.11 0.14 1.93 Raleigh- Ourhaf 2.42 0.17 0.19 0.09 0.11 1.86 )illmingtoa 2.42 0.15 0.19 0.10 0112 1.88 Mobile Howes Asheville 2.59 0.31 0.2S 0.09 0.13 1.90 Burlington 2.53 0.24 0.20 0.10 0.14 1.95 Charlotte 2.59 0.26 0.24 0.10 0.14 1.95 Fayetterille 2.60 0.35 0.24 0.10 0.09 1.91 Greensboro-r6-Ha 2.52 0.24 0.23 0.10 0.13 1.61 Hickory- Morganton 2.56 0.26 0.26 0.10 0.12 1.92 Jacksonville 2.66 0.36 0.25 0.08 0.11 1.66 Raleigh- Ournam 2.68 0.24 0.24 0.09 0.11 1.80 Wilmington 2.40 0.21 0.26 0.07 0.33 1.73 C.F. (D /A) Asheville 2.49 0.13 0.17 0.10 0.12 1.97 Burlington 2.55 0.13 0.20 0.11 0.13 1.97 Charlotte 2.71 0.18 0.22 0.11 0.16 2.04 Fayetteville 2.97 0.23 0.29 0.14 0.19 2.02 Gretnsboro-WS-HP 2.61 0.16 0.21 0.10 0.14 2.00 Hickory- Morganton 2.65 0.15 0.19 0.11 0.17 2.04 Jacksonville 2.99 0.3S 0.34 0.14 0.16 2.00 Raleigh - Durham 2.65 0.19 0.22 0.11 0.14 2.00 Wilmington 2.60 0.14 0.10 0.12 0.14 2.01 3.1'. Detached Asheville 2.50 0.13 0.10 0.10 0.12 1.90 Burlington 2.53 0.14 0.20 0.11 0.13 1.99 Charlotte 2.73 0.19 0.22 0.11 0.16 2.05 Fayetteville 2.87 0.22 0.20 0.14 0.19 2.04 Greensboro- 143 -HP 2.64 0.16 0.21 0.11 O.IS 2.02 Hickory- Morganton 2.66 0.15 0.19 0.11 0.17 2.05 Jackaonville 2.96 0.20 0.34 OAS 0.18 2.01 Raleigh- Durham 2.70 0.18 0.21 0.12 0.15 2.03 Wilmington 2.62 0.14 0.18 0.12 0.14 2.03 S.F. Attached Asheville 1.88 0.10 0.07 0.05 0.03 1.64 Burlington 2.24 0.06 0.21 0.07 0.04 1.06 Charlotte 2.20 0.12 OAS 0.07 0.12 1.74 Fayetteville 2.99 0.44 0.43 0.14 0.16 1.02 Greensboro -NS -HP 1.92 0.10 0.11 0.04 0.04 1.63 Hickory- Morganton 2.34 0.10 0.24 0.11 0.21 1.67 Jscks00+111& 3.11 0.73 0.34 0.06 0.03 1.96 Raleigh- Ducham 2.09 0.13 0.11 0.05 0.07 1.71 Wilmington 2.24 0.22 0.18 0.10 0.11 1.62 Apt. tall units) AMovillo 1.82 0.16 0.12 0.06 0.07 1.41 Burlington 1.94 0.10 0.17 0.03 0.04 1.S2 Charlotte 2.02 0.10 0.15 0.06 O.OB 1.54 Fayetteville 2.36 0.34 0.29 0.06 0.06 1.62 Gr..nsbere -WS -61P 1.90 0.16 0.13 0.06 0.04 1.46 Hickory- Mocgsntoa 1.96 0.19 0.17 0.06 0.08 1.47 Jacksonville 2.18 0.30 0.17 0.05 0.03 1.63 Raleigh - Durham 1.91 0.13 0.12 0.05 0.06 1.56 Wtledngton 1.97' 0.17 0.13 0.05 0.04 1.50 Apt. WaitacS07 Ambeville 1.99 0.17 0.14 0.07 0.08 1.43 Sur11m9ton 1." 0.10 0.17 0.03 0.04 1.53 Charlotte 2.05 0.19 0.26 0.06 0.09 2.3s reyettevllle 7.36 0.34 0.29 0.07 0.06 1.63. Greensboro-w-H! 1.91 0.17 0.14 0.06 0.00 1.47 elokory- Moc9antoe 1.90 0.19 0.17 0.06 0.09 1.47 Jackawavlllt 2.17 0.29 0.17 O.OS 0.03 1.62 malelga -Derive 1.94 0.14 0.13 6.0s 0.06 1.61 Milalagton 1.99 0.14 0.14 0.05 0.07 1.59 Apt. funitG >50) IAsl.evllte 1.24 O.OS 0.01.1 0.00 0.00 1.20 $'rliflgton 1 71 0.00 0.00 0.00 0.00 1.21 1 O charlotte 1.34 0.04 0.02 0.00 0.01 1.27 Fayetteville 1.00 0.00 0.00 0.00 0.00 1.00 Greensboro -MS-Hp 1.15 0.02 0.02 0.00 0.01 1.14 Hickory- MOryanton 1.45 0.00 0.30 0.00 0.00 1.15 Jacksonville 3.11 1.11 0.00 0.00 0.00 2.00 Aalelah- Durham 1.23 0.01 0.00 0.01 0.01 1.21 MSlmingLOn 1.46 0.00 0.00 0.00 0.00 1.46 Othar Aehev111e 1.42 0.11 0.00 0.00 0.00 1.31 aurlington 1.62 0.00 0.00 0.17 0.15 1.30 ' Charlotte 1.25 0.)0 0.14 0.12 0.10 1.79 rayettoville 3.72 0.69 0.36 0.12 0.24 1.81 Greensboro -W-HP 2.46 0.20 0.25 0.09 0.12 1.80 Hickory - Morganton 2.22 0.09 0.04 0.00 0.11 1.89 .Tackvo vllle 3.06 0.14 0.31 0.20 0.43 1.97 Halvigh- Duchar 2.21 0.16 0.08 0.06 0.11 1.79 V411rington 22.14 0.06 O.TO 0.10 0.12 1.75 4 0 Table 1 below summarizes the restilii,ig maxinruut supporutble• iutpacr fces by type cif' housing unit. These fees total $3,529 for new single family housing units, $1,104 fi_,r new 12:u!!ilj:tZily units, and $3,714 for new mobile hornc�, Multifaitiily fees are signifiicanily lower than singtc fartuly and mobile home fees due to the relatively low nuiliber of students per multifamily household in Chatham County. Impact fees for mobile hot ics ure higher than single fans ly impact fees because of a greater property tax credit off -sct for single family homes. (Single family homes are more expensive than mobile homes and therefore receive more credit in the forni of future property tax payments used to pay off school debt service.) The impact fees ultimately adopted by the elected officials could be less than the ttiaxirnum justifisable and /or they could be phased. Table 1: Public School impact Fees Public School Students Per Household Single Family (SF) Multifamily (MF) Mobile Home & Other (MN) Level Of Service Nigh Building Sq_ Ft. Ptr Student K • 9 Net Local Capital Cost Per Sq. H. Grades Portable Classrooms per 100 students Cost Per Portable Classroom Maximum Gross Impact Fee Per /lousing Unit Ref ore Credits 0.28 Single Family 0.40 Multifamily 0.04 Mobile Home & Other 0.29 Less Property Tax Credits 0.37 Single Fancily 5902 Multifamily $ 178 Mobile Home do Othtr S195 Less Sales Tar Credits 0.00 Single Family $932 Multifamily 5704 Mobile Home & Other 5944 Marimum Cross Ito act Fee per Homing Unit ISingle Family $3,5n. Muhifamtly $1.164 MOW)c Home A Other S3,7141 Elcmcntary/Mid4tc Nigh All K • 9 9 -12 Grades Standards: 0.28 0.11 0.40 0.11 0.04 0.15 0.29 0.09 0.37 113 163 $102.19 5107.19 0.39 0.00 $30.000 $30,000 $3,313 $1,945 $5,263 SI,270 $776 52,046 $3,339 $1,513 $4,9.53 11 Page 3 Tischler g ASSOCialeS, MC • 12 The Impacts of Recent Residential Development on Public Schools in North Carolina Urban Areas: Single - Family Housing Compared to Apartments prepared by: Dr. Emil E. Malizia Department of City and Regional Planning The University of North Carolina at Chapel Hill May, 1997 13 May 1, 1997 Prepared by: Emil E. Malizia AANC Report: The Impacts of Recent Residential Development on-Public Schools in North Carolina Urban Areas: Single - Family Housing Compared to Apartments Background In May 1996, the Principal Investigator (PI) completed a survey of the academic literature and trade publications pertaining to development impact fees for the Apartment Association of North Carolina (AANC). The specific concern was to identify empirical estimates of development impacts by housing type -- apartments compared to single - family residences. The review supported the idea that apartments generally have lesser impacts compared to single - family housing primarily because the former have fewer persons per household and fewer children per household. Survey Work In September 1996, the PI agreed to conduct a random sample survey of housing units in the five urban areas of North Carolina — Asheville, Charlotte, Piedmont Triad (Greensboro and Winston - Salem), Research Triangle (Raleigh, Durham, Cary and Chapel Hill), and Wilmington. The survey was designed to estimate the differences between the household characteristics of apartment dwellers compared to the occupants of single- family housing. The sample size was supposed to be large enough to generalize the results to all metropolitan areas of North Carolina. The survey was specifically intended to determine the number of children being sent to public schools by households living in recently built apartments and single - family dwellings. The questions pertained to household size; number, age and grade level of children; public, private or home schooling; tenure of the household in the dwelling, county, urban area or state; and housing size, value and age. Exhibit 1 is the survey questionnaire. In October 1996, staff at the Center for Urban and Regional Studies conducted a telephone survey of 630 households in the five North Carolina urban area's in order to generate the 14 information needed to test for statistically significant differences between occupant characteristics of recently constructed apartments and single - family housing. The minimum sample size was determined by reviewing previous apartment surveys. The amount of variation in household size and in number of children were used to estimate sample size. The PI determined that a sample of 150. apartment units and 150 houses would be sufficient to generate statistically significant difference -of -means tests for North Carolina urban areas. The goal of having results that could be generalized to North Carolina urban areas ideally required a listing of all recent construction in all metropolitan counties. Of course, this ideal sample frame does not exist. Instead, expert sources were used to identify recent apartment complexes and subdivision locations in the five North Carolina metropolitan areas surveyed. AANC identified 19 apartment complexes constructed since 1990 in these areas. Local planners, economic developers and real estate brokers were contacted for information on recent residential subdivisions and on additional apartment complexes. This process essentially randomized the selection of street locations in each metropolitan area. The sample was drawn from street address and phone number listings derived from Select Phone Deluxe — software that includes a set of 6 CD -ROMS and essentially compiles the telephone listings available for every area of the U.S. The software was satisfactory for the task although it was not entirely current nor error free. It was superior to the available R.L. Polk & Company's City Directories, which were outdated and not available for all areas. AANC also provided telephone number listings from member tenant lists for the apartment complexes. A sample of households was randomly drawn from these listings. An equal number of apartment units and houses was drawn. The number from each urban area roughly corresponded to that area's relative size. The range was from 69 in Wilmington to 118 in Charlotte. Thus, the sample could be described as a cluster random sample. However, the sample size is far too small to generate any reliable statistics for each area. Instead, the PI compiled information from the 1990 Census to generate the demographic profiles for each metropolitan area in North Carolina and has forwarded this "PUMS" data to AANC. From the sample of 630 households, responses were received from 465 households. The response,rate of 74% was very high 15 primarily due to follow -up telephone calls made by the interviewers and project managers. Of these, 455 surveys were ultimately found to be usable in the analysis. Exhibit 2 is the disposition sheet used to track each survey. The completed telephone interviews were checked, and the information was carefully coded in order to stress quality control. The PI also examined every survey for coding errors and double checked questionable entries. Therefore, the information on which the analysis is.based is highly reliable. The data were compiled in a SAS data set and analyzed on a univariate and multivariate basis. The univariate analysis uncovered some questionable values in the data set, and again the original survey information was reviewed and corrected as appropriate to eliminate these discrepancies. Results The variable names and summary statistics are in Exhibit 3. The statistics were run on all variables for convenience although they are not meaningful for some variables such as the first four listed. The first 25 listed variables, through Interviewer Number, come from the compiled surveys. The next eight unlabeled fields convert the residential tenure variables from year -month values to decimal values in years. This conversion is accomplished by dividing "months of residence" by 12 to create a decimal value and then adding this amount to "years of residence." For example, the values for YRSRES and MNTRES are 3.24 years and 2.07 months, respectively. The decimal value, TOTRES, is 3.41 years which equals 3.24 years plus another 2.07/12 years. Approximate monthly rent was assigned to each apartment unit on the basis of its urban -area location and number of bedrooms. Assigning approximate rent level was preferable to asking respondents about rent they paid because survey questions about monetary values always drive down response rates to that question and to others as well. For example, 218 owner- occupants indicated the year their residence was built, but only 169 of them indicated its approximate value. The last two variables listed were computed from the data set. The number of adults is the difference between number of persons in the unit and number of children aged 18 or younger. The number of children in pre- school, private or home school is the difference between number of children and the number in public school in one of three grade levels. 16' The most important task was to produce an exhibit similar to the one provided by the Illinois School Consulting Service. This information is the population per dwelling unit by age group for all units, for single - family, for apartment units, and for units of different size. The PI had to complete a preliminary analysis to determine the number of owner - occupied dwellings to include in the analysis. Ideally, he wanted to include all units. However, the sample frame for single - family housing was not sufficiently precise to capture only recently built homes. In fact, 37 units surveyed were built before 1985 and 19 before 1975. The results on number of persons and children in the unit were higher for all units than for more recently-built housing, but the differences were very small. This result does support the idea that, on average, household size tends to increase slightly with years of tenure in the house. Difference -of -means tests were run comparing housing with two, three and four bedrooms. The analysis showed that no statistically significant differences exist between all single- family units, the subsample of houses built after 1985 and the subsample of houses build after 1975 for units with the same number of bedrooms. Thus, the PI was able to retain all units in the analysis on the basis of this analysis. The analysis pertains to 216 apartments and 239 single - family units. In fact, eight single- family units are mobile homes (2), condos ( 5 ) or townhouses (1) . The rest are single - family detached units. The PI analyzed these eight records individually and decided to leave them in the sample because their demographic characteristics matched the single - family detached and owner - occupied unit characteristics quite well. Of course, these units tended to be smaller and less valuable than detached single - family housing. All apartment units were rented while seven single - family units were rented. The PI compiled the table of average generation rates from separate analyses of sample subsets. The results are shown in Exhibits 4 and 5. The nine rows give housing type and are listed below with the number of observations (n) for that category: 1) all units (n = 455), 2) all single - family units (n = 239), 3) single - family units with two bedrooms (n = 15 which includes one one - bedroom house), 4) single- family units with three bedrooms (n = 126 which" includes three units not reporting number of bedrooms), 5) single- family units with four (n = 78), five (n-,= 18) or six bedrooms (n = k 2), 6) all apartment units (n = 216), 7) apartment units with one bedroom (n = 50), 8) apartment units with two bedrooms (n = 117) and 9) apartments with three bedrooms (n = 49) . The eight•coluinns distribute the total number of persons per dwelling unit by age and school status. Pre - school children per unit listed in the first column are not old enough to attend public schools. In the next three columns, children in public schools per unit are assigned by grade, either to elementary school, junior high school or`high school. School -age children attending private schools or receiving schooling at home per unit are listed in the fifth column. The sixth column lists the number of children 18 years old or younger per dwelling unit. Adults per unit, which includes 18 year -olds who are not attending high school, usually because they have begun college, are in the seventh column. The total number of persons per dwelling unit is shown in the final column. Except for rounding errors, the rates are additive. The rates for public - school attendees by grade give the rate of public school attendance. The sum of this rate and the pre- school plus private /home school rates is the average number of children in the units. This rate added to the rate for number of adults gives the number of persons per unit. Exhibit 4 presents the average rates for each cell, which represent the best point estimates for the sample. Exhibit 5 contains the estimates of standard error, which can be used to calculate confidence intervals for each average rate and to test for the significance of differences between single - family and apartment units. The results in Exhibit 4 generally confirm our expectations on the basis of previous surveys. Single- family houses definitely have more persons per unit and more children per unit than apartments. Moreover, these differences are highly statistically significant (t statistic for persons per unit = 7.37, for children per unit = 6.21). The rates for single - family houses are higher than the apartment rates for every category. Therefore, one can say that recent apartment development in urban areas of North Carolina generates less demand for public education and for other demographically - driven public services than single - family housing developed in these urban areas. One major result from the survey is that all apartment units in the sample generate 0.435 persons 18 years old or less compared to 0.975 children from single - family houses. 18 The results for units by number of bedrooms are interesting. As expected, the rates for apartments with one bedroom, the smallest dwelling units, are the lowest while the rates for houses with four or more bedrooms are the highest. The overall difference is about one adult and one child more living in single - family houses with four or more bedrooms compared to one - bedroom apartments. On the other hand, the rates for two- and three - bedroom apartments compared to two- and three - bedroom houses are quite similar. Two - bedroom apartments appear to generate more population and school -aged children than two - bedroom houses. However, these differences. are not statistically significant. For example, the t statistics comparing number of persons per dwelling unit and number of children per dwelling unit in two - bedroom apartments and in two - bedroom houses are 1.00 and 0.30, respectively. Significant t statistics have values of about 2.00 or more. The average values for three - bedroom apartments are higher than the values for three - bedroom houses and usually lower than the values for houses with four or more bedrooms. The statistical analysis indicates that the former differences are significant while the latter differences are not. For example, three - bedroom apartments generate greater impacts than three bedroom houses for persons per unit and children per unit. The t statistics are 3.35 and 2.13, respectively. On the other hand the comparable t statistics for three - bedroom apartments compared to houses with four bedrooms or more are 0.83 for persons per unit and 1.35 for number of children per unit. Also, three - bedroom apartments and houses with four or more bedrooms have the same impact on the public schools. (The average rates for grades K -12 are 0.755 for both.) Although the survey collected no information on income, households in three - bedroom apartments are probably less affluent than households in three - bedroom or four - bedroom houses. The results for apartments compared to single - family houses are consistent with these detailed comparisons on the basis of number of bedrooms because of the mix of dwelling units found in most recent developments. Apartment complexes are dominated by one - bedroom and two - bedroom units, which are developed much more frequently than three - bedroom units. On the other hand, new subdivisions rarely contain two - bedroom dwellings. Residential development is dominated by houses with three bedrooms or more. Thus, the weighted average impact of apartment complexes is significantly less than the impact of the same number of single - family dwelling units. CONCLUSIONS 19 The survey achieved its intended purpose. Apartment units that have been built in North Carolina urban areas since 1990 house significantly fewer persons per unit and fewer children per unit than single - family houses. As long as apartment developers and managers own complexes with predominantly one- or two - bedroom units, they can argue, with confidence, that their complexes are generating fewer persons per unit and children per unit than the same number of recently built, detached single - family houses with three bedrooms or more. Three - bedroom apartment units have demographic impacts that are greater than the impacts of three - bedroom houses. Households in three - bedroom apartments are living at higher densities and are more likely to send children to public schools. This information should be useful to AANC as they discuss the application of impact fees and other forms of development exactions with local and state officials. EXHIBITS 20 - EXHIBIT 1 Apartment Association of North Carolina Survey EXHIBIT 2 Disposition Sheet EXHIBIT 3 Variables, Labels and Descriptive Statistics EXHIBIT 4 Average Generation Rates by Housing Type EXHIBIT 5 Standard Errors of Generation, Rates by Housing Type (10/3) Apartment Association of North Carolina Survey ID. No. Urban Area County Apart= l/House =2 (01 -03) (04 -04) (05 -06) (07 -07) Hello, my name is . I'm calling from the University of North Carolina at Chapel Hill, Center for Urban and Regional Studies. This month we're calling people in different North Carolina cities to obtain information that school planners can use to better predict the number of students. The questions will take less than 5 minutes to complete, and they deal with the size of your home and number of people in your family. We have only a phone number and we will not record any names, so all answers are anonymous. If I have your permission, let me begin by asking... 1. How many people are in your household including yourself? (How many people live in your household 50% or more of the time ?) (10 -11) 2. Do you have any children aged 18 or younger in your household? Yes..................................... ..............................1 No............................. ...... ..............................2 Refused................................. ..............................9 (12 -12) If YES: For each child, could you tell me his or her age; whether he or she attends a public school, a private school or is home schooled; and what school grade he or she is in? Let's start with the youngest child -- what is his or her age? School grade? Type of school (public, private, home)? ASK FOR EACH CHILD. If NO, go to Question 3. Code after the completion of the interview: Number of children aged 18 or younger (13 -13) 21 Age; Grade Public Private Ho' lme: School, , ;:School > :School>:.' Child #1 1 2 3 Child #2 1 2 3 Child 43 1 2 3 Child #4 1 2 3 If NO, go to Question 3. Code after the completion of the interview: Number of children aged 18 or younger (13 -13) 21 22 Number of children in public elementary school (grades K -5) r (14 -14) Number of children in public middle school (grades 6 -8) (15 -15) Number of children in public high school (grades 9 -12) (16 -16) 3. Do you own or rent your home? Own.................................... ......................:.......1 Rent.................................... ..............................2 Other.................................... ..............................3 Refused................................. ..............................9 (17 -17) 4. How many bedrooms do you have in your home? (18 -18) 5. How many bathrooms do you have in your home? (There can be half - baths, so can have the values 1.0, 1.5, 2.0, 2.5, etc) (19 -21) 6. How long have you -lived in your home? years months (Don't know =88, refused =99) (22 -23) (24 -25) tom. 7. How long have you lived in county name? years months (Don't know =88, refused =99) (26 -27) (28 -29) 8. How long have in you lived in the urban area name area? years months (Don't know =88, refused =99) (30 -31) (32 -33) 9. How long have you lived in North Carolina? years months (Don't know =88, refused--99) 1 (34 -35) (36 -37) (Homeowners only) 10. In what year was your home built? (Don't know--8888, refused =9999) (38 -41) 11. Can you give the approximate square footage- of the heated area of your home? (Don't know --8888, refused =9999) (42 -45) (Note: Most single - family, 3 bedroom houses in North Carolina are from 2,000 - 2,500 sq. fQ - 12. What is the approximate value of your home? (Don't know= 888888, refused= 999999) (46 -51) Thank you for your help. Interviewer Number (52 -53) W W DISPOSITION SHEET Urban Area: County: Apartment House (circle one) Telephone No. Date Called Day Called Time Called Disposition* Interviewer (name /no.) 1. 2. 3. 4. 5. 6. 7. 8. * 1 = completed interview 2 = refused 3 = no answer 4 = answering machine -- note number of rings before switches to message 5 = nonworking number 6 = inconvenient time / call back / appointment: Appointment Day Date Time Name to ask for 23 numbed by numkids _ .06 Tuesday, January. 21, 1997 55 .•.r• Variable Label ' ''� N Mean Std Dev - ----- ---------------------`------------------------------------------ -------- - - - - -- ID URBANAR COUNTY APTHOUS NUMHOUS KIDS18 NUMKIDS KIDSK5 KIDS68 KIDS912 OWNRENT NUMBED NUMBATH YRSRES MNTRES YRSCOUN MNTCOUN YRSURBA MNTURBA YRSNC MNTNC YRBUILT SOF00T VALUE INTVNUM PERMNRES TOTRES PERCNRES TOTCORES PERURRES TOTURRES PERMNCRE TOTNCRES RENT NONPUBPR ADULTS ID. No. Urban Area County Apartment or House Number in Household Presence of Children 18 or Younger Number of Childrem 18 or Younger Number Public School grades K -5 Number in Public grades 6 -8 Number in Public grades 9 -12 Own or Rent Home Number of Bedrooms in Household Number of Bathrooms in Household Years Lived in Home Months Lived in Home Years of Residence in County Months of Residence in County Years of Residence in Urban Area Months of Residence in Urban Area Years of Residence in North Carolina Months of Residence in North Carolina Year House was Built Approximate Square Footage of Home Approximate Value of Home Interviewer Number approximate monthly rent no. children in private /home school /pres number of adults in household 455 476.8483516 454 3.3920705 453 4.3509934 455 1.5252747 454 2.6585903 455 1.5912088 455 0.7186813 455 0.2373626 455 0.0879121 455 0.0879121 454 1.4911894 452 2.8141593 448 2.1584821 451 3.2394678 451 2.0687361 450 9.0688889 450 1.4311111 450 10.0555556 450 1.3311111 449 15.6481069 449 1.0534521 218 1987.78 198 2326.97 169 173437.28 448 3.8370536 451 0.1723947 451 3.4118625 450 0.1192593 450 9.1881481 450 0.1109259 450 10.1664815 449 0.0877877 449 15.7358946 216 729.3055556 455 0.3054945 454 1.9383260 285.2359968 2.0067279 7.5728310 0.4999104 1.3055168 0.5054000 0.9731995 0.5440200 0.3059020 0.3130197 0.5004739 0.7656561 0.7444767 4.7995688 3.0378608 11.8152882 2.6370516 12.6656165 2.5890228 16.5560736, 2.4214062 9.9377943 933.5140635 92221.11 1.5766082 0.2531551 4.7315896 0.2197543 11.7436413 0.2157519 12.5934044 0.2017838 16.4869097 126.0490092 0.6270978 0.7282886 N .A numbed by numkids'' _ .06 Tuesday, January 21, 1997 56 • Wtriable Cabel Minimum Maximum , ---------------------------------------------------------------------='-------- ID ID. No. 1.0000000 948.0000000 URBANAR Urban Area 1.0000000 7.0000000 COUNTY County 1.0000000 92.0000000 APTHOUS Apartment or House 1.0000000 2.0000000 NUMHOUS Number in Household 1.0000000 7.0000000 KIDS18 Presence of Children 18 or Younger 0 2.0000000 NUMKIDS Number of Childrem 18 or Younger 0 3.0000000 KIDSK5 Number Public School grades K -5 0 3.0000000 KIDS68 Number in Public grades 6 -8 0 2.0000000 KIDS912 Number in Public grades 9 -12 0 2.0000000 OWNRENT Own or Rent Home 1.0000000 2.0000000 NUMBED Number of Bedrooms in Household 2.0000000 4.0000000 NUMBATH Number of Bathrooms in Household 1.0000000 6.0000000 YRSRES Years Lived in Home 0 45.0000000 MNTRES Months Lived in Home 0 11.0000000 YRSCOUN Years of Residence in County 0 66.0000000 MNTCOUN Months of Residence in County 0 11.0000000 YRSURBA Years of Residence in Urban Area 0 74.0000000 MNTURBA Months of Residence in Urban Area 0 11.0000000 YRSNC Years of Residence in North Carolina 0 74.0000000 MNTNC Months of Residence in North Carolina 0 11.0000000 YRBUILT Year House was Built 1916.00 1996.00 SQFOOT , Approximate Square Footage of Home 900.0000000 6000.00 VALUE Approximate Value of Home 6000.00 700000.00 INTVNUM Interviewer Number 2.0000000 9.0000000 PERMNRES 0 0.9166667 TOTRES 0.0833333 45.0000000 PERCNRES 0 0.9166667 TOTCORES 0.0833333 66.0000000 PERURRES 0 0.9166667 TOTURRES 0.0833333 74.0000000 PERMNCRE 0 0.9166667 TOTNCRES 0.1666667 74.0000000 RENT approximate monthly rent 500.0000000 950.0000000 NONPUBPR ho. children in private /home school /pres 0 3.0000000 ADULTS number of adults in household 1.0000000 ,v 7.0000000 "' Exhibit 4 Population, Age - Cohorts and Schooling Status by Housing Type Average Generation Rates per Unit Type of Unit Pre - School (0-4 yrs.) Elementary Grades K -5 Jr. High Grades 6 -8 High School Grades 9 -12 Private or Home School Children (18 yrs or less) Adults Persons per dwelling unit All Units 0.2102 0.2374 0.0879 0.0879 0.0953 0.7187 1.9383 2.6586 Single Family 0.3002 0.3264 0.0921 0.1130 0.1432 0.9749 2.0840 3.0630 Two BR or less 0.2000 0.0667 0.0000 0.0000 (2) 0.2667 1.4667 1.7333 Three BR 0.3333 0.2857 0.0556 0.0714 (2) 0.7460 2.0320 2.7840 Four BR or more 0.6224 0.4184 0.1531 0.1837 (2) 1.3776 2.2449 3.6224 Apartments 0.1106 0.1389 0.0833 0.0602 0.0422 0.4352 1.7778 2.2130 One BR 0.0000 0.0200 0.0000 0.0000 (2) 0.0200 1.3400 1.3600 Two BR 0.1282 0.1026 0.0598 0.0342 (2) .0.3248 1.7350 2.0598 Three BR 0.3673 0.3469 0.2245 1 0.1837 1 (2) 1.1224 1 2.3265 3.4490 (1) Average generation rates for Elementary, Jr. High and High School pertain to public schools only. (2) Included with Pre - school children P Exhibit 5 Population, Age- Cohorts and Schooling Status by Housing Type Standard Errors for Average Generation Rates per Unit Type of Unit Pre - School (0 -4 yrs.) (1) Elementary Grades K -5 Jr. High Grades 6 -8 High School Grades 9 -12 Children (18 yrs or less) Adults Persons per dwelling unit All Units 0.029 0.026 0.014 0.015 0.046 0.034 0.061 Single Family 0.047 0.040 0.020 0.024 0.067 0.049 0.085 Two BR or less 0.145 0.067 0.000 0.000 0.182 0.165 0.316 Three BR 0.055 0.052 0.021 0.023 0.083 0.060 0.106 Four BR or more 0.083 0.071 0.039 0.049 0.106 0.083 0.123 Apartments 0.032 0.029 0.021 0.016 0.056 0.049 0.079 One BR 0.000 0.020 0.000 0.000 0.020 0.068 0.074 Two BIV 0.039 0.035 0.025 0.017 0.063 0.054 .0.083 Three BR 0.095 0.085 0.067 0.056 1 0.156 0.089 0.168 (1) Children in private or home schools included with pre - school children. N V