PRQ09-1-Percival-Representation and Local Policy-Relating County-Level Public Opinion to Policy Outputs

8/3/2019 PRQ09-1-Percival-Representation and Local Policy-Relating County-Level Public Opinion to Policy Outputs

1/15

University of Utah

Representation and Local Policy: Relating County-Level Public Opinion to Policy OutputsAuthor(s): Garrick L. Percival, Martin Johnson, Max NeimanSource: Political Research Quarterly, Vol. 62, No. 1 (Mar., 2009), pp. 164-177Published by: Sage Publications, Inc. on behalf of the University of UtahStable URL: http://www.jstor.org/stable/27759854 .

Accessed: 07/02/2011 20:11

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=sage. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Sage Publications, Inc. and University of Utah are collaborating with JSTOR to digitize, preserve and extend

access to Political Research Quarterly.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=sagehttp://www.jstor.org/action/showPublisher?publisherCode=utahhttp://www.jstor.org/stable/27759854?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=sagehttp://www.jstor.org/action/showPublisher?publisherCode=sagehttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/27759854?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=utahhttp://www.jstor.org/action/showPublisher?publisherCode=sage


2/15

Political Research Quarterly

Representation and Local PolicyVolume 62 Number 1March 2009 164-1772009 Universityof Utah10.1177/1065912908316341

http://prq.sagepub.comRelating County-Level Public Opinion toPolicy Outputs hosted athttp://online.sagepub.comGarrick L. PercivalUniversity ofMinnesota, DuluthMartin JohnsonUniversity of California, Riverside

Max NeimanPublic Policy Institute of California, San Francisco

Students of local politics have arguedAmerican federalism implies littlerole for local tastes inpolicy making. Peterson(1979) anticipates thepursuit of a productive taxbase will depress subnational government spending on social services, while promoting developmental policies. We investigate the role public opinion plays incounty-level redistributive, developmental, and allocational program spending inCalifornia, using a novel measure of county politicalideology.Our findings challenge expectations that local governments are uniformlybiased against redistribution.Wefind that social service spending varies across counties as a function of ideological orientation. In several policy areas,institutional structuremediates the responsiveness of officials.Keywords: policy representation; local politics; political ideology; public policy

Scholarsof comparative state politics have longbeen interested in the extent towhich the public

policies of state governments reflectmass public opinion in the states (Weber and Shaffer 1972; Morehouse1973). Erikson, Wright, andMclver (1993) and othershave established an association between states' generalpolitical attitudes and the choices of state policy makers.We extend the investigation of subnational opinionpolicy linkages by examininghow local politicalideology shapes policy within states, asking towhatextent does localized ideological dispositions influence local-level policymaking? We argue that localideology should have a much greater influence onlocal governments' policy choices than is often presumed. To test this expectation, we follow methodsemployed at the state level to create a reliable and stablemeasure of local ideology by aggregating statewideCalifornia Field Poll surveys (1990-1999) to thecounty level and test the relationship between localpolitical orientations and a broad range of policy outputs at the county level.We find counties' expenditure patterns vary as afunction of ideology across a number of issues; however, the ideology-policy linkage isnot straightforward.Specifically, the impact of county ideological dispositions are more likely to impact redistributive policies?where political conflict is most intense?and the164

strengthof this association is conditioned by differingcounty institutional structures.As awhole, these resultsclarify the circumstances under which local governments will support redistributive policies, despite thestrong inclination for localities not to do so. By contrast,our findings support the view that there is a meaningful local politics, despite the substantial hemming-inof local choice bymandates, constitutions, and resourceconstraints.

Local Policy Making as"Limited Politics"Peterson's (1979, 1981) seminal examination of city

policy making argues local governments' policy choicesare confined by the structural constraints of U.S. federalism, and as a result, local politics is best characterizedas a "limited politics," where local governments' policyagendas are relatively narrow,with an almost exclusiveGarrick L. Percival, Assistant Professor of Political Science,University ofMinnesota, Duluth; e-mail: [email protected] Johnson, Associate Professor of Political Science, Universityof Cahfornia, Riverside; e-mail: [email protected] Neiman, Associate Director, Public Policy Institute ofCalifornia; e-mail: [email protected].


3/15

Percival, Johnson, and Neiman /Representation and Local Policy 165

focus on enacting policies aimed at improving economic growth. In his account, there is littleroom for therepresentation of local interests given the economicimperatives that limit local policy makers.Peterson's (1981) primary argument that localgovernments' main goal is set on delivering economic growth policies has shaped the expectationsmany scholars have about the local policy makingprocess. According to the "city limits" framework,local governments tend to shy away from policiesthat could potentially hurt communities' economicstanding. This works to create an economic and political bias against redistributive policies; those policiesthat redistribute wealth from those who are better offto thosewho areworse off. Instead, local governmentspursue what Peterson classifies as "developmental"(such as investing in an industrial park, subsidizing anew shopping center, or building local schools) and"allocational" (including street sweeping, garbagecollection, community policing, and fire protectionpolicies) that together work to enhance the local taxbase and generate additional resources that can beused to help thewelfare of the city.He assumes thispursuit of economic growth is popular among decision makers, and because of this, there is generallyless political conflict at the local level.Overall, Petersonclaims thatbecause local governments are less likelyto play an active role in areas of policy where political conflict is high, and more likely to adopt andimplement policies where political conflict is low,local political forces like ideology and partisanshipshould have minimal influence on policy making atthe ocal level (1981, 128).In fairness toPeterson, his framework does not preclude having local governments produce policies thatbenefit the lesswell-off at the expense of better-heeledlocal populations. There is ample evidence to suggestlocal institutions and local political circumstancesmatter in shaping policies (Clingermayer and Feiock1995). However, empirical analysis of actual policyoutcomes among cities demonstrates a strong andconsistent inclination to benefit economic and com

mercial interests, and the challenge is to ascertainwhen the imperative of economic development givesway to redistributive policies (Lewis 2001a, 2001b).

Reconsidering the Role of Local PublicOpinion inLocal Policy Decisions

We develop a "county limits" variation on thePeterson theme and investigate the responsiveness ofcounty governments to local circumstances. Historically,

county governments have held a unique position inthe American governmental system, designed asadministrative arms of state governments that couldconveniently and efficiently deliver state programswithin their geographic boundaries (Berman andSalant 1996; Benton 2002a). As a result, theirfunctionalresponsibilities often differed from those of municipalities. From this perspective, as counties areseverely restricted by top-down constraints and trendtoward producing a narrow set of state-imposed allocational and developmental policies, one ought not toexpect a significant explanatory role for such factorsas the ideology of their residents.An alternative view advanced here posits that localideology should have a significant impact on countypolicy processes and policy choices that elected officials make. This flows from an emerging body ofliterature on county government that documentstremendous change in the role and functions of countygovernments in the United States. Growing demandsfrom the bottom up, driven by increasing urbanizationand suburbanization, in addition to top-down forcessuch as a decline in federal revenue sharing to localgovernments in the 1980s and state governments' fiscalplights in the 1990s, has not only caused counties toincrease the range of services theyprovide, but has alsoprovided counties with significantlymore authority anddiscretion (Benton 2002a). As many counties havebecome increasingly urbanized, especially those in theAmerican South andWest, the demand forcity-like services has increased (DeSantis and Renner 1996). Inresponse, thenature of services offered by county governments has grown since the early 1980s frommoretraditional services like property tax assessment, lawenforcement, and elections (Benton and Rigos 1985;Cigler 1990), to other additional services such as healthcare, educational services, pollution control, and masstransit among others (Duncombe 1977; Schneider andPark 1989; Benton 2002a; 2003). Inmany cases, counties have begun to rival or even become the predominant provider ofmany municipal and regional services(Benton 2002a). To carry out increasingly complexfunctions, state governments have tended to increasethe amount of policy discretion and hence decision

making authority to county governments (Bowman andKearney 1986;Martin andNyhan 1994). Ideologicaldivisions often characterize many of these "new" policies being adopted and implemented at the countylevel. These ideological divisions, coupled withincreased county authority and discretion provides asound basis forour central expectation that there is significant "room" for local ideology to shape policy at thecounty level.


4/15

166 PoliticalResearchQuarterlyThe ideological disposition of county residentscould be linked to public policy in several ways.Street-level bureaucrats in counties with distinct ide

ological profiles, for example, might be more likelyor feel compelled to implement policies inways thatclosely reflect prevailing ideological preferences(Lipsky1980).Of course,the staff nd leadership flocal agencies might also be recruited inways thatreflect the views of the local legislature (Board ofCounty Supervisors, School Boards, orCity Councils).In addition, local elected officials (e.g. County Boardsof Supervisors) are likely to try to adopt policies thatreflect or are consistent with the ideological preferences of their constituents. In short,as counties becomeinvolved in such controversial or divisive issues asmanaging growth, implementing controversial programs in such areas as welfare, parolees, homelessness, and health care for the poor, it is plausible toexpect that the ideological inclinations of the localpopulation will play a role. Based on this, our centralexpectation is that local ideological orientations shouldimpact county-level policy outputs.Students of representation have found public opinion to have a strong influence across much of theAmerican political system, including thevoting behavior ofmembers of Congress (Miller and Stokes 1963;Erikson 1978), on nationalpolicy outputs Wlezien1995, 1996; Stimson, MacKuen, and Erikson 1996),and between state policy decisions and general measures of political ideology and more specific publicdemands of state residents (e.g., Erikson, Wright, andMclver 1993; Hill and Hinton-Andersson 1995; Hilland Leighley 1996). In contrast, however, littleworkhas systematically investigated policy responsivenessto ideology at the local level.One major reason the relationship between ideology and public policy remains underexplored bystudents of local politics is the inherent difficulty ofmeasuring local ideological orientations. In particular, it is extremely difficult tofind comparative measures of ideology across several localities. Earlierattempts atmeasuring shared beliefs among communities have led to concepts such as local "politicalculture" or "political ethos" (see Banfield andWilson1963; Eulau 1973). Although there were efforts tomeasure political culture or the "ethos" of local residents, these studies involved only indirect indicatorsof voters' views (Hawkins 1971).In limited cases, scholars have utilized publicopinion data sampled frommultiple local communities by a single research organization using comparable sampling and survey administration techniques(e.g., Pierce, Lovrich, and Moon 2002). However,

these kinds of comparable data are only narrowlyavailable. Scholars have also relied on the presidential vote totals as a proxy for local ideology (DeSantisand Renner 1996) and most recently have used innovative simulation techniques to create measures ofschool board ideology across the states (see Berkmanand Plutzer 2005). The former approach brings thepossibility of confounding attitudes toward nationaloffice with local issues, while the lattermeasures relyon the assumptions built into simulations, and it isdifficult to ascertain the extent towhich they are representative. We use a different technique informed byan approach developed for the study of public opinion and policy at the state level.

Measuring Political Orientations at theLocal Level inCaliforniaIn this research we measure local ideology usingmethods similar to those of Erikson, Wright, andMclver (1993). Their research significantly advancedtheunderstanding of state-level public opinion by cre

ating reliable and valid measures of political ideologyand partisanship y pooling 1976-1988 nationallysampled CBS/New York Times polls and aggregatingthem at the state level. In the years since Wright,Erikson, andMclver (1985) developed the approach,a number of scholars have studied opinion-policylinkages at the state level, exploring patterns of femalerepresentation (Arceneaux 2001), the use of the deathpenalty (Norrander 2001), and environmental policyresponsiveness (Johnson, Brace, andArceneaux 2005).A recent edited volume (Cohen 2006) provides anexcellent overview of these techniques and applications to state politics. We use a similar approach, reaggregating to the county-level statewide CaliforniaField Poll surveys conducted 1990-1999.1Established in 1947, and continuing every yearsince, the Field Poll routinely fields surveys questionstoCalifornia residents on a wide range of public policy issues and questions regarding their support forvarious political candidates at the national, state, andlocal levels of government.2 For this research, datawere gathered from forty-eight Field Poll surveystotaling 51,930 individual respondents. The Field Pollconsistently asks respondents to place themselvesalong a three-point political ideology continuum.Specifically, respondents were asked, "Do you consider yourself to be politically conservative, liberal,middle-of-the road, or don't you think of yourself inthisway?" Conservatives were coded 100,middle-ofthe-road 0, and liberals -100. In addition, the Field


5/15


Table 1IdeologyScores byCountywith Sample Sizes

County Namedeology Sample Size CountyName Ideologyample SizeSierra

MaderaTulareShasta

MariposaKernInyoSutterEl Dorado

MonoFresnoSan BenitoNevadaTehamaSan BernardinoAmadorRiversideKingsOrangeSan JoaquinYubaVenturaCalaverasPlumasPlacerTrinityColusaSiskiyouStanislaus

50.0041.2741.2337.4037.1436.3535.2535.1434.9534.7834.3833.7832.6131.8231.7431.5830.3630.3629.7429.3928.9227.8627.7827.7827.5927.2726.9226.9226.56

1412735025337666341162082377474184891,727571,4601163,451570838547237296142678441

MercedGlennDel NorteLassenButteTuolumneSolano

MontereyNapaSanDiegoSanta BarbaraSacramentoSan Luis ObispoImperialHumboldtLakeModocLos AngelesSanta ClaraContra CostaMendocinoSan MateoYoloSonomaSanta CruzAlamedaMarinAlpineSan Francisco

25.1225.0024.2424.0723.7523.6621.0121.0020.7519.9419.7319.3518.1815.0013.2710.5910.5310.388.418.366.453.270.000.00

-5.81-7.78

-11.90-20.33-25.35

215373355361934044301663,0445241,4593961202008621

10,3262,0201,1001279731935653481,740338151,449

Poll asks each respondent his or her county of residence, allowing us to link each response to a givencounty. Individual responses were then aggregated tocreate ideological scores for California's fifty-eightcounties. The number of cases in each county rangedfrom 10,326 inLos Angeles county to fourteen inSierraand Trinity counties (mean = 659.01). Ideology scoresranged from themost conservative Sierra county (50.00)to themost liberal San Francisco county (-25.35) witha mean = 21.28. Ideology scores and sample sizes foreach of thefifty-eight counties are listed inTable 1.

Auditing theCounty-LevelMeasure of IdeologyIndividual responses are treated here as aggregate

data, and therefore it is not appropriate to use standard measures of individual-level reliability likeCronbach's alpha (Brace et al. 2002). Because of this,Jones and Norrander (1996) recommend testing reliability analysis on the basis of aggregate units, andnot individuals. To first test the generalizability of the

ideology measure, we use the O'Brien coefficient(O'Brien 1990). Presented by Jones and Norrander(1996), the O'Brien generalizability test compareswithin-unit variance to the across-unit variance whiletaking into account sample size (Norrander 2001,113).3Measures of ideology will be more generalizable across units with less intra-county variation andmore variation in ideological dispositions intercounty. An O'Brien generalizability coefficient thatexceeds .70 is considered to be highly generalizable,and values between .60 and .70 are considered to bemoderately generalizable. The O'Brien coefficientfor the county-level ideology measure is .96.An additional test of reliability is the split-halfapproach used by Erikson, Wright, and Mclver(1993). The split-half approach involves splitting theField Poll sample into two subsets by assigning oddyear surveys to one subset and even years to theother.Mean scores for county ideology were calculated foreach subset and correlated using Pearson's r coefficients. The Spearman-Brown prophesy formula wasused to assess the reliability of each measure:


6/15

168 Political ResearchQuarterly2rv1+rv

where r12 thePearson's r correlation between the splithalves. Reliability scores of .70 and above are considered reliable, those between .60 and .70 are consideredmoderately reliable, and those below .60 are considered unreliable (Jones and Norrander 1996). TheSpearman-Brown coefficient for the reliability of thecounty-level ideology measure equals .60.To test the stability of themeasure, the Field Pollsample was divided into "early" and "late" subsets. Theearly subset included survey years 1990-1995, and thelate subset 1996-1999. Mean scores for county ideology were calculated and correlated. The SpearmanBrown coefficient for the stability of county-levelideology was .62, close towhat is considered theminimum reliability level. In sum, the assessment ofthe reliability of the ideology measure is mixed. TheO'Brien measure is highly reliable, although theSpearman-Brown coefficients using the split-halfapproach are at the low end of scores considered tobe"moderately" reliable. Given this, in a final attempt toassess the validity and reliability of themeasure, weconstructed an alternative measure of county-level ideologyusingpooled survey atafrom he ublicPolicyInstituteof California (PPIC). PPIC conducted seventyone surveys from 1998 to September 2006, and eachsurvey included questions about political ideology thatsubstantively mirrors the Field Poll question we usehere.We constructed the same ideology measure usingthis PPIC data, aggregated to the county level. Thesecounty-level ideology scores are highly correlated withtheField Poll county scores (r= .82,p < .01), suggesting thatwith repeated sampling, the survey-based measure we use reliably gauges local-level politicalorientations.We have chosen to use theField Poll ideology measure here, but at the same timemake note ofitspossible deficiencies, which in the analysis wouldlikely tend to attenuate findings.

How Well Do Field Poll-Based MeasuresRepresent County-Level CaliforniaDemographics?

California counties are not thepopulation of interest for the Field Poll, and thuswe cannot assume thatthe sampling frame employed by that organizationproduces representative estimates of county population.As notedbyHill andHurley (1984), a samplebias might be introduced when creating a nonrandomsample from state residents. To test the validity of the

Field Poll sample, a series of demographic characteristics were derived from the Field Poll sample andcorrelated with county demographic characteristicscollected by theU.S. Census (see Brace et al. 2002).Results presented inTable 2 show that county samplesobtained from theField Poll are substantially representative. Specifically, we find a strong correlation betweenthe educational attainment of the sample and educational attainment reported by theU.S. Census in 1990and 2000 (U.S. Census Bureau). A similarly strong relationship is found among between the income of FieldPoll respondents and U.S. Census statistics. Racial characteristics of respondents, although showing a slightlyweaker correlation toU.S. Census figures than do theeducation and income figures, are moderately strongnonetheless. Importantly, the strong correlations foreducation and income and themoderately strong correlations for the race variables suggest that the Field Pollsamples adequately reflect county populations.

Expectations about Public Opinion,Government Institutions, andPolicy Making

Using themeasure of county ideology discussedabove, our primary expectation is that counties' ideologicalmakeup will impact policy outputs at thecounty level;however, we expect the relative impact of county ideology tovary across issue type and differing county structures. Among different issue types, local politicalorientations should affect redistributive policies ratherthan those policies considered to be developmental orallocational in nature (Wong 1988). Policies associatedwith welfare payments and public health care are considered by Peterson (1981) as redistributivepolicies, and asnoted above, thesepolicies often raise issues thatare ideologically divisive and more likely to elicit support oropposition along ideological lines.Traditionally, conservatives push formore restrictivewelfare policies and lessgovernment spending on public health care, reflectingtheir views toward limitingboth the size of governmentand the scope of government intervention. In contrast,liberals have traditionallypushed forgreaterwelfare benefits with fewer restrictions and a more active role forgovernment in health care,which reflects theiroverarchingbelief in the social benefits thataccrue from a larger,more active role for government in fighting social illsrelated to poverty (Rom 1999). These associationsbetween ideology and policy outputs have been empirically demonstrated at the state level,where research hasfound greater social welfare spending among stateswithmore liberal publics and political elites (Erikson,Wright,


7/15


Table 2Representativeness of Field Poll County Samples

2000 1990 Average 1990/2000EducationIncomeWhiteBlack

AsianDemocratic Party

registration**p < .01Note: Education is measured by the percentage of county residents who have earned a bachelor's degree or higher. Income ismeasured by correlating the percentage of Field Poll respondentswho mentioned their total household income was between$20,000 and $40,000 dollars and themedian household incomeof the respondent's county reported by the U.S. Census. Racialcharacteristics are based on sample estimates drawn from selfreported information from the Field Poll and are correlated withU.S. Census data. Democratic Party registration is based on thepercentage of Field Poll respondents who identified themselvesas members of the Democratic Party and correlated with voterregistration data housed by the California Secretary of State.

and Mclver 1993; Hill and Hinton-Andersson 1995).Conversely, allocationai and developmental policies areless likely toengender political conflict, and thereforewemight expect ideology tomatter lesswith respect to theseand other allocationai and developmental policies.

Following this, our first hypothesis is thatmoreliberal counties will produce higher levels of spending on redistributive social programs likewelfare andpublic health care relative tomore ideologically conservative counties. Because allocationai and developmental policies are generally associated with lowerlevels of political conflict, we expect ideology tohave little or no effect in these areas.

Formally, the following hypothesis is tested:

HI: Ceteris paribus, ideologically liberal countieswill produce higher levels of spending in redistributive policy areas like welfare and publichealth care relative tomore ideologically conservative counties.A growing body of literature focused on the impactof county form or structureon policy suggests thatany

relationship between ideological orientations and policy outputsmay also be influenced by county structure(DeSantis and Renner 1996). Contemporary reforms in

American county structures have been designed toincrease professionalism and centralize executive leadership tomore effectively carry out their expanding

91** 91** .91**90** 92** 92**.75** 76** .76**g7** 88** 88**.86** 87** .87**.76** 78** 77**

service delivery roles (Benton 2002b). Thus, "reformed"county governments include those with an elected orappointed executive (rather than the traditional countycommission form, which lacks a singular executiveauthority), home rule charters, and nonpartisan elections. Early tests of the effects of local governmentstructureson policy decisions placed emphasis on citiesrather than counties. Lineberry and Fowler (1967)found that reformed city governments (at the city levelreformed structures are those with council-managergovernment and at-large and nonpartisan elections) hadlower taxes and expenditures than unreformed cities(i.e., citieswith mayor-council government and partisanelections); however, Clark (1968) found reformed governments tohave higher expenditures.More recent work examining the structure-policylink at the county level has consistently found thatreformed county executive structures have higher percapita expenditures than do unreformed county commission forms (see Schneider and Park 1989).Extending this line of work, DeSantis and Renner(1996) argue that the impact of county structure isnotnecessarily additive (or direct) but rather conditionsthe influence of county contextual factors like socioeconomic and political factors on policy choices at thecounty level. That is, particular county governmentstructuresmay facilitate or hinder counties' abilitiesto respond to specific policy demands. For example,using a measure of the percentage of the two-party(Democratic) vote in the 1988 presidential campaignas a proxy for county ideology, DeSantis and Renner(1996) find ideology isunrelatedtoexpendituresneither county commission or county executive structuresbut has a positive and significant impact on expenditures for county administrator forms. Overall thisevidence highlights the importance of testing for conditional effects between county ideology, countystructure, and local policy outputs.

InCalifornia, county government structuredoes notdiffer dramatically?all counties can be considered"reform" governments in that they have nonpartisanelections and appointed executives rather than countycommissions with no executive. However, the state ofCalifornia has two classifications of counties basedon whether a county is considered to be a home rule"charter" county or a "general law" county. AmongCalifornia counties, fourteen of fifty-eight are considered "charter" counties (Connell 2001). Generally, ahome-rule charter grants a county a greater degree ofself-rule and self-determination that frees itfrom somelegal restrictions imposed by the state (Duncombe1977). Chartered governments can also leverage


8/15

170 Political ResearchQuarterlydifferent fiscal reforms thatmake it easier to respondto increasing resident demands for an expanded menuof services as well as higher levels of current services(Benton 2002b). The California Constitution, however,does not allow officials in charter counties extra regulatory functions or added revenue-raising abilities, butit does allow them the ability to consolidate or segregate different county administrative offices, providefor the election or appointment of county officials, andset powers and duties of all officers.4 Although theState of California provides a limited amount of freedom to its charter counties vis-a-vis other state-countygovernment arrangements, given their increased discretion and capacity relative to general law countystructures,we might expect the ideology-policy linkage to be further strengthened in those counties wherea charter is present.

H2: The relationship between county ideology andredistributive policy should be strengthened incounties where a charter is present. Among countieswith home rule charters, ideologically conservative counties will produce lower levels ofspending on redistributivepolicy areas likewelfareand health care while more ideologically liberalcounties will produce higher levels of spending.

Ideology and Local PolicyOutputs inCalifornia

To examine the relationship between policy and theideological disposition of California counties, we usedata from California's Office of the Controller oncounty-level spending across six policy areas: publicassistance, health and sanitation, public ways and facilities, education, public protection, and general governmental expenditures. These measures are the study'sdependent variables. The policy measures describe anumber of differentpolicy areas?redistributive policies(public assistance, health care), developmental spending(public ways and facilities, education), and allocationaiexpenditures (public protection and general governmental spending).We use thesepolicy categories for twoprimary purposes. First, these are all policy areas inwhichcounty governments are actively involved in regard topolicy formulation and implementation. Second, theyallow us to employ Peterson's issue typology to explorethe relationship of ideology with policy outputs across asubstantivelywide range of issues. Following Peterson(1981), each of these policy areas is likely associated

with differentdegrees of political conflict, depending onwhether itmanifests primarily a redistributive,developmental, or allocational dynamic.Policy Indicators

The Office of the State Controller in Californiareleases an annual report on county revenues andexpenditures, separate from fiscal data of all other localgovernments (e.g. municipalities, school districts, orcommunity college districts).We draw our policy indicators from this report.5We identify six areas of publicexpenditures (Connell 2001), two in each of Peterson'sissue typology. For allocational policies, we use generalgovernment expenditures, which includes budget itemsfrom day-to-day county administration (e.g., legislativeand administrative expenses, finance, counsel, personnel, elections, property management, etc.) and publicprotection (e.g., judicial services, police protection,detention and correction, fire protection, etc.). Thedevelopmental policies we examine are education(school administration, library services, and agricultural education) and public ways and facilities (including roads, transportation systems, and parking). Finally,our redistributive policies are health care (public health,medical care, mental health, drug and alcohol abuseservices) and public assistance (welfare, social services, general relief, etc.).To match the timing of our ideology data(1990-1999), we use county expenditure data from the1998-1999 fiscal year. We transform each of thesebudget lines intomeasures of per capita spending, bydividing each county's expenditures in each area byestimates of the county population in 1999 (CaliforniaInstitute of County Governments 2001). Consequently,the dependent variables in themodels that follow areestimates of per capita expenditures across six policyareas at the county level for 1998-1999.6Alternative Explanations

As noted, ourmain independent variable is countylevel political ideology, but we also consider otherpossible explanatory variables. For example, the literature on local policy determinants has long suggestedthe importance of a host of social, economic, anddemographic factors that shape public policy (for anexcellent compilation of relevant factors see Kantorand David 1983).

Socioeconomic characteristics. It is intuitive thatfiscal capacity is a strong predictor of policy outputs


9/15


(Dye 1979).How much a jurisdiction evotes to anactivitymust, inpart, reflect the ability to support suchservices. Studies of subnational governments demonstrate thateconomic factors like the per capita incomeof residents predict public sector expansion (Hawkins1971; Feiock andWest 1993). Education is also relatedto local expenditures: where more educated publics arepolitically involved they tend to finance public programs. In the models presented below, we includemeasures of median household income and the percentage of residents who have earned a high schooldiploma or higher. Data are drawn from the 2000 U.S.Census. Itmight be expected thatcounties with greatermedian incomes and educational attainment levelswould have a greater capacity to increase spendingacross all thepolicy areas.

Intergovernmental revenues. Although countyspending levels can be viewed as partly a functionof the socioeconomic characteristics of its residents, we also have to take into account that countyexpenditures are constrained by intergovernmentalgrants-in-aid or revenue sharing from the state or federal government. For example, counties may receivefunding from the state or the federal government (orboth) forwelfare or health care programs, but levelsof funding are often set by how many people live in acounty who meet eligibility criteria rather than howmuch the county wants to spend. Because it is ourgoal to show that local ideological forces shapecounty spending choices, it is important to control forany effects driven by federal or state revenue sentdown to county governments. To account for countyspending choices influenced by state and federalfunding levels, we include in the regression modelstwo variables labeled state funds and federal funds,which are per capita measures of state and federalrevenues directed to each county in 1998-1999(Connell2001).We expecthigher evelsof spendingin those counties that receive higher per capita levelsof state or federal funding.

Racial politics. Increasingly, political scholars pointto racial diversity within any given environment as asignificantredictorf publicpolicy and how publicprograms are distributed (Hero 1998). Racial diversitymay impact policy in a couple ways. Geographic proximity to large number of racial minorities may increasesentiments of racial threat among whites (Key 1949;Stein, Post, and Rinden 2000). Perceptions of racialthreat tend to decrease support of policies perceived tohelp minoritymembers (Stein, Post, and Rinden 2000).

Moreover, when public programs are perceived to target minority groups, program allocations tend tobecome less generous (Katz 1989). Research of mostrelevance here shows local governments tend to imposetougher sanctions and fewer benefits towelfare recipients living in racially diverse contextual environments(Keiser, Mueser, and Choi 2004). To control for thepossible influence of race on the dependent variables,we include ameasure of the percentage of black residents and thepercentage of non-white Hispanics residing in each county. Data are drawn from the 2000 U.S.Census (U.S. Census Bureau).

FindingsTable 3 provides OLS regression models for eachof the six policy output dependent variables. In eachof the regression models, we include a county structuredummy variable (1 = charter, 0 = no charter) anda charter*ideology interaction term to test the second

hypothesis that county structure conditions theimpact of ideology on county-level policy outputs.

Among the redistributivepolicy models, county-levelideology has a negative and statistically significantinfluence on redistributive policy outputs after controlling forother possible predictors; however, the relationship is only significantwhen consideringpublicassistance expenditures. The negative coefficient for theideology measure in the public assistance model indicates that ceteris paribus, liberal counties aremore likelyto expend greater funds on welfare and other social services than counties that aremore conservative. A standard deviation change in ideology exemplifiesmovement from a relatively liberal place like LosAngeles County to a substantially more liberal placelike lamedaCounty with ities ncludingakland andBerkeley), or alternatively a shift fromOrange Countyin southernCalifornia toHumboldt County in the northern portion of the state.The standardized coefficient forcounty ideology in the public assistance model is .28,suggesting that a standard deviation shift in ideologywould be associated with an additional $90.24 spentpercapita on a county's public assistance programs in1998-1999. The debate over assistance to thepoor andthe scope of government involvement aremore oftenpolitically contentious and structured by ideologicalbeliefs and attitudes. Given this, it is not surprising tofind that local ideological dispositions relate to outputsin thispolicy

area.In addition, the charter*ideology interaction termhas a negative and statistically significant association


10/15

172 PoliticalResearchQuarterlyTable 3

Modeling Local Per Capita Spending across Six PolicyAreas, OLS with Robust Standard ErrorsAllocationai Policies Developmental Policies Redistributive olicies

IdeologyIncomeEducation

County structure

Charter*ideologyState fundsFederal fundsBlackHispanicConstant

General Public Protection Education PublicWays andFacilities Health Care PublicAssistance-2.441(1.91)0.005**(0.002)7.784***(2.80)

-140.22***(69.40)

4.890*(2.55)1.359***

(0.251)0.233(0.185)10.76

(392.95)391.281(217.60)

-1306.16***(282.25)N=51

F947= 5.64***\ff2=.84

-0,(1.0,(0,4,(1

-72,(49,

1(11

(00(0

-30(320211(162

-681(185NF =

All112)003***.001)403***.50)99,62).96.77).106***.142).074116).485,44).61268)630.11)= 579.85***= .87

-0.390*(0.226)0.000(0.000)0.546*(0.291)

-12.43***(5.03)0.467**(0.212)0.080***(0.019)0.022(0.013)14.556

(45.93)25.954(22.81)

-73.244***(25.51)N=51

F947=4.99***#2=.78

2 37***(0.847)0.006***(0.001)

-2.14(1.40)

-33.22(43.83)-0.223(1.57)0.675***(0.066)0.522***(0.123)

-390.392(244.13)

-254.638**(121.02)

-412.604***(110.80)

N=51F9A1= 147.15***#2=.88

-0.093(0.964)0.000(0.001)

-0.509**(2.016)

307.212**(154.196)-12.102*

(5.875)0.348***(0.090)0.097(0.095)

-402.125(522.447)

0.195117.470-14.158(136.071)

N=581 9,48~ J-^yR2=.66

-2.21***(0.708)

-0.006***(0.001)3.998**(1.69)40.555

(33.78)0.151(1.23)0.110t(0.063)0.408***(0.111)

373.299**(169.52)329.334***(117.91)125.490***

(114.890)N=51

F947= 70.49***R2=M***/? < .01, **p < .05, *p < .10Note: Robust standard errors shown in parentheses below each coefficient. Except for the health care model San Francisco County isexcluded from the analysis because expenditures across these policy areas are not recorded in a comparable manner.

with counties' health care expenditures, with moreconservative counties with charters spending less, onaverage, thanmore ideologically liberal counties withcharters. Conversely, no relationship is found betweenthe charter*ideology interaction term and publicassistance expenditures. At least when consideringhealth care expenditures, this finding suggests countystructure has amediating impact on county ideology,where as expected, a charter strengthens the association between counties' ideological dispositions andpolicy outputs.Several additional predictor variables are associated with patterns of county-level spending acrossthe different policy domains. Higher levels ofincome are associated with more general government, public protection, and public ways and facilities spending, but less spending on publicassistance. Higher educational achievement is associated with more spending on general government,public protection, and public assistance, but lowerlevels of health care expenditures. Counties withhigher levels of per capita state funding expend

greater funds across each policy indicator, whilefederal funds increases counties' spending in thepublic ways and facilities and public assistance policy categories. The county structure variable alsoappears to influence spending choices, with those countieswith home-rule charters spending more on general government, education, and health care. Finally,larger black and Hispanic populations are associated

with more spending on public assistance.As expected, controlling for these alternativespending explanations, we find no significant relationship between ideology and thebudget lines in thedevelopmental policy areas and allocational policieslike general government expenditures; however, contraryto our expectations, ideology had a negative andstatistically significant relationship with public protection spending. Overall, these findings lend supportto Peterson's (1981) assertion that allocational policies like legislative and administrative expenditures,and developmental policies like highway construction cause little political conflict, and as we advancehere, less likely to be influenced by local ideological


11/15


Table 4Modeling Local Per Capita Spending across Six Policy Areas, Heteroskedastic Regressions

Allocational Policies Developmental Policies Redistributive oliciesPublicublicWays Health Public

General Protection Education and Facilities Care ssistanceChoice modelIdeologyIncomeEducation

County structure

Charter*ideologyState fundsFederal fundsBlackHispanicConstantVariance modelEstimated margin of error

Constant

-0.306(0.59)

-0.000(0.000)1.29

(1.129)-35.83**(18.721)0.563(0.769)0.260**(0.129)-0.100(.088)

-184.517(155.738)-33.183(91.840)

-36.987(110.304)

30.205***(2.827)5.410***(0.315)n=57

%2= 192.67***ps. r2=.23

0.924(0.811)0.001(0.001)0.479(1.589)

-28.659(29.243)-0.345(1.170)0.647***(0.141)0.000(0.113)

-75.041(228.802)

8.494(126.35)-86.642(145.004)

18.392***(2.827)6.752***(0.315)n= 51

%2= 142.90***ps. #2=.18

-0.065(0.081)0.000(0.000)0.165(0.157)

-6.420**(2.770)0.143(0.111)0.016(0.016)0.012(0.011)19.768

(22.220)-12.594(12.617)

-12.986(14.915)22.699***(2.827)1.874***

(0.316)n=51%2= 130.02***

ps. r2=.25

-0.165(0.462)0.000(0.000)0.150(0.861)

-20.734(13.645)

0.463(0.568)0.063(0.102)0.144**(0.070)

-210.452*(116.591)

-134.158**(70.81)19.88

(84.165)37.362***(2.827)4.503***(0.315)n=51

%2= 140.285***ps. r2=a9

-0.938*(0.559)

-0.000(0.001)1.324

(1.786)309.748***(59.681)

-12.016***(2.121)0.201***(0.049)0.120(0.087)

-661.484(414.806)-16.577(140.117)

1.666(127.453)-13.886***

(2.804)10.058***(0.310)n=5s

x2= 85.301***ps. r2=.u

?1 923***(0.605)

-0.006***(0.000)3.210***(1.247)34.242

(26.815)0.155(1.041)0.105(0.067)0.460***(0.077)304.186

(195.553)294.708***(97.025)150.518

(101.104)7.757***(2.827)6 999***(0.315)n=51

x2= 100.256***ps. r2=a4

***/? < .01, **p < .05, *p < .10

dispositions. Significantly, these results suggest thatthe importance of ideology on local-level policy makingwill depend to an extent on the issue under consideration and the structure of county government.The results presented thus far

are mixed, havingfound evidence of a direct relationship between ideology and redistributive policy but only when taking intoconsideration public assistance expenditures like welfare and other social services. The absence of a statistically significant association between county ideologyand both redistributivemodels may be a result fromnot having accounted adequately for the variation insample sizes used to compute the ideology measure atthe county level. In Table 1,we see sample sizes ranging from fourteen (Trinity and Sierra Counties) tomore than ten thousand (Los Angeles County). Giventheobvious source of heteroskedasticity in thesemodels,we would not trust conventional standard errors, sowe estimated themodels in Table 3 calculating robust

standard errors. Still, we are concerned about potentially biased coefficients in thesemodels and hypothesis testsmarred by theuncertainty associated with thesestandard errors.Because we know that sample size willsystematically affect the variance of the disturbancesaround the regression lines, we deal with thismoresystematically, by modeling thevariance of the regression line using heteroskedastic regression.7Table 4 shows models similar to those inTable 3,reestimated with the variance models described in

Note 7. We see that, as expected, themargin of errorwe computed for each county subsample has a systematic influence on the disturbance term of thechoice model for each regression. The larger themargin of error, the greater the variance is around theregression line for a given observation.

Closely resembling the results in Table 3, thechoice models show county ideology is a significantpredictor of public assistance spending, but not health


12/15

174 PoliticalResearchQuarterlycare expenditures. In thepublic assistance model, thenegative and significant coefficient suggests thatmore liberal counties spend more on things like welfare programs than do more ideologically conservativecounties. Again, the charter*ideology interaction termfailed to reach statistical significance suggesting thatthe relationship between ideological dispositions andpublic assistance programs is not conditioned bycounty structure.

However, the charter*ideology interaction term isagain statistically significant in the health care model.Support for these types of health care services, similar tothose connected to other forms of public assistance likewelfare, tend to break down along political ideologicallines. The finding here suggests that the influence ofcounty ideology on per capita spending on public health,medical care,mental health, or drug and alcohol abuseservices is conditioned by county structure?more conservative counties with home rule charters spend, onaverage, fewer dollars on health care compared tomoreideologically liberal counties with home rule charters. Inthecase of health care expenditures, the influence of ideology is only apparent when counties have greater structural freedom from the arms of the stategovernment andwhere policy makers have greater administrative capacity tomatch thepolicy to the preferences of their constituents.As expected, the choice models show that therelationship between ideology and public protectionexpenditures found inTable 3 disappears, with countyideology having no significant impact on any of the allocationai or developmental policies.Taken together, these results suggest that ideological cleavages at the local level help drive differences incounties' spending choices across redistributive policies. However, when considering the relationshipbetween local ideology and local policy outputs, theimpact of counties' ideological makeup differsdepending on the issue type under consideration, inaddition to county structure.As hypothesized, countyideology influences redistributive policy areas?wherepolitical conflict ismore likely to be felt?rather thanthose that are less politically contentious at themasslevel like developmental and allocation policies. Giventhis, even among redistributive policies the ideologypolicy linkage is not necessarily straightforward?county structurematters.When considering health carespending, the impact of county ideology is only evidentwhen coupled with a home rule charter that providespolicy makers greater capacity to attune particular policy expenditures to ideological beliefs held amongcounty residents.With public assistance policy, countyideology appears to influence policy outputs regardlessof county structure with more liberal/conservative

counties producing more liberal/conservative policyoutputs. These differences may be partly a function ofthenature of political conflict surrounding welfare andhealthcarepolicies in the ate 1990s.The late 1990smarked a politically contentious era inwelfare policyas controversial welfare reform efforts, driven by theadoption of the federal 1996 TANF law, were wellunder way at the county level. Thus, relative to healthcare policy, the high degree of political conflict connected towelfare may be behind themore robust relationship between ideology and public assistanceexpenditures; however, more research is needed tounderstand thesemore nuanced effects among the tworedistributive policy areas.

DiscussionThe findings presented here suggest students of subnational politics would benefit from paying additionalattention to the ideological variation within individualstates as well as variation between states.Using awellestablished, multiyear state survey instrument like theCalifornia Field Poll to create reliable, valid, and stable measures of local ideology as we do here indicatesthat subnational researchers can use similar method

ological tools to advance our understanding of thewaylocal ideology influences local policy making and theextent towhich policy represents thepolitical interestsof diverse, localized populations.We find that local policy outputs are influenced bycounties' ideological dispositions where more liberal/conservative counties produce more Uberal/conservativeoutputs across a range of policy areas including publichealth care and welfare. Importantly, the influence ofcounties' ideology varies across differentpolicy issueswith ideology playing amore importantrole on redistributive policy areas?those issues characterized by ahigher degree of political conflict?and a less importantrole on issues where littlepolitical conflict occurs. Onissues with little political conflict, intergovernmentalexpenditures and county-level demographic and socioeconomic contextual characteristics tend todrive expenditure patterns across the counties. The relationshipbetween ideology and policy also appears to be conditioned by county structure. In California, a home rulecharter appears to strengthen the relationship betweenideology and health care expenditures than the maineffects of county ideology alone would suggest.Moreconservative counties with charters are found to spendless on public health care relative tomore ideologicallyliberal counties. This set of findings challenges previousassumptions connected to the local policy-making


13/15


process. Prior research suggests that local economic considerations are themain driving force behind local policydecisions and local governments should be biasedagainst redistributive policies. Taken as a whole, weshow thatthatunder theright political and structuralconditions, there is still room for local politics to influencepolicy choices at the local level of government.

Finally, it is important to consider the extent towhich our findings here are representative of the localideology-policy relationship in other county governments across theU.S. political system. Does the factthatwe find associations between ideology and policyinCalifornia suggest that these same forces spill overto other county governments in other regions such astheNortheast, South, orMidwest?DeSantis and Renner (1996) note significant variation in county spending levels throughout differentregions of the United States. Traditionally, westerncounties spend more (in absolute terms) than countiesin other regions?especially those counties in theSouth, which tend to have more powerful stategovernments relative to local governments. TheProgressive movement of the early twentieth centuryhas left a lasting legacy onWestern county governments as most are considered "reformed" governments and, with that, have a greater capacity todeliver an increasing array of services. Our study,which examines the influence of local political ideology on policy in the State of California, with itsreformed county structures and relatively activecounty governments ismost likely to be representative of local policy making in the western UnitedStates. As the population of California and the rest ofthe western states continues to grow and as localcommunities form unique political attitudes causedby increasing racial and ethnic diversity, finding associations between local ideological attitudes and localpolicy is an important contribution by itself.

Although absolute spending levels have been documented to be highest in theWest, there remains astrong likelihood that creating reliable and valid measures of county ideology can be shown to shape relative spending levels in counties in regions of thecountry thathave traditionally spent fewer dollars inabsolute terms. In the end, this is an empirical question that is beyond the scope of this research. Anumber of regular, institutionalized state-level surveys (NationalNetwork of StatePolls 2007) suggestthepossibility that scholars can create ideology measures like those used here to test the influence of ideology in counties outside thewestern region of theUnited States. Importantly, we believe the analysishere adds to the growing body of work suggesting

that a variety of local political forces continues toshape policy and provides a framework for a productive and rewarding research agenda in the area oflocal government policy making.

Notes1. Cohen (2006, 6-10) describes a varietyof alternatives o

measuring state public opinion using pooled national sample surveys. In principle, these alternatives are also feasible formeasuringcounty-level ideology. These alternatives include pooling surveyswith subunitsamples (Jonesand Norrander 1996), combiningindependent surveys taken within comparable subnational geographicunits (Beyle,Niemi, and Sigelman 2002) and simulatingpublic opinionusing surveydata or otherpolitical information(Weberet al. 1972-1973; Berry et al. 1998; Park,Bafumi, andGelman 2003). Given the absence of California surveys with comparable county-level subunit samples or any comparable independent surveys of California counties, we are unable to pursue two ofthese alternatives. Simulating county-level ideology is a reasonablealternative (see Berkman and Plutzer 2005), but these techniquesgenerally rely on extensive assumptions about connectionsbetween individual demographic attributes and opinions. We makedifferent assumptions, namely that the coverage of county residents in these pooled statewide samples of California allow us tocapture a meaningful county-level measure of ideology. Scholarswho simulate public opinion measures share our assumptionsaboutforming ublic opinionfromtheaggregation f individualopinions. In sum, the simulation of public opinion is complementary to the approach we take, but not clearly preferable.2. The Field Poll uses samples of theCalifornia telephonehouseholdpopulationdrawn from andomdigitdial (RDD) samples of Survey Sampling Incorporated. The sample is a stratifiedsample of California counties where samples are systematicallystratified to all counties in proportion to each county's share oftelephone households in the survey area. Further sampling information can be referenced from the California Field Poll CodeBooks 1990-1999.3. O'Brien's (1990) generalizability oefficient or theR:Adesign contemplates themean square, an estimate of the population variance between aggregate units, MS (a), and the meansquare for individual-level scores within the aggregated units,

MS(r:a), using the formula:

[MS(a)-MS(r:a)]Ep2= - MS(a)MS (a) andMS(r:a) were estimatedusing theone-wayANOVAprocedure in SPSS.4. However, in both home rule charter and general law counties, the county sheriff, district attorney, and assessor must beelected.5. The definitions f thebudget itemsprovidingthepolicyindicators are discussed in an appendix to the State of CaliforniaCountiesAnnualReport (Connell 2001, 145-7).

6.We follow other researchers who have investigated policyoriented dependent variables from a cross-sectional measure, orbased on a small number of years using an independent variableconstructed from data from an inclusive larger set of years (e.g.,


14/15

176 Political ResearchQuarterlySchneider and Jacoby2006; Uslaner 2006). The indicators eand other researchers use are contemporaneous.7.Harvey (1976) develops a regressionmodel that llows usto systematically model sources of heteroskedasticity. A choicemodel is used to test hypotheses about the dependent variable,and a variance model is used to explore systematic variance in thechoice model's disturbance term:yt= xfi + u. (i=l,2,3,...,?), (1)

cf=eziai=l,2,3, ...,w), (2)

where y. is a vector of observations on the dependent variable, x. is avector of observations on independent variables, (3 s a vector of parameters, ?. is the disturbance term of the choice model, o/2 is the variance of the disturbance term, zt is a vector of observations ofindependent variables, and a is a vector of parameters. In the traditional regression model, we assume the disturbances around theregression line are distributed Normal with a mean of 0 and a fixedvariance (u. ~ N[0, a2]). Harvey's multiplicative heteroskedasticity

model, we anticipate systematic variance in the disturbance term.Here we model the variance of the residuals of the choice modelas a function of a computed margin of error, which takes intoaccount sizes of the county subsamples, as well as the size of theunderlyingopulation f the ounty ampled. his isroughly quivalent to themargin of error discussed with polling results, plus orminus some percentage that depends primarily on the number ofpeople surveyed, as well as the size of the population of interest.Given thewide variance of county sizes as well as wide variance ofsubsample sizes, we thought it important to take both into account.We calculate margin of error (see Weisbert, Krosnick, and Bowen1996,74) usingthefollowing ormula:

where t= 1.96 for a 95 percent confidence interval, p = the sample proportion which we set at .5 to calculate a conservative margin of error, N = number of observations in county subsample,and/= the sampling fraction, or the number of respondents in thecountysubsampledividedby the totalcountypopulation.Usingthe formula, we compute margins of error for the county-levelideology scores that vary from ?1.0 percent (Los AngelesCounty) to 27.2 percent Trinityounty).

ReferencesArceneaux, Kevin. 2001. The "gender gap" in state legislative

representation: New data to tackle an old question. PoliticalResearch Quarterly 54:143-60.Banfield, Edward, and James Q. Wilson. 1963. City politics.

Cambridge, MA: Harvard and MIT Press.Benton, Edwin J. 2002a. Counties as service delivery agents:

Changing expectations and roles. New York: Praeger.Benton, Edwin J. 2002b. County service delivery: Does government structurematter? Public Administration Review 62:471-9.Benton, Edwin J. 2003. County governments. In Encyclopedia of

public administration and public policy, ed. Jack Rabin. NewYork: Marcel Dekker.

Benton, Edwin J., and Platon Rigos. 1985. Patterns of metropolitan service dominance: Central city and central county service roles compared. Urban Affairs Quarterly 220:285-302.Berkman, Michael B., and Eric Plutzer. 2005. Ten thousand

democracies: Politics and public opinion inAmerica's schooldistricts. Washington, DC: Georgetown University Press.Berman, David R., and Tanis J. Salant. 1996. The changing roleof counties in the intergovernmental system. In The American

county: Frontiers of knowledge, ed. Donald C. Menzel.Tuscaloosa: The University ofAlabama Press.Berry, William D., Evan J. Ringquist, Richard C. Fording, andRussell L. Hanson. 1998. Measuring citizen and government

ideology in theAmerican states, 1960-93. American Journalof Political Science 42:327-48.

Beyle, Thad, Richard G. Niemi, and Lee Sigelman. 2002.Gubernatorial, senatorial, and state-level presidential jobapproval:The U.S. officialsjob approvalratings JAR)collection. State Politics and Policy Quarterly 2:215-31.

Bowman, Ann O'm, and Richard C. Kearney. 1986. The resurgence of the states. Englewood Cliffs, NJ: Prentice Hall.Brace, Paul, Kellie N. Butler, Kevin Arceneaux, and MartinJohnson. 2002. Measuring public opinion in the Americanstates: An expanded range of aggregated measures, 19741998. American Journal of Political Science 46:173-89.California Institute of County Governments. 2001. TheCalifornia county fact book. Sacramento: California Instituteof County Governments in conjunction with California StateAssociation of Counties, http://www.cicg.org/data (accessedMarch 10,2003).

Cigler, Beverly A. 1990. County government: A century of change. InTheMunicipal Yearbook1989.Washington C: ICMA, 55-65.Clark, Terry N. 1968. Community structure, decision making,

budget expenditures, and urban renewal in 51 cities. AmericanSociological Review 33:576-93.Clingermayer, James C, and Richard Feiock. 1995. Councilviews toward targeting developmental policy benefits. TheJournalofPolitics 57 (2): 508-20.Cohen, Jeffrey. 2006. Introduction:Studying ublic opinion intheAmerican states. In Public opinion in state politics, ed.

JeffreyE. Cohen. Stanford, CA: Stanford University Press.Connell, Kathleen. 2001. State of California counties annual

report,fiscal year 1998-1999. Sacramento: Office of CaliforniaState Controller, http://www.sco.ca.gov/ard/local/locrep/counties/9899/index.shtmlaccessedApril 17,2006).DeSantis, Victor S., and Tari Renner. 1996. Structure and policy

expenditures inAmerican counties. In The American county:Frontiers of knowledge, ed. Donald C. Menzel. Tuscaloosa:The University ofAlabama Press.Duncombe, Herbert S. 1977. Modern county government.

Washington, DC: National Association of Counties.Dye, Thomas. 1979. Politics and economics: The development ofthe literature on policy determination. Policy Studies Journal

(June): 652-62.Erikson, Robert S. 1978. Constituency opinion and congressionalbehavior: A reexamination of theMiller-Stokes representationdata. American Journal of Political Science 22:511-35.Erikson, Robert S., Gerald C. Wright, and John P.Mclver. 1993.Statehouse democracy: Public opinion and policy in theAmerican states. Cambridge: Cambridge University Press.Eulau, Heinz. 1973. Ixibyrinths of democracy: Adaptation, linkages,

representation, and policies in urban politics. Indianapolis, IN:Bobbs-Merrill.


15/15


Feiock, Richard C, and Jonathan West. 1993. Testing competingexplanations or local policy adoption:Municipal solidwasterecycling programs. Political Research Quarterly 46 (2): 399-419.

Harvey, A. C. 1976. Estimating regression models with multiplicative heteroskedasticity. Econometrica 44:461-65.

Hawkins, BrettW. 1971. Politics and urban policies. Indianapolis,IN: Bobbs-Merrill.Hero, Rodney E. 1998. Faces of inequality: Social diversity inAmerican politics. New York: Oxford University Press.

Hill, Kim Quaile, and Angela Hinton-Andersson. 1995. Pathwaysof representation: A causal analysis of public opinion-policylinkages. American Journal of Political Science 39:924-35.

Hill, Kim Quaile, and Patricia A. Hurley. 1984. Estimating congressional district attitudes with national election study data:A reliability assessment. Political Methodology 10:447-63.Hill, Kim Quaile, and Jan . Leighley. 1996.Political partiesandclass mobilization in contemporary United States elections.American Journal of Political Science 40:787-804.

Johnson, Martin, Paul Brace, and Kevin Arceneaux. 2005. Publicopinion and dynamic representation in the American states:The case of environmental attitudes. Social Science Quarterly86:87-108.

Jones, Bradford S., and Barbara Norrander. 1996. The reliabilityof aggregated public opinion measures. American Journal ofPolitical Science 40:295-309.

Kantor, Paul, and Stephen David. 1983. The political economy ofchange in urban budgetary politics: A framework for analysisand a case study. British Journal ofPolitical Science 23:251-7r4.

Katz, Michael R. 1989. The undeserving poor: From the war onpoverty to thewar on welfare. New York. Pantheon.

Keiser, Lael R., Peter R. Mueser, and Seung-Whan Choi. 2004.Race, bureaucratic discretion, and the implementation of welfare reform. American Journal of Political Science 48:314-328.Key, V. O. 1949. Southern politics in state and nation. New York.Alfred A. Knopf.

Lewis, Paul. 2001a. Looking outward or turning inward?Motivations for development decisions in central cities andsuburbs. Urban Affairs Review 36:296-320.

Lewis, Paul. 2001b. Retail politics: Local sales taxes and the flscalization of land use. Economic Development Quarterly 15:21-35.Lineberry, Robert L., and Edmund P. Fowler. 1967. Reformismand public policies in American cities. American PoliticalScience Review 61:701-16.Lipsky, Michael. 1980. Street-level bureaucracy: Dilemmas ofthe individual in public services. New York: Russell Sage

Foundation.Martin, Lawrence L., and Ronald C. Nyhan. 1994. Determinantsof county charter home rule. International Journal of PublicAdministration 17:955-70.

Miller, Warren E., and Donald E. Stokes. 1963. Constituencyinfluence in Congress. American Political Science Review57:45-56.

Morehouse, Sarah McCally. 1973. The state political party andthe policy-making process. American Political ScienceReview 67: 55-72.

National Network of State Polls. 2007. NNSP: National networkofstatepolls. Last updatedApril 14,2007. Lexington:SurveyResearch Center, University of Kentucky, http://survey.rgs.uky.edu/nnsp/accessedAugust 29, 2007).Norrander, Barbara. 2001. Measuring state public opinion withthe senate national election study. State Politics and PolicyQuarterly 1:111-25.

O'Brien, Robert M. 1990. Estimating the reliability of aggregatelevel variables on individual-level characteristics. SociologicalMethods and Research 18:473-504.

Park, David K., Joseph Bafumi, and Andrew Gelman. 2003.Bayesian multilevel estimation with post stratification:State-level estimates from national polls. Political Analysis12:375-85.Peterson, Paul E. 1979. A unitary model of local taxation andexpenditure policies in the United States. British Journal ofPolitical Science 9:281-314.

Peterson, Paul E. 1981. City limits. Chicago: University of ChicagoPress.Pierce, John, Nicholas Lovrich, and C. David Moon. 2002. Social

capital and government performance: An analysis of twentyAmerican cities. Public Productivity and Management Review25:381-97.Rom, Mark C. 1999. Transforming state health and welfare pro

grams. In Politics in the American states, ed. Virginia Grayand Herbert Jacobs. Washington, DC: CQ Press.Schneider, Mark, and Kee Ok Park. 1989. Metropolitan countiesas service delivery agents: The still forgotten governments.Public Administration Review 49:345-52.Schneider, Saundra K., and William G. Jacoby. 2006. Citizeninfluences n statepolicy priorities: he interplayf public

opinion and interest groups. In Public opinion in the states, ed.Jeffrey E. Cohen. Stanford, CA: Stanford University Press.

Stein, Robert M., Stephanie Shirley Post, and Allison L. Rinden.2000. Reconciling context and contact effects on racial attitudes. Political Research Quarterly 53:285-303.Stimson, James A., Michael B. MacKuen, and Robert S. Erikson.

1996. Dynamic representation. American Political ScienceReview 89:543-65.U.S. Census Bureau. State and county quick facts.http://www.quickfacts.census.gov/qfd/states/06000.html

(accessed March 21, 2003).Uslaner, Eric M. 2006. The civic state: Trust, polarization, andthe quality of state government. In Public opinion in the

states, ed. Jeffrey E. Cohen. Stanford, CA: StanfordUniversity Press.

Weber, Ronald E., Anne H. Hopkins, Michael L. Mezey, andFrank Munger. 1972-1973. Computer simulation of state electorates. Public Opinion Quarterly 36:49-65.Weber, Ronald E., and William R. Shaffer. 1972. Public opinionand American state policy-making. Midwest Journal ofPolitical Science 16:683-99.Weisbert, Herbert F., Jon A. Krosnick, and Bruce D. Bowen.1996. An introduction to survey research, polling, and data

analysis. Thousand Oaks, CA: Sage.Wlezien, Christopher. 1995. The public as thermostat: Dynamicsof preferences for spending. American Journal of PoliticalScience 39:981-1000.Wlezien, Christopher. 1996. Dynamics of representation: Thecase of U.S. spending on defence. British Journal of PoliticalScience 26:81-103.Wong, Kenneth K. 1988. Economic constraint and politicalchoice in urban policymaking. American Journal of PoliticalScience 32:1-18.Wright, Gerald C, Robert S. Erikson, and John P. Mclver. 1985.

Measuring state partisanship and ideology with survey data.Journal of Politics 47:469-89.

Documents

PRQ09-1-Percival-Representation and Local Policy-Relating County-Level Public Opinion to Policy Outputs