EconC175_California+Demographics+and+Support+for+Public+Education

EconC175_California+Demographics+and+Support+for+Public+Education

Info iconThis preview shows pages 1–20. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 20
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: fLKST 'DKMLT California Demographics and Support for Public Education Final Project for Economics C175 Allison Day Allison Day California, once able to offer all of its residents high quality public education and fiill tuition at its public universities, is now struggling to offer an adequate primary and secondary education to its students in the face of a growing budget crisis. In my discussions with students and other California residents about the source of California’s problems financing public education, people quickly assign blame to one of three sources: the I strain of a growing number of students, the lack of support for education from the elderly, and the fiscal burden imposed by the growing Hispanic immigrant population. If these conjectures regarding the relationship between demographics and education fimding are true, the current crisis is unlikely to pass as the number of students, elderly, and Hispanic immigrants continues to grow. In this paper, I will analyze variations in political support and spending on education across California’s demographically diverse counties in order to test these claims and provide evidence for the potential relationships between California’s demographics and fimding for education. In my analysis, I will represent support for education with two variables. First, I will use the California Secretary of State’s records to construct a variable measuring the proportion of education bond and parcel tax measures passed in each county between 1996 and 2006.1 Though education bonds may be placed on the ballot by local government bodies rather than the electorate itself, bond measures must be approved by a majority vote, so that the bond’s passage represents public support even though the choice to place measures on the ballot may not represent public opinion. This variable isolates political support for education from other determinants of school finance and when examined over a ten year span removes some potential bias due to the sporadic nature of such votes. However, this variable does not take into account whether such initiatives are rejected for reasons beyond political support, such as preexisting levels of spending or the availability of alternative fundraising methods. Second, I will use the RAND Corporation’s California Education Statistics measure of annual per pupil spending by county in 2006 to represent both the financial constraints on existing resources, as per pupil spending decreases as existing revenues are divided between more expenditures, and ' California Elections Data Archive. “Vote Totals, Elections Outcomes and Text for School District Ballot Measures.” California Secretary of State, http://www.sos.ca.gov/elections/county-city-school-district—election-results.htm (accessed April 1, 2010). Allison Day political support for education, as people choose to live in a town and pay existing property taxes in part based on the value they place on existing levels of per pupil spending.2 This variable may better represent support for education than the proportion of ballot measures passed in that it removes the issues of preexisting funding and alternative funding sources, but is weakened by the number of factors beyond political support that can influence per pupil spending such as financial constraints, redistribution from other counties, and federal and state funding. In order to test the common belief that demographics influence support for education, I will compare variations in these measures of support to variations in three demographic variables measured by county from the RAND Corporation’s California Population and Demographic Statistics database: the percent of the population under 14 in 2006, representing the size of population legally required to attend school; the percent of the population over 65 in 2006, representing the size of the elderly population; and the percent of the population under 14 that is of Hispanic origin in 2006, approximately representing the size of California’s largest immigrant population.3 To compare these variations I will create a scatterplot of the measure of support and demographic measure for each of California’s 58 counties, calculate the correlation coefficient to determine the direction of the relationship, and calculate the R2 value to determine the amount of variation in the data that is captured in the pr0posed relationship.4 One important limitation of this analysis is that there are a number of factors that could influence both the demographic composition and support for education in a county. Income is potentially the most important underlying influence on both demographics and support for education, as certain groups may be more likely to settle in high income areas that choose to spend more on education. To attempt to control for income’s influence on my variables, I will include the 2000 Census data on median family income for each county in my analysis. Median family income shows only a small negative correlation with each demographic variable, with 2RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http://ca.rand.org/stats/educationlperpupil.html (accessed April I, 2010). 3 RAND California Population and Demographic Statistics. “Bridged-race Postcensal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http://ca.rand.org/stats/popdemo/popraceage.html (accessed April 1, 2010). 4 For summary statistics of each variable and discussion of outliers, please see Figure l. Allison Day R2 values of 1% or less for each relationship.5 Though these relationships are weak, I will include an analysis of income against each measure of support to indicate the extent to which income is playing a role in determining both measures of support and demographics. When comparing the proportion of ballot measures passed to the four independent variables, the only relatively strong relationship was between measures passed and median family income. Figure 2, which plots median family income against the percent of measures passed, shows a relatively strong positive correlation between income and measures passed and a R2 value of 0.124797, indicating that roughly 12% of the variation in the proportion of measures passed can be accounted for by variations in median family income. Though this is a relatively small R2, it is far greater than the R2 between measures passed and the demographic variables, indicating that though these demographic factors may influence political support, financial constraints could play a more important role in determining political support for increases in taxation.6 This constraint on political support makes sense, as voters must consider not only their political preferences but also whether they can afford to pay more taxes when voting. When comparing armual per pupil spending to the four independent variables, the relationship between each of the demographic variables and aimual per pupil spending is stronger than the same relationship for median family income. Figures 3 through 6 show each of these relationships and make clears the relative strength of the relationships between spending and the demographic variables when compared to the relationship between spending and income. Though control of education finance has somewhat shifted from the local to the state level due to school finance equalization reform, there are still a number of ways in which local preferences influence fimding for education. School districts receive roughly 45% of their revenue fiom local property taxes and annual per pupil spending still varies widely between districts.7 Given the largely local nature of education finance, changes in local demographics can impact education funding for education in two 5 National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census).” National Center for Education Statistics CES Common Core of Data, http://nces.ed.gov/ccd/bat/ (accessed April 1, 2010). R2 with percent under 14:0.00851, with percent over 65=0.01141, and with percent under 14, Hispanic=0.012014 6R2 for population under 14, the percent of the population over 65, and the percent of people under 14 of Hispanic origin, are 0.014098, 0.012456, and 0.005713 respectively. 7 Jonathan Gruber, Public Finance and Public Policy, 3rd Edition (New York, NY: Worth Publishers, 2010), 282. Allison Day ways: first, by imposing strain on existing local resources, and second, by changing the level of political support for funding public education. These two mechanisms will provide potential explanations for the relationships observed between annual per pupil spending and the demographic variables. Figure 3 shows a negative correlation between the percent of the population under 14 and annual per pupil spending with a fairly strong R2 equal to 0.152773 after two outliers with extremely small populations under 14 are removed.8 This negative relationship indicates that the common conjecture regarding the financial strain created by a large number of students competing for resources potentially outweighs the additional political support of a large population of parents. Though this is data is in no way conclusive, it indicates that there could be some merit to this inference. Figure 4 shows a positive correlation between the percent of the population over 65 and annual per pupil spending and an R2 of 0.206023, the strongest R2 examined in this paper, indicating that roughly 1/5 of the variation in annual per pupil spending is explained by variation in the percent of the population over 65. This positive relationship and the relatively high R2 provides evidence against the common opinion that large elderly populations place financial strain on finding for education and provides evidence for a positive correlation between levels of political support for education and the size of the elderly population. This political support may arise from a sense of community cohesion and a desire to provide an intergenerational loan to younger members of the community to support education.9 However, the elderly might be attracted to areas that provide high levels of support to education, possibly through a suburban effect, whereby the elderly are drawn to neighborhoods that promote family values and therefore support spending on public education. Another consideration is that the elderly may be drawn to wealthy areas that have more money to give to education, but given the relatively low correlations between income and the elderly population, it does not appear that this is 8 Kings andSolano County are removed with 3% of the population under 14. 9 For more information on intergenerational loans and political support for education, see Claudia Goldin and Lawrence F. Katz. (1999). Human Capital and Social Capital: The Rise of Secondary Schooling in America. 1910-1940. Journal of Interdisciplinary History 29(4): 692. Allison Day the case. Though this data is not conclusive, the positive direction of the relationship calls into question the common conjecture that the elderly have a negative impact on education finance. Figure 6 shows a slight negative correlation between the percent of the population under 14 that is of Hispanic origin and annual per pupil spending and a relatively weak R2 of 0.089699. The negative relationship could result from the additional financial burden of a high number of low income immigrants, the high cost of English language education, or a decline in political support for education as a relatively high immigrant population weakens community cohesion.10 However, this measure’s relative weakness compared to the negative relationship between the total population under 14 seems to indicate that the immigrant population’s influence on political support for education is small and provides suggestive evidence against the common claim that the children of Hispanic immigrants are a greater burden on the education system than the children of natives. Though this analysis is in no way conclusive, given-the inability of correlation to indicate causation, the relative weakness of the R2 measures reported, and the potential influence of other factors such as federal fimding and local industry composition and the expected returns to education, this information provides some basis for further analysis. If the relationships posited in my analysis are correct, California could be the perfect storm of demographic factors potentially constraining the state’s ability to find education, with California ranking 9th with regards to the proportion of the population under 18, 45th with regards to the proportion of the population over 65, and 2nd with regards to the proportion of the population that is of Hispanic origin.ll The importance of these issues for California’s future merits further study, perhaps in the form of an analysis of the changes in support for education within districts over time as demographics within that district change, providing a greater indication of causality through the timing of these changes. '0 For more information on these impacts see Ronald Lee, “Immigration: Consequnces of Fiscal Developments in the Receiving Population,” in International Encyclopedia of the Social & Behavioral Sciences, ed. David Sills (Free Press, 1968), 7217-7220, and Goldin and Katz, 695. ” US. Census Bureau Statistical Abstract 2007. “The 2010 Statistical Abstract: State Ranking.” US. Census Bureau, http://www.censusgov/compendia/statab/rankings.html (accessed April 1, 2010). Fi » ure 1: Summa Statistics for California’s 58 Counties _mwlmm Median Famil Income] $34 ,101 $49,635 $49,653 $88,934 Percent Ballot Measures Passed 0% 59% 100% Annual Per Pu il S - endin 3 $6,171 $8,894 $8389 $22,492 Po ulation Under 14" 3% 19% 19% 27% Population Over 655 1% 13% 12% 22% P0 ulation Under 14, His anic Ori in‘5 10% 37% 33% 85% Sources: See footnotes or graphs below for data sources. lYuba County has the lowest and Marin County has the highest, with no clear outliers. 2 Alpine, Del Norte, and Sierra County have passed 0% of the ballot measures as a result of having 1 or no ballot measures voted on in the past 10 years. Even though these could be outliers, having no ballot initiatives considered appears significant enough to be included. Inyo, Lake, Plumes, and San Francisco each passed 100% of ballot measures and given that these counties considered a number of initiatives they seem reasonable to include as well. 3Alpine County has the highest per pupil spending, far higher than the next highest, lnyo County at $12,392. Alpine County appears to be an outlier. Alpine County does not stand out in terms of any of the other variables, but perhaps the high per pupil spending is a result of the sparsc population and high busing costs in this remote county. 4 Solano and Kings County have the lowest population under 14, with the next lowest at 13%. These values are unusual and perhaps result from the predominance of the agricultural industry in this area, so that there are a number of migratory adults residing here. Tulare County has the highest proportion, but does not seem to diverge from general figures. 5 Kings County has the lowest with Solano County the next lowest at 9%. Once again, this potential outlier may be the result of the predominance of the agricultural industry in this county. 6 Imperial County has the highest proportion with the Tulare County the next highest at 69% in Tulare. Imperial County is very close to the Mexican border, providing a potential explanation for this potential outlier Figure 2: Median Family Income (2000) vs. Percent Ballot Measures Passed (1996-2006) 100% r . I . I . : I- - ' ‘ f 7 g -_ , R = 0.353266 ;' in . 3 i ' R2 = 0.124797 Percent of Measures Passed (mun-mus) .520, “052:3.00333.0(ka 00333.0(”). 53.000233110333.“(10:530. (ng 00 Median Family Income (2000) Sources: Data on Median Family Income: National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census).” National Center for Education Statistics CES Common Core of Data, http://ncesedgov/ccd/bat/ (accessed April 1, 2010). Data on Ballot Measures: California Elections Data Archive. “Vote Totals, Elections Outcomes and Text for School District Ballot Measures.” California Secretary of State, http:f/www.sos.ca.gov/elections/county-city—school—district-election-results.htm (accessed April 1, 2010). Figure 3: Percent of the Population Under 14 (2006) vs. Annual Per Pupil Spending (2006) by County 513,000.00 7- -— - - 7 7 p 7 . l i E . ‘ n i ‘ Without Outliers: A $12,000.00 —————— —,—-- -- 7" 7 ; 777 R: _0 39036 E i : ' : ‘ i l 2 ' 5 . : n I ‘ ‘ R =0.152773 if, 5“=°°°'°° '" g" " ' """' ""3 l WithOutliers: g - R =0.033756 ; 59,000.00 - ' 157 E $8,000.00 7' :5: $7,000.00 7. i f < 56.00000 7:777:7777 757 7'7? 7 7 $5,000.00 E - - - - ?- - --i -- 0% 5% 10% 15% 20% 25% 30% Percent Population L'nder 1 4 (2006) Note: Darker line represents linear trend when two outliers are removed. Source: Data on Percept Population Under14: RAND California Population and Demographic Statistics. “Bridged- race Postcensal Population Estimates by Race/Ethnicity and Age Group.“ RAND Corporation, http://ca.rand.org/statslpopdemo/popraceage.html (accessed April I, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http://ca.rand.org/stats/cducationfperpupil.html (accessed April 1, 2010). Figure 4: Percent of the Population Over 65 (2006) vs. Annual Per Pupil Spending (2006) by County $13,000.00 77— ——_— .. . _ ,I i, é 7 i . g . R=0.453897 51290090 i" " i 3 f ' ' R2=0.206023 311,000.00 77777 77.. :3 .. l_§__.____ ' 510,000.00 77777777 . 7- $9,000.00 %——--—-i -- 1 $8,000.00 5 77 3“, 57,000.00 . ,7 7 Annual Per l’upll Spending (2006) $6,000.00 -:-— - 77 W. i -7- u, .%_fi $5,000.00 7777 --———--——- -—-------—-1 --—-‘ 0% 5% 10% 15% 20% 25% Percent Population Over 65 (2006) Source: Data on Percept Population Over 65: RAND California Population and Demographic Statistics. “Bridged- race Postcensal Population Estimates by Race/Ethnicity and Age Group." RAND Corporation, http:/fcarand.org/stats/popdemo/popraceage.html (accessed April 1, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http:ffca.rand.org/stats/education/perpupil.html (accessed April 1, 2010). Figure 5: Percent of Population Under 14 of Hispanic Origin (2006) vs. Annual Per Pupil Spending (2006) by County $13,000.00 . a _ . - _ ”‘ 51000000 ml ' E 3 ‘ -R= -0.2995 g -= - i 1 i a : aR2=0.089699 "‘ i i I ;l 5 ; : '3' 511.000.00 "g. g a . a. . at ' ' I i 5 a 1:: 51000000 =‘ $9,000.00 ; — a : "a | : 58.00000 ; _ n” : E $7,000.00 5 E a i 36.00000 -' 35.00000 . . _ 0% 20% 40% 60% 30% 100% Percent Population Under 14 of Hispanic Origin (2006) Source: Data on Percept Population Underl4 of Hispanic Origin: RAND California Population and Demographic Statistics. “Bridged—race Postcensal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http:r'lca.rand.0rg/stats/popdemor‘popraceage.html (accessed April 1, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http:l/ca.rand.org/stats/education/perpupiLhtml (accessed April I, 2010). Figure 6: Median Family Income (2000) vs. Annual Per Pupil Spending (2006) by County 313.000.00 E 7i I”--. ---. , I .- : - - i I a i = 0.07404 : i : 1 1 - = -, a 312900-00 ' ‘i"""‘@ '5 IR‘=0.005482 g l f j 3 ‘ E 9; 31100000 j”*"i'”i!*f** 7;" :I .1 ..-___. at : i Ii i i 1 1 : ; g 51000000 -i————--.. _.-i_.. i.'.'. a .0-.- .:....,-.F i ' E E : : 3 i - é 39.00000 -£——-——-—-i- ———f1- I, gin. ; r We . . g l l i 1, I a: 3 5 1 : E $8,000.00 ; ——_.... . fin - I...'....-..i. n... .. . W“: E 1 lll- I. i 1 . I 2 $7,000.00 : '— ' rirrll~7777 -.r 7 _ < 36,000.00 3 ———-i-- .11-- .5 ___i 5 3.--- _ 35.00000 - 1 2 1 ' : g- ya” 33 s s S s 9: Sc 0 c 0. r):jg?300(}g‘fgor)(lgfi90 0(155%)!)nofipadigoriaodg‘ooagf 30m 0“ Median Famin Income (2 012K!) Source: Data on Median Family Income: National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census)” National Center for Education Statistics CES Common Core of Data, http://ncesedgov/ccdfbatf (accessed April 1, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http:l/ca.rand.org/stats/education/perpupil.html (accessed April l, 2010). References: Goldin, Claudia and Lawrence F. Katz. (1999). Human Capital and Social Capital: The Rise of Secondary Schooling in America. 1910-1940. Journal of Interdisciplinary Histozy 29(4): 692 Gruber, Jonathan. Public Finance and Public Policy, 3rd Edition (New York, NY: Worth Publishers, 2010). '3 Lee, Ronald. “Immigration: Consequnces of Fiscal Developments in the Receiving Population,’ in International Encyclopedia of the Social & Behavioral Sciences, ed. David Sills (Free Press, 1968): 7217—7220. U.S. Census Bureau Statistical Abstract 2007. “The 2010 Statistical Abstract: State Ranking.” U.S. Census Bureau, htlp://www.census.gov/compendia/statab/rankings.html (accessed April 1, 2010). C175 SPRING 2010 ( J rd AUTHOR: TITLE: Allison Day California Demographic and Support for Public Education TOPIC - Clear statement of research question and good motivation for your interest in this topic Good description of your data and choice of variables. Well-cited sources Thoughtful discussion of limitations and potential biases associated with your indicators. It would be helpful to include a bit more background information on how measures are plaCed on the ballot to begin with. What is the process required? Who usually initiates the process? Are there variations across counties? How might this affect your findings? Good use of secondary sources to provide additional evidence and background information. DATA ANALYSIS Clear, detailed explanation of your methodology. Again, good discussion of potential limitations Well done for including a brief explanation of how/why you expect local v. /demographics to affect education finance (p. 4) . It would be informative to include a table of summary statistics to I. provide a richer description of your sample. E.g. average % Hispanics, average % population under 14, average number of measures passed, average population size etc. Your R2 are pretty low which suggests that there are other factors which i' \ 7 are primarily responsible for the variation in education expenditures. Do you have any ideas what some of these factors ight be? I p ’ 2: Are there any interesting outliers in your data’gg. areas with "I L .13: ' ' particularly high education expenditures)? Do you notice anything about 4;} ' .c j these places? Do older people seem to be concentrated in certain areas? / r - \ -‘ Do you have any theories about what could be the main factors - 4506.; g. influencing the distribution of the elderly? f// .’ t. / / ' ‘- 7..}(‘O .41.!- K,__ is in CHARTS/GRAPHS CONCLUSION Well—labeled, clearly presented figures. Good recognition of the difference between causation and correlation. Although you appear to lose sight of this at some points in your discussion earlier in the paper. Thoughtful suggestions of ways in which you could improve your analysis. With regard to the idea of looking at changes in education support and changes in demographics, are there any counties which have experienced particularly dramatic changes in recent times that you could examine? WRITING Overall, an interesting, well written paper. Please proofread for spelling and grammatical errors. GRADE: A- GRADER: .Kenny Ajayi (kajayi@berkeley.edu) i / - 'I ,. A 1 r ‘ r . ‘ (l i \Vfifiimt. (flurry/Lilia e-W‘U'LfiVJ iL/‘LfijlfiEfl/J . him 95w“- Office Hours: Fri 5y April 29m, 10am-12pm in the Peixot‘to Room (611! floor of Evans Hall). Email me by Friday if you n-.. ......L'I.. Lac an...“ an LLnnn hm...— L...._. "-J ....-...Irl III... on "ALA—lulu __ ..__n:..l_...__c L“..— ._....I.L..... 4.1.... gr -' vi -. [A0 of” l 1 l . y,“ ‘ _ 7L California Demographics and Support for Public Education Term Project for Economics C175 EconomiC’S’C‘l’i'S Allison Day Term- Paper'" SI-BT‘186672-2-i3 -. California Demo‘graphics:and:Support~forPublic Education California, once able to offer all of its residents high quality public education and full tuition at its public universities, is now struggling to offer an adequate primary and secondary education to its students and slashing funding for the artsand higher education in the face of growing budget crisis. In my discussions with students and other California residents about the source of California’s problems financing public education, people quickly assign blame to one of three sources: the strain of a growing number of students, the lack of support for education from the elderly, and the fiscal burden imposed by the growing Hispanic immigrant population. If these conjectures regarding the relationship between demographics and education funding are true, the current crisis is unlikely to pass as the number of students, elderly, and Hispanic immigrants continues to grow. In this paper, I will analyze variations in _r___,__————._ political support and spending on education across California’s demographicallyédiverse counties in fin— __ __..._._——ra order to test these claims and provide evidence £01; the potential‘relationfiships between California’s demogrjip_hics and fundin for education. In my analysis, I will represent support for education with two variables. First, I will use the California Secretary of State’s records to construct a variable measuring the Mion of education EndandparceLtaxsmeasures p_a_ssed in each county b_etwe_en___l 979767 zatncgljlflot').l Though this variable .. , ,- ,. ,._\ 1,.\_,..,,).~,.._ ' ,.,,L:.'l J; “Li; . I: «d; gukkrgr - ,}_.\‘L -\,~UL/Ilr_l ic\_ _ _ isolates polltlcal suppbrt foir educatlon and over a ten year span removes some potentIal bias due to the sporadic nature of such votes, this variable does not take into account whether such initiatives are rejected for reasons beyond political support, such as preexisting levels of spending or the availability of alternative fundraising methods. Second, I will use the RAND Corporation’s Califomia Education Statistics measure DignguijflpupiLspepding _by_ county in 20.06 to represent both the financial constraints on existing resources, as per pupil spending decreases as existing revenues are divided ‘ California Elections Data Archive. “Vote Totals, Elections Outcomes and Text for School District Ballot Measures.” California Secretary of State, http://www.sos.ca.gov/elections/county—city—school—district-election—results.htm (accessed April I, 2010). i‘l' ,1-._ .7 .‘d‘lil‘ A. , r ’-r '. ,Iiwf _UL~;I . ‘ --' /t\1_',(.."l 1‘ l" ' ’2'" u‘ L\ i. ”' ' J _-‘ 1 . I . I, f), I ("L f If; I»! LLJ LVJ " - H h Allison Day between more expenditures, and political support for education, as people choose to live in a town and pay existing property taxes in part based on the value they place on existing levels of per pupil spending.2 This variable may better represent support for education than the proportion of ballot measures passed in that it removes the issues of preexisting funding and alternative funding sources, but is weakened by the number of factors beyond political support that can influence per pupil spending, including financial constraints, redistribution from other counties, and federal and state funding. In order to test the common conjectures regarding demographics and support for education f discussed earlier in the paper, I will compare variations in these measures of suppgrt to variations in ___—.._.‘_.._._.,.._____b 7571-“:4 l I',‘ ll!“ . a“. \ llb“ , '7, three demographic variables measuredib- 'Lcioun f'lu'siri RAND Co oration’fliifomiafiopulation 31nd __ Demographic Statistics database: the percent of the population undilél in 2006, appi‘dximately representing the size of population legally required to attend school; the percent of the population over 55 in 2006, representing the size of the elderly population; and the percent of the populatigmundeLEl, ‘ that is of Hispanic origin, approximately representing the size of California’s largest immigrant population impacting school funding.3 To compare these variations I will create a scatterplot of the measure of support and demographic measure for each of Califomia’s 58 counties, calculate the 1 correlation coefficient to determine the direction of the relationship, and calculate the R2 value to :. " N- (‘5‘ _ i _ t} I ‘4 . ip \_ [0' l y . . . . . . . . /lt‘-;‘- determme the amount of vanatron 1n the data that IS captured 1n the proposed relatronship. 5‘: ’: ' _ ;r a , o ' 2‘ R?“ A One important limitation of this analysis is that there are a number of factors that could influence both the demographic composition and support for education in a county, Wimflhfimgflpbwfi _ confiding factorrisjncerHe, as elderly people may be both more likely to have a high income and more likely to settle in wealthier areas. To attempt to control for income’s influence on my variables, I will include the 2000 Census data on median family income for each county, which shows only a small 2RAND California Education Statistics. “Per Pupil Spending." RAND Corporation, http://ca.rand.org/stats/education/perpupil.html (accessed April 1, 2010). 3 RAND California Population and Demographic Statistics. “Bridged-race Postcensal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http://ca.rand.org/stats/popdemo/popraceage.html (accessed April 1, 2010). 7 il- ' l A Allison Day negative correlation and-between median family income and the demographic variable caehémeasur with R2 values of 1% of less for each relationship.4 Though these relationships are weak, I will include an analysis of income against each measure of support to indicate the extent to which income is playing a role in determining both measures of support and demographics. When comparing the proportion of ballot measures passed to my four dependent variables, the only relatively strong relationship was between measures passed and median family income. Figureji- which plots median family income against the percent of measures passed, shows a relatively strong positive correlation between income and measures passediwitili a R2 value of 0124797, indicating that roughly 12% of the variation in the proportion of measures passed can be accounted for by variations in median family income. Though this is a relatively small R2, it is far greater than._th§_1i2_lEM¢n measures passed and the demographic variables, indicating that though these demographic factors may 4)» 3 int influence political support, financial constraints could play an important role in determining political 0L _ £le ‘ support for increasgin taxation.3 This constraint on political support makes sense, as voters must a Milli/4'7; I. , “.i flea" { consider whether they can afford to pay more taxes when makingihis—decision in addition to their own .W fiat l political preferences. When comparing annual per pupil spending to the independent variables, the relationship between each of the demographic variables and annual per pupil spending is stronger than the same relationship for median family income. Figures % through fishow each of these relationships and make clears the relative strength of the relationships between spending and the demographic variables when compared to the relationship between spending and income. Though control of education finance has somewhat shifted from the local to the state level due to school finance equalization reform, there are still a number of ways in which local preferences influence funding for education. School districts 4 National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census).” National Center for Education Statistics CES Common Core of Data, http:/lncesedgov/ccdfbat/ (accessed April 1, 2010). R2 with percent under l4=0.00851, with percent over 65=0.0| I41, and with percent under 14, Hispanic=0.0|2014 5R2 for population under 14, the percent of the population over 65, and the percent of people under 14 of Hispanic origin, are 0.014098, 0.012456, and 0.005713 respectively. 3 Allison Day receive roughly 45% of their revenue from local property taxes and annual per pupil spending still varies widely between districts.6 Given the largely local nature of education finance, changes in local demographics can impact education funding for education in two ways: first, by imposing the strain on LI (u existing local resources, and second, by changing the level of political suppori funding public education. It isithroughtheclocalfundingmodel for education and these two mechanisms that local demographics cantinfluenceedueatiorrfinanceand— it--is’these mechanisms that'will provide potential explanations for the relationships observed between annual per pupil spending and the demographic variables. Figureéshows a negative correlation between the percent of the population under 14 against annual per pupil spending with a fairly strong R2 equal to 0.152773 after two outliers with extremely small populations under 14 are removed.7 This negative relationship indicates that the common conjecture regarding the financial strain created by the greater number of students competing for resources potentially outweighs the additional political support of a large population of parents. Though this is data is in no way conclusive, it indicates that there could be some merit to this inference. Figure Bishows a positive correlation between the percent of the population over 65 against annual per pupil spending and an R2 of 0206023, the strongest R2 examined in this paper, indicating that roughly 1/5 of the variation in annual per pupil spending is explained by variation in the percent of the population over 65. This positive relationship and the relatively high R2 provides evidence against the common opinion that large elderly populations place financial strain on funding for education and Mitt. n. , i .i‘ i‘, i/-’.'.-i;‘ div. L. L'LL|-L1{ Li .41..“ . iridicateskt‘hat the elderly may r ctually have higher levelsoflpoditicfisupPOrt—for—education—than .4— * ‘ 1‘ .‘ . ‘,r_ _" I: I]. ‘ [Jib-‘ll 'r'r LIL; -. ._./ I, l I f. ‘ ,,_,r M13311 commpglyibilieifiécl; This political support may arise through asehse of community'cohesion and a to wt" ftp/“fl HAL desire to provide an intergenerational loan to younger members of the community to support education.8 (’1, .1536 but?! 5. I) Iii/Iv.- LJ’, (rL-k \( l; a (j I,- L {if Li”!!! [. ,1 (w, ‘— l, I, .IE-t f Lid Liz. I f A” , _. ,__ tn; I Inih FM”. Given the relatively low correlations between income and the elderly population, it does not appear i (“i \L. I h‘ ‘= M a ,.‘_~\=. - ‘ ' ‘: l ‘-‘ " , uni/“x I w“ -, [l’ 5% (i ' ' [W (all 1&1]: I [W' 6 Jonaflian Gruber, Public Finance and Public Policy, 3rd Edition (New York, NY: Worth Publishers, 2010), 2823 ‘ if! i- '_ U 4W 94]} 03 7 Kings and Solano County are removed with 3% of the population under 14. 1-5 ” ‘ Niki ‘ til ‘90 8 For more information on intergenerational loans and political support for education, see Claudia Goldin and LawrenceLF. Q 1 3h ‘ Katz. (1999). Human Capital and Social Capital: The Rise of Secondary Schooling in America. 1910-1940. Journal of ‘ witty, Interdisaplinary History 29(4): 692. ‘ -.J_/-J£\_l We" 0' W) z4'1' - Pi“ ..': . 4 {mill 1 Pill.) ‘ '1 “El = " l “\i it?” {Ll' y. Allison Day conclusive that this positive relationship is a result of higher levels of income in relatively old communities and calls into question the common conjecture that the elderly have a negative impact on education finance. Figure 5 shows a slight negative correlation between the percent of the population under 14 that is of Hispanic origin against annual per pupil spending and a relatively weak R2 of 0.089699. The negative relationship could result from the additional financial burden of a high number of low income immigrants families high cost of English language education and a decline in political support for education as a relatively high immigrant population weakens community cohesion.9 However, this measure's relative weakness compared to the negative relationship between the total population under 14 seems to indicate that the immigrant populationh influence on political support for education is small signers and provides vidence against the common claim that the children of Hispanic immigrants are a greater burden on the education system than the children of natives. Though this analysis is in no way conclusive, given the inability of correlation to indicate 1" \ f h causation and the relative weakness of the R2 measures reported, this information provides some basis for further analysis. If the relationships posited in my analysis are correct, California could be the ’ " be 7 ' perfect storm of demographic factors potentially constraining the state’s ability to fund education, with w ‘ 2 California ranking 9”‘ out of 50 states with regards to the proportion of the population under 18, 45th with regards to the proportion of the population over 65, and 2m with regards to the proportion of the population that is of Hispanic origin.10 The importance of these issues and the potential importance of these results for California’s future merits further study, perhaps in the form of an analysis of the changes in support for education within districts over time as demographics within that district change, providing a greater indication of causality through the timing of these changes. 9 For more information on these impacts see Ronald Lee, “Immigration: Consequnces of Fiscal Developments in the Receiving Population,” in International Encyclopedia ofthe Social & Behavioral Sciences, ed. David Sills (Free Press, 1968), 7217—7220, and Goldin and Katz, 695. '0 US. Census Bureau Statistical Abstract 2007. “The 2010 Statistical Abstract: State Ranking.” us. Census Bureau, http://www.census.gov/compendiafstatab/rankingshtml (accessed April 1, 2010). Figure 1: Median Family Income (2000) vs. Percent Ballot Measures Passed (1996-2006) 100% r 3 I 7; n ; _. ‘ I _ a 90w - i i é § -, ' R=0.353266 9 /° .. . 2 ’5 ' 3 3 3' I a g R =0.124797 g 30% ; i . i . __ _ 5 3. 70% '8 a 60% , E 50% — j I— . i 40% =3 = E : a: 30% i —— - —— i — —— u i I ‘ g 30% W .7 7 WW 7 ‘ W, , '- : :J D’ — 10/0 I I I 0% .‘ $20 $30 340 350 .360 S 70 $80 $90 .5 10 .000. “(fungflgugol 00.00000.(mam).(JUO'OUflUQ 00.0w) 00 (1000.00 Median Family Income (2000) Sources: Data on Median Family Income: National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census)" National Center for Education Statistics CES Common Core of Data, http://nces.ed.gov/ccdfbat/ (accessed April I, 2010). Data on Ballot Measures: California Elections Data Archive. “Vote Totals, Elections Outcomes and Text for School District Ballot Measures." California Secretary of State, http:l/www.sos.ca.gov/electionsfcounty—city—schooI—dislrict-election-results.htm (accessed April 1, 2010). Figure 2: Percent of the Population Under 14 (2006) vs. Annual Per Pupil Spending (2006) by County $13,000.00 ; ‘ - p — 1 ‘ i f . 3 u f i Without Outliers: A 512.000.00 { referee": 7 -- -~ - --—- ‘ ——--- R: 4139036 31100000 R2=0-152773 1;; " ' ' ‘ 1 ' ' i ' : With Outliers: E ' ‘ . n . . : _ E .. 7 . l ,l: I .. . .I . .. R2” :1 ‘ - - i R =0.033756 2 59.00000 7 e "2 $8,000.00 — .9 3 $7,000.00 — 5 . . { 36.00000 A } " ‘ ——- ---—--: ———; .__._...___i_ . . . . . i ,ii:,,, , ,7, 0% 5% 10% 15% 20% 25% 30% Percent Population Under 14 (2006) Note: Darker line represents linear trend when two outliers are removed. Source: Data on Percept Population Under14: RAND California Population and Demographic Statistics. “Bridged— race Postcensal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http://ca.rand.org/stats/popdemo/popraceage.html (accessed April 1, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http://ca.rand.org/stats/education/perpupil.html (accessed April 1, 2010). Figure 3: Percent of the Population Over 65 (2006) vs. Annual Per Pupil Spending (2006) by County 513,000.00 f 7777 a 7:22, 2 . ‘ ? : . ‘ _ R=0.453897 A 512,000.00 - 7 -— g. R2=0_206023 «2 y : 3 $11,000.00 -#- at E 510,000.00 9 :9" =‘ $9000.00 ~ ~ 2 E ._ 53,000.00 Po g- $7,000.00 < $6,000.00 3--——-—---1 .‘ . ._!_3 55000.00 J—' 2 2,21 2 in, 0% 5% 10% 15% 20% 25% Percent Population Over 65 (2006) Source: Data on Percept Population Over 65: RAND California Population and Demographic Statistics. “Bridged— race Postcensal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http:f/carand.org/stats/popdemo/popraceage.html (accessed April I, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http://ca.rand.org/stats/education/perpupiLhtml (accessed April 1, 2010). Figure 4: Percent of Population Under 14 of Hispanic Origin (2006) vs. Annual Per Pupil Spending (2006) by County 513,00000 ; fl . W ; 7 a . - 2 I ; .R= 41.2995 ."‘ , .rwiiiinfi 2,7, ,itiifi , i ,7 iii, , g SL900“) . i ; ; ; 5R2=0.089699 2 -.i ' I ' 23‘ $11,000.00 ; WW 3. r- 5 : : I : : . ":5’ $10,000.00 ~71r~~7+i~ { 7-7577 E. . _ 1 s . =‘ 39,000.00 : 3 i -; 38,000.00 2 . E 37,000.00 3 a . a : : E 86,000.00 773+ r “fin j i a 1 1 : ; 55,000.00 1 é ‘ 1 0% 20% 40% 60% 80% 100% Percent Population Under 14 oinspanic Origin (2006) Source: Data on Percept Population Underl4 of Hispanic Origin: RAND California Population and Demographic Statistics. “Bridged-race Postccnsal Population Estimates by Race/Ethnicity and Age Group.” RAND Corporation, http://ca.rand.org/stats/popdemo/popraceage.html (accessed April I, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http:l/ca.rand.org/staLs/education/perpupil.html (accessed April 1, 2010). Figure 5: Median Family Income (2000) vs. Annual Per Pupil Spending (2006) by County ' ' "___1" " '_'E_"' '_"'I_" "" % - 1 I E i 1 5 i 1 R: -0.07404 A W V V Vii-iii 7'" 7' ' ""j’ ""” ' r i W“ R2 = c I i V I i ' 1 g 1 I :i I : j 1 «:3 $11,000.00 —é-————--——.,——a——- - ..._:._.. 1.. .._..___._ u ' i 3 ' E i ' 1 E 1 - I I l 1 g 510,000.00 3 g. .. .....__ I a; - . :‘ 39.000.00 :— $8,000.00 -;—-——— -- ,9 7; 31000.00 - —.—._-... ._.._ 1 < $6,000.00 i 5' ....__ . . 11 $5,000.00 : $30 in in” is 56 97 5‘8 St) Wt .0m5000050000.30000.5min.(51000030000. 30000 451.000 0” Median Family Income (2000) Source: Data on Median Family Income: National Center for Education Statistics Common Core of Data. “Median Family Income (2000 Census)” National Center for Education Statistics CES Common Core of Data, http://nces.ed.gov/ccdfbalf (accessed April 1, 2010). Data on Per Pupil Spending: RAND California Education Statistics. “Per Pupil Spending.” RAND Corporation, http:l/carand.org/stats/education/perpupil.html (accessed April 1,2010). References: Goldin, Claudia and Lawrence F. Katz. (1999). Human Capital and Social Capital: The Rise of Secondary Schooling in America. 1910-1940. Journal of Interdisciplinary History 29(4): 692 Gruber, Jonathan. Public Finance and Public Policy, 3rd Edition (New York, NY: Worth Publishers, 2010). 1 Lee, Ronald. “Immigration: Consequnces of Fiscal Developments in the Receiving Population,’ in International Encyclopedia of the Social & Behavioral Sciences, ed. David Sills (Free Press, 1968): 7217-7220. US. Census Bureau Statistical Abstract 2007. “The 2010 Statistical Abstract: State Ranking.” US. Census Bureau, http://www.census.gov/compendiafstatab/rankings.html (accessed April 1, 2010). ...
View Full Document

This note was uploaded on 02/07/2011 for the course ECON C175 taught by Professor Traeger during the Spring '09 term at University of California, Berkeley.

Page1 / 20

EconC175_California+Demographics+and+Support+for+Public+Education

This preview shows document pages 1 - 20. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online