Local Governance and Redistributive Policy: Explaining Local Funding for Public Defense

Abstract

In many American states, public defense is provided at the county rather than state level (Langton & Farole 2009). Local governments have discretion over implementing and funding the right to counsel, resulting in considerable variability in programs and funding levels. Placing this issue in the theoretical context of redistributive policies and politics, we investigate decisions on funding this service across upstate New York counties. Using as a point of departure Paul Peterson's classic explication of community politics, we first model variation in funding as a function of counties' fiscal capacity, need for services, and costs of supplying legal representation. We also test Peterson's prediction that local political factors will play little if any role in budget decisions. Second, through interviews with program administrators we explore the characters of twelve defender programs in which expenditures departed from the model's predictions. We find that three factors—which we term “influence,” “infrastructure,” and “ideas”—also vary directly with levels of funding. We conclude with a discussion of the implications of these findings for theoretical thinking about due process policies and local politics, and for policy debate over how best to ensure adequate counsel in criminal court.

Scholars and advocates have long argued that counsel for indigents accused of crime is underfunded (Lefstein 1982; Lucas 2005; Moran 1982; President's Commission on Law Enforcement and the Administration of Justice 1967; Spangenberg 2000; The Constitution Project 2009). Historically, even as a series of Supreme Court decisions expanded the scope of the right to counsel, the Court left responsibility for organizing, administering, and funding indigent defense to state and local governments. The resulting patchworks of policies, experts say, results in services of widely disparate quality, often falling short of Constitutional standards.

In the last 5 years, however, indigent defense has been added to a growing list of criminal justice reform targets. Former U.S. Attorney General Eric Holder has spoken out on the problems facing defenders, and the Department of Justice has allocated funding for research on the topic (Frederique et al. 2015; Herberman & Kyckelhahn 2014; Holder 2012). At the state level, researchers have investigated the quality and costs of defense programs (Anderson & Heaton 2012; Carmichael et al. 2015; Hopkins & Labriola 2014); and lawsuits challenging the adequacy of locally funded programs have been resolved with new resources, and new oversight, for defense providers (e.g., Hurrell-Harring et al. v. State of New York 2014; Wilbur v. Mount Vernon, WA 2013). Yet, there has been little systematic investigation of the economic and political dynamics surrounding the funding of public defense.

Differences in defense services have been linked to changes in case outcomes (Abrams & Yoon 2007; Miller et al. 2015) and clients' satisfaction with the legal process (Campbell et al. 2015). Recent research suggests that funding is important: in Houston, Brooklyn, and the state of Washington better funding was associated with improved service delivery, reduced defender caseloads, speedier case processing, reductions in uncounseled guilty pleas, improved trial outcomes for defendants, and improved defender repute among local judges (Fabelo et al. 2013; Labriola et al. 2015; Luchansky 2010). While we know less about how resources shape less readily quantifiable aspects of court process (such as defendants' experiences, and the cultures of defense programs and courthouses) ethnographic studies have suggested that defense services may be inflected by race (Richardson & Goff 2013; Van Cleve 2016), professional norms (Corbin 2015), and institutional history (Mayeux 2016).

We investigate, across the 57 upstate New York counties, variation in public defense expenditures.¹ We frame our analysis around the key observation that public defense is essentially a redistributive policy—a policy that draws on public resources to benefit disadvantaged populations. While legal scholars and advocates frame indigent defense as a Constitutional right, in practice the provision of counsel is, in many states, a policy problem involving organizational, administrative, and allocational decisions (Davies & Worden 2009; Peterson 1981; Pruitt & Colgan 2010).

We report our findings as follows: In Part I, we describe New York as a state whose indigent defense policies represent the range of conditions that obtain across the nation. We develop hypotheses about the relationships between county characteristics and defense funding based on longstanding theoretical insights about local decisions on redistributive policies. In Part II, we describe the data and measures used to test these hypotheses, and report findings from a quantitative analysis of expenditures. In Part III, we analyze interviews with indigent defense agency heads in 12 counties which were outliers in our quantitative analysis. We conclude by addressing the theoretical and policy implications of our findings.

Part I: Dimensions of Indigent Defense Policy

The Parameters of Indigent Defense in New York and the Nation

The Sixth Amendment to the United States Constitution states that “In all criminal prosecutions, the accused shall … have the Assistance of Counsel for his defense.” The Supreme Court has, over the years, interpreted this to mean that any defendant who is unable to afford to pay for counsel, and who may face the possibility of incarceration if found guilty, has the right to a lawyer free of charge. (Gideon v. Wainwright 1963) and Douglas v. California (1963) established states’ responsibility to provide counsel for indigent defendants facing felony charges and direct appeals of conviction. Subsequent decisions expanded the right to include most misdemeanors, juvenile proceedings, certain pretrial appearances and limited postconviction proceedings. Most states also provide counsel for indigents in some family law cases such as state claims for termination of parental rights, child custody, abuse and neglect, child support enforcement, and domestic violence, though these policies vary across the states (Abel & Livingston 2008; see also the National Coalition for a Civil Right to Counsel 2016).² The Bureau of Justice Statistics estimates around 80 percent of criminal defendants are entitled to counsel at government charge (Harlow 2000; see also Smith & DeFrances 1996). To our knowledge, there are no reliable national estimates of the percentage of family court parties who receive free representation.

New York is one of 22 states that depend on counties to fund public defense (Constitution Project 2009). The counties outside New York City represent the range of social, economic, and political conditions of counties across the nation, and also the diversity of defense program structures and funding levels. Their population per square mile varies from less than 3 to 4,705; median household income in 2010 ranged from $40,179 to $92,450; and in the 2012 Presidential election upstate New York counties were almost exactly split in their support for the two major candidates.

New York has consistently ranked high among states for spending per capita on indigent defense (Davies & Worden 2009), though this is an artifact of the inclusion of the high costs of providing programs in New York City. Among upstate counties, spending on public defense from local funds ranged from $6.79 per capita to $28.50 in 2012, a range similar to that across all states (Davies & Worden 2009).

Like many states, New York gives counties discretion in organizing defense programs, and leaves funding largely up to county governments.³ Counties are directed by NYS County Law §722 to provide indigent legal services in one or more of four ways: a public defender office, a conflict defender office, a legal aid society, or assigned counsel.⁴ Public defender and conflict defender programs typically comprise offices with staffs of salaried attorneys (conflict defender offices represent defendants whose cases present conflicts of interest to the primary program) (DeFrances & Litras 2000). Legal Aid Societies are private not-for-profit corporations which contract with counties to provide legal services; structurally they often resemble public defender offices. Assigned counsel programs draw on panels of privately practicing attorneys who accept clients assigned by judges or administrators, and who bill the county at fixed hourly rates.⁵ There are few policies that standardize programs across counties. In short, the scope of the right to counsel has expanded in the past 50 years, but the responsibility for organizing and funding it remains highly decentralized.

The Economics and Politics of Redistributive Policymaking

Each of the 50 states in the United States is divided into counties, or in some places “boroughs” or “parishes.” These comprise over 3,000 separate administrative units, covering the entire territory of the country, the vast majority of which have some form of government, typically a locally elected board or commission. These local governments have considerable responsibilities for maintaining local infrastructure such as transportation, public safety, and social services programs, as well as legislatively and constitutionally created functions such as law enforcement, prosecution, adjudication, and criminal punishment (Benton 2001). Yet they are also obliged to maintain balanced budgets; their actions are regulated by state laws; and they are limited in their powers to tax, make policy, and regulate behavior. In short, county governments have many obligations, but limited discretion, regarding the programs for which they collect and spend money. As a result, county officials must make tradeoffs among competing demands.

An early generation of public policy researchers theorized that local governments are under pressure to strategically allocate their limited resources to protect their communities' economic viability and potential for growth (Dye 1966; Hofferbert 1970; Lineberry & Fowler 1967).⁶ While much of this theorizing involved public ideology and political culture, (Paul Peterson 1981) argued that while ideology and culture might drive policy decisions (and resource allocations) at state and national levels, local officials were constrained by economic imperatives that rendered political disagreements somewhat irrelevant. We frame our attempt to account for variation in indigent defense funding with Peterson's theoretical argument.

Peterson distinguished between developmental, allocational, and redistributive policies. Developmental spending goes for projects that increase the attractiveness of the community to taxpaying residents and businesses. In competing with similar counties, officials seek to create favorable conditions for business, industry, workers and residents, in hopes that this spending may pay off in greater public revenues and healthier economies. Allocational spending maintains critical infrastructure and provides services to residents, such as police and fire protection, transportation systems, and schools. Redistributive policies provide assistance to residents or groups who cannot secure critical resources themselves; as such, they neither maintain nor grow local economies but instead drain them of resources.⁷

Peterson hypothesized that three sets of economic factors shaped expenditures for these three types of policies, although not in identical ways. Fiscal capacity is the level at which a community can generate resources, primarily through taxation. States place caps on counties' ability to tax property, sales, and income, and hence fiscal capacity is largely a function of the wealth of its residents and businesses. Demand is the level at which a service or product is sought. It fluctuates with the cost of that service or product; hence a community's excellent parks, schools, and bus systems will be attractive only insofar as its taxes do not discourage residents and businesses from living there. Peterson further distinguished between demand as defined in this way, and need in the context of redistributive policies: need is the number of people who are eligible for essential services, conditioned by how much of those services they draw. Finally, the cost of supplying services or products varies with the market price of labor, material, and infrastructure as well as with the effort needed to make services accessible to beneficiaries.

Officials' investments in developmental policies are shaped largely by demand: policies that attract more interest will get greater investments. These expenditures will be somewhat constrained by the costs of supplying these programs, but little affected by fiscal capacity (since these are investments in future growth so even poor communities have incentives to fund them). Allocational policies are moderately shaped by all three sets of factors, since they are provided to the general public and are more often necessary than optional. Redistributive policies, Peterson argued, present a different calculus. He suggests that demand is not relevant, insofar as these policies provide near-essential benefits to people who cannot contribute much to the economy; they represent instead varying levels of need. But he further suggests that (even in 1981) because federal and state governments provided so much of the resources for redistributive policies (such as unemployment compensation, welfare, and disability benefits), neither the level of need nor the costs of supplying it would play much of a role in such policies. Instead, fiscal capacity would determine the rates of local investments in these services. State or federal funds, for example, might pay for free school lunches for children, but a wealthy county might find the funds to cover a neighborhood lunch program for children during summer vacation. Other scholars offer further explication (Hajnal & Trounstine 2010; Lieberman & Shaw 2000; Sapotichne et al. 2007; Stone 2006), suggesting that wealthier county governments—those with higher per capita tax revenues—have more slack resources and may feel less pressure to minimize redistributive program costs. They may even see merit in more generous programs for their least advantaged residents.

Missing from this framework, of course, are political and cultural explanations. Peterson does not overlook the possibility that organized interests, voters, and parties might take an interest in local policy, but he predicted that their preferences would have limited impact on outcomes, particularly regarding redistributive policies. Parties more accurately reflect national than local issues; citizen participation in local politics is limited and sporadic; and local concerns seldom crystallize into organized groups (1981: 115–129).

How might these propositions apply to public defense? Public defense is an essentially redistributive policy: it allocates critically needed assistance for people who must pass a means test. But some of Peterson's stipulations do not clearly apply to public defense. First, in many states public defense remains largely a county responsibility, financed primarily with tax revenues collected from local residents and businesses. Second, unlike school lunches, the services provided are not uniform, so supply costs vary across counties as well as across clients: a felony case defense is more demanding than a misdemeanor case. Third, many redistributive policies benefit sympathetic populations, such as elderly and disabled residents, children, and laid-off workers but indigent defense provides legal counsel—an expensive commodity—to an unpopular population, people accused of crimes. Despite early theorists’ expectations that ideology and politics would play little role in the tight budgets of cities and counties, research on an array of local redistributive policies suggest that political ideology can have least modest effects on programs and funding (Hajnal & Trounstine 2010; Kelleher & Yackee 2004; Martin 2001; Percival et al. 2008; Wald et al. 2001). Furthermore, public defense, although a redistributive policy, also lands squarely in the criminal justice policy arena. There is some evidence that jurisdictions with more conservative populations, and those with larger minority populations, adopt punitive practices in law enforcement and sentencing because community residents, leaders, or both feel more compelled to sanction offenders, particularly those upon whom they impose racial stereotypes (Garland 2001). With these caveats in mind, we propose the following hypotheses, aligned with Peterson's theory.

Fiscal Capacity

All else equal, taxable resources will predict expenditures on indigent defense, through two possible mechanisms: wealthier counties may elect to provide more generous funding than absolutely necessary, while economically stretched counties will reduce expenditures to the bare minimum.

H1: Counties that generate higher tax revenues per capita will spend more on public defense, per case.

Need

The need for indigent defense is best measured as a weighted combination of the number and seriousness of cases that reach the courts. Total caseload—a variable that is largely beyond counties’ control—should emerge as a substantial correlate of absolute expenditures.⁸ However, one should expect that, adjusted for population, caseload should be negatively correlated with expenditures. If two otherwise identical counties have different caseloads (and similar resource levels), the county with the higher caseload must spend less per case, because caseloads must be processed within available resources, not the other way around, particularly when the services provided do not have mandated or unavoidable costs per unit (or client).⁹

H2: Counties with heavier caseloads (and hence greater need) will spend less on public defense, per case.

Demand for public defense must be balanced with demand for other county-supported social services when resources are finite. Counties with different proportions of disadvantaged populations face different demands on all redistributive policies, and hence fewer resources available for each program. We predict that counties with larger economically disadvantaged populations will spend less on public defense, all else equal, because they face higher levels of demand for other mandated social services (see Craw 2010; Martin 2001).

H3: Counties whose populations include a larger proportion of disadvantaged residents will spend less on public defense, per case.

Cost of Supplying Services

We examine three dimensions of the costs of supplying defense: the market value of legal expertise, the burden of providing services to dispersed populations and courts, and the differences in costs based on structure of service delivery and program types. Indigent defense programs purchase lawyers’ services in local markets, whether they are paying lawyers on a case-by-case basis (as would be the case in an assigned counsel panel program) or maintaining a public defender office. Hence programs may have to pay more to attract attorneys in counties where attorneys can earn higher salaries in the private sector. Although assigned counsel systems are guided by legislated state standards for hourly rates of reimbursement, judges retain discretion in approving vouchers for time worked and services rendered. Prior research suggests the amounts attorneys claim for similar cases varies considerably across counties (King, Unpublished report). Moreover, the salaries for attorneys in public defender and legal aid programs are not fixed by state law. Accordingly, we expect that in counties where attorney incomes are higher, spending on public defense will necessarily also be higher. Hence:

H4: Counties in which attorneys, overall, earn higher salaries will spend more per case on public defense services.

Providing counsel usually requires in-person contact and consultation. Arrestees hauled into court for arraignments must hope their assigned lawyers will get to the courthouse on time to confer with them; jailed defendants must wait for their lawyers to visit them; and otherwise clients usually must travel to attorneys' offices to discuss their cases. Hence defense programs are constrained by the geographical dispersion of clients, courthouses, and providers.

New York's court system represents the complex patchwork of jurisdictions found in many states (National Center for State Courts 2010). County Courts handle most felony and grand jury matters. Family Courts handle custody, visitation, and support claims, as well as some family abuse cases. Upstate New York's 61 cities have City Courts, whose elected judges handle preliminary felony hearings, misdemeanors, and violations. But outside those city limits, misdemeanor cases, felony arraignments, and bail hearings are heard by 2000 part-time lay magistrates who preside over 1,200 village and township justice courts. Their jurisdictions range from a few hundred acres to 450 square miles, with populations that range from less than 40 to over 750,000 (Office of the State Comptroller 2010).

This diversity has implications for defenders, who must appear in courts scattered throughout their counties. This need to travel long distances (often during evenings and weekends, when justice courts meet) reflects a challenge that (Pruitt & Colgan 2010) refer to as spatial inequality. Where populations are concentrated in cities, law offices are efficiently located near city courthouses. Where populations are scattered across suburban and rural jurisdictions, however, attorneys must travel among several courts, which may be separated by significant distances and may hold sessions outside ordinary business hours. Hence, we hypothesize that geographic dispersion of courts and populations increase the costs of supplying representation.

H5: Counties with more geographically dispersed courts and populations will spend more on public defense, per case.

As indigent defense caseloads grew in the 1980s and 1990s, so did debate about the relative costs and quality of program types. Advocates have argued the merits of public defender offices over assigned counsel, and although researchers have seldom found significant differences in case outcomes (Hanson et al. 1992; Hartley et al. 2010), many argue that public defender offices foster shared expertise and greater commitment to clients than is obtained in assigned counsel programs (Bohne 1978; Emmelman 1993; McIntyre 1987; Sudnow 1965). Whether public defender programs cost more, or less, than assigned counsel programs is an open question: the few studies that directly address this have produced mixed findings (see, e.g., Cohen et al. 1983; Grier 1971; Houlden & Balkin 1985; Worden 1993). Moreover, simple distinctions between program structure may not adequately capture the many factors within each category that may inflate or decrease costs (Anderson & Heaton 2012).

That said, we have some evidence that upstate New York county officials believe that public defender offices may prove more economical: when the state legislature raised the hourly rates for assigned counsel programs in 2004 a number of counties replaced their assigned counsel programs with public defender offices (Spangenberg Group, 2006b: Appendix L). Hence, we hypothesize tentatively that public defenders (and legal aid societies, which are structurally similar) may prove more cost-efficient than traditional assigned counsel programs. Though public defender programs pay for salaries, benefits, overhead, management and support services, assigned counsel programs may not benefit from the economies of scale and ability to absorb higher caseloads. We also note that most counties have several programs, often of different types, both to deal with conflicts and sometimes to specialize in different types of cases. Hence:

H6: The percentage of a county's caseload that is handled by assigned counsel lawyers is associated with higher expenditures per case.

County Political Values

Though Peterson doubted the effect of local political values in the adoption and funding of redistributive policies, he did so with the understanding that most such policies were largely subsidized by state and federal programs, and he did not address the possibility that communities would judge beneficiaries as undeserving. The public defense function serves a population of destitute people accused of crimes—a ready target for conservative politicians who would prefer to spend resources on services for law-abiding citizens and development. We note as well that racial animus can steer state and local criminal justice policy and practices (Jacobs & Tope 2007; Soss et al. 2001; Stucky et al. 2005). The racial threat hypothesis states that where minority populations are higher, criminal justice practices will be more severe, a function of whites’ anxiety about minorities.

H7: Counties that are more politically conservative will spend less, per case, on public defense.

H8: Counties with larger minority populations (per capita) will spend less, per case, on public defense.

Summary

We expect county indigent defense expenditures to be shaped by the budget dynamics that Peterson considered central to redistributive policies, but we also consider the ways in which public defense differs from many other redistributive policies. We report the findings of a quantitative analysis of these factors in Part II below.

Part II: Analysis of County Contextual Factors

Data and Measures

To test the hypotheses outlined above, we used data compiled from publicly available sources and 2012 defense program budget information collected by the (Office of Indigent Legal Services 2013). The unit of analysis is the county, excluding the five boroughs of New York City. Table 1 reports descriptive statistics for these counties.

Table 1. Characteristics of New York Counties and Their Indigent Defense Programs

The Dependent Variable: Expenditures per Case

The primary dependent variable, expenditures, could be measured in several ways: total local expenditures, expenditures per capita, or expenditures per case. Our interest here is how much money county officials make available to fund programs, adjusted for caseload, so the dependent variable is measure of dollars spent per weighted case.¹⁰ In 2012, the Office of Indigent Legal Services collected program level data, which included counts of cases of all types (violation, misdemeanor, felony, and appeals; and family court cases for which state law provides a right to counsel, see N. Y. Family Court Act, §262).¹¹ We were fortunate to have access to detailed data on caseload, disaggregated by case types, and expenditures, from the 129 programs operating in the 57 upstate counties. In 2012, total expenditures at the county level ranged from $103,999 to $18,218,415; in about a third of counties, less than $700,000 was spent, and in another third spending was over $2,000,000.

Counties that process larger volumes of complex cases such as felonies and appeals will likely spend more per case; and Family Court cases must also be weighted appropriately. Hence, we concatenated data across all programs, within each county. In order to derive a standard measure of expenditures per case, we weighted cases in accordance with national standards for defender workloads. National standards recommend that felonies be counted as equivalent to 2.67 misdemeanors, and appeals be counted as equivalent to 16 misdemeanors (National Advisory Commission 1973, standard 13.12; see Office of Indigent Legal Services 2013). In keeping with guidelines adopted elsewhere, we counted a Family Court case as the equivalent demand, in terms of effort, of a felony case representation (Office of Indigent Legal services 2013). Hence our measure of expenditures per case is expenditures per weighted case:

$\frac{Total county expenditures}{\left(\# misdemeanors\right)+\left(2.67 x \# felonies\right)+\left(16 x \# appeals\right)+\left(2.67 x \# family court cases\right)}$

Using this formula, we found that counties spend an average of $246 per misdemeanor case (ranging from $118 to $529), or $656 per felony or family case (ranging from $315 to $1412).¹² On average, counties provided representation for approximately three misdemeanors for every felony; appeals were rare. It was not possible to isolate criminal cases from family cases for the purposes of estimating any real disparities in expenditures per case, since many programs handle both and expenditures were reported at the program, not case or case category, level. These expenditures are of course averages, and surely mask considerable variation in actual effort and time spent on cases. However, these figures differ starkly from the price of retaining counsel. One recent estimate for private representation in a misdemeanor case ranged from $500 to $4000, and from $1200 to $10,000 for felony representation (Bergman & Bergman 2013). In one county, attorneys agreed that the fee for a petty larceny would be at least $1000, and the going rate for an “uncomplicated” DWI would be between $1000 and $2500.¹³

The Independent Variables

County Revenue

In New York, as in many states, counties’ tax revenues are largely determined by sales and property taxes, and the state imposes some caps on both of these. Counties with more valuable properties and businesses, and healthier economies, generate more revenue per capita. We measured county resources as county tax revenue per capita (New York State Comptroller's Office 2015).¹⁴ The wealthiest county, by this measure, takes in 2.5 times the revenue per capita as the poorest county.

Caseloads

We expect that counties’ expenditures per case will depend on both the volume and complexity of caseloads. We initially calculated separate measures for criminal and family caseloads. In 2012, criminal court caseloads (weighted as described above) range from 16 to 79 per 1000 residents, with a mean of 41; Family Court caseloads range from 6 to 100 per 1000 residents, with a mean of 30. New York counties, it appears, are on average handling roughly equivalent numbers of criminal and family cases. Interestingly, however, criminal and family court caseloads are correlated at only. 398, suggesting that the dynamics that produce impoverished criminal defendants are somewhat different from the dynamics that produce indigent clients in eligible civil proceedings. Hence, we initially included two caseload variables, criminal caseload per 1000 population and family court caseload per 1000 population. Preliminary analyses indicated that these two variables had very similar coefficients, and that when they were combined into an aggregate weighted caseload measure the effect was approximately double that of each individual coefficient, so we have included the aggregated measure in results reported below.

Disadvantaged Population

We measured economic disadvantage with a factor scale of four 2010 census variables: percent population above poverty; percent population with health insurance; median income per household; and percent residents holding at least a high school degree (alpha = 0.71). The scale was inverted, such that higher values indicate more distressed populations. Where a larger proportion of the population is economically distressed, more demands will be placed on services for children, the unemployed, elderly, and ill.

Market Value of Legal Services

We obtained attorney salary data from the Bureau of Labor Statistics and Census data, published by the American Bar Association (American Bar Association 2011). Median values for attorneys’ self-reported salaries, at the county level, ranged from $72,150 to $156,580. Not surprisingly, median salaries were statistically associated with the overall economic well-being of the county and with proximity to New York City, where the cost of living is significantly higher than in counties farther upstate.

Spatial Inequality

We explored two common measures of geographic dispersion of service needs: population per square mile and percentage of population living in incorporated cities. Population per square mile does not, in upstate New York, completely capture an urban-rural distinction, since many counties’ populations, while highly concentrated, are not in cities. Concentration of city population is also of limited use, since many low-population counties have significant proportions of their populations in scattered but very small cities.

We elected instead to use a measure of population concentration that was developed specifically to capture the challenges of providing public services such as health care to populations in areas of varying concentration (Serdar & Darling 2012). The Economic Research Service of the U.S. Department of Agriculture's Office of Rural Health Policy developed the Rural-Urban Continuum Codes (a nine-point scale) to distinguish among counties that are uniformly metropolitan, counties that exist in metropolitan areas of varying population sizes, counties that have significant urban areas but are not in metropolitan areas, counties with small cities that are adjacent to metropolitan areas, and counties that have small city populations and are not near metropolitan areas, and those that are completely rural and are, or are not, adjacent to a metropolitan area (U.S. Department of Agriculture 2004). In New York's upstate counties, the most common scores (1, 2, and 3, representing more spatially condensed populations) included counties with between 0 and 50 percent city populations. Several counties in these categories were densely populated exurban places with few incorporated cities. Scores 4 and 5 had city populations ranging from 24 to 38 percent. Most counties with scores 6 through 9 had no city populations at all, with two counties coming in at 14 percent and 22 percent, respectively. We dichotomize this scale, such that rural counties are those that have scores of 4 or above, indicating counties with small cities or even lower concentrations of population, outside metropolitan areas.

Program Structure

We operationalized program structure as the percentage of weighted caseload that is handled by assigned counsel (rather than public defender or legal aid society) programs, and we use that variable to test the hypothesis that assigned counsel programs are associated with higher per-case costs. Ten counties used assigned counsel as the primary means of providing representation; in eight it was the only program. The remaining counties had a public defender or legal aid society as the primary provider of representation, supplemented by an assigned counsel program, with some places also employing a conflict defender. Over a third of counties fund three or more distinct programs.¹⁵ By quantifying the percentage of a county's caseload handled by assigned counsel, which ranges from 3 to 100 percent, we captured the degree to which counties relied upon that approach for providing representation, allowing us to control for the additional costs that we hypothesized might result.

County Political Climate

We measure political culture with voting data from a polarized gubernatorial election in 2010, one that highlighted sharp differences in crime and justice policy debates. In this election, Carl Paladino ran as a Tea Party candidate, campaigning against state and local government excess but in favor of greater criminal justice enforcement. He was opposed by Democrat nominee Andrew Cuomo, who espoused reduced use of incarceration, closing of some prisons, and greater attention to righting erroneous convictions. Although Paladino carried only 11 upstate counties, his support at the county level ranged as high as 69 percent (Leip 2013).¹⁶

We measure racial threat as the percentage of the population that reports as “nonwhite” in 2010 Census data. New York counties range from 3 to 31 percent nonwhite according to the 2010 census.

Analyses and Results: Correlates of Program Expenditures

Table 2 reports the results of OLS regression analysis of the dependent variable, expenditure per weighted case.¹⁷

Table 2. Correlates of Public Defense Expenditures in New York Counties

Coefficients statistically significant at the ^*.10 level; ^**.05 level; ^***.01 level.

R2 = .620 n = 57.

This model explains 62 percent of the variance in county spending per case, and it provides significant support for Peterson's central thesis about redistributive policies.¹⁸ The highest tax-generating county spends about $100 more per case than the county with the lowest tax revenues per capita. The county with the most economically disadvantaged population spends about $80 less per case than the least challenged county.

The more cases a county receives (and the more serious those cases are), the less it spends per case. Statistically, adding 133 additional cases per 1000 population (the differential between the lowest and highest caseloads) translates into $187 less in expenditures per case.

We also found, however, that variables measuring the costs of supplying legal counsel predict expenditures. Counties in which lawyers enjoy the highest range of salaries spent significantly more per case than those where lawyers get by on the lowest range of salaries; rural counties spend about $50 more per case than those coded as urban and metropolitan. We also found some support for policy makers' intuition that assigned counsel, under current laws, cost more than public defender offices. This finding may seem somewhat counterintuitive: after all, defender offices should cost more given a salaried staff including not only lawyers but also, perhaps, investigators, paralegals, and interns, as well as benefits and overhead. First, perhaps public defense offices are indeed more efficient: they provide similar services at lower cost per case, maybe because their lawyers have more criminal casework experience, better relationships with other court actors, or more streamlined protocols for processing cases. Second, more pessimistically, perhaps they invest less effort in each case, consistent with early stereotypes of defenders who serve the courts’, not their clients’, interests, or who suffer under such high caseloads that they ration little time for each case. Third, perhaps the mechanisms for funding institutionalized programs and assigned counsel programs are different in ways that have implications for budget over-runs. We found evidence that some public defense offices receive fixed allocations each year (and are expected to stay within those budgets). Assigned counsel programs, in which a number of attorneys and firms submit vouchers as cases are concluded, may run out of money before the fiscal year ends, which requires that counties authorize supplementary appropriations.

Within these parameters, county politics and culture, as measured here, play a negligible role, as Peterson predicted. Counties with larger nonwhite populations spent marginally more on indigent defense, contradicting the prediction of the racial threat hypothesis, but this effect is weak and does not reach statistical significance. There was no significant difference in expenditures for politically conservative counties, net of all other variables. (We also note that if these two variables are removed from the model the remaining effects do not change in substance or significance.) Although public defense does not square with some of Peterson's descriptions of redistributive policies—it is a county burden, and its beneficiaries are not well regarded by the public—it does not appear to vary with ideological leanings in upstate New York.

While estimates of effect size are, of course, only statistical approximations of real-world differences, we note that the magnitudes of these effects are substantial in the context of the actual ranges of expenditures. Further, for most of these counties, this model predicted expenditures are between 80 and 120 percent of actual expenditures per case. But we also recognize that these variables represent relatively fixed constraints: counties’ resources, caseloads, labor costs, and population dispersion are quite stable over time. Previous research suggests that courthouse and bar association politics, relationships among court actors and county authorities, and program leadership steer outcomes as well. These dynamics do not readily lend themselves to quantified measurement, but they may be instructive for theoretical and policy perspectives. Hence, we turn to an investigation of defenders’ accounts of their experiences with the budget process.

Part III—Exploring Unexplained Variation in Spending: Influence, Infrastructure, and Ideas

Given the constraints of economics on indigent defense expenditures, what can we learn about the differences between programs which spent significantly more, and others significantly less, than those who were similarly situated? Following (Yin's 2011) “two-tailed” strategy, we explored this question in 12 counties that were significant outliers: six spent at least 20 percent less than the regression model predicted, and six spent at least 20 percent more.¹⁹ Our investigation draws on an archive of interviews conducted with program administrators by ILS staff in 2011–13. These interviews were undertaken to learn about respondents' experiences in managing their programs, and followed a protocol including questions about basic program operations (e.g., staffing levels, case assignment processes) and about the obstacles they faced in performing their duties.²⁰ Interviewers took notes and prepared memoranda of the conversations. These memoranda provide information about the ways in which defense administrators navigate their political and economic environments. We analyzed memoranda detailing 49 interviews conducted with program staff in the 12 outlier counties, and we summarize our findings below.²¹

In the six counties that spent significantly less than predicted, four programs identified themselves as public defenders. Two were public agencies (the others were private organizations with renewable county contracts). The other two counties relied primarily upon assigned counsel programs and spent, respectively, $216 and $233 per case, on average. Of the six counties spending at higher levels than the model predicted, five provided counsel primarily through public defender offices; one relied entirely on an assigned counsel program.

We identified three dimensions on which the two groups of defender organizations differed which we term influence, infrastructure, and ideas. As with any exploratory inquiry, we present these findings cautiously, noting that they are based on a small number of counties, and our observations and inferences based on interviews with defense administrators and not others with stakes in this policy in these counties.²²

Influence

We define influence as administrators' voice in political and budgetary decisions that affected their programs. In the counties that outspent their peers, there was evidence that local decision processes routinely included agency heads' input and concerns. Several occupied ex-officio positions on local committees. Others described the importance of the close relationships they had forged with local criminal justice and political actors which allowed them to insert themselves into policy discussions and assure their agencies’ voices were heard.

These administrators reported that their programs were treated even-handedly by county government, largely based on professional grounds. Three of the five chief defenders reported that their lawyers' salaries were at or near parity with those in the District Attorney's offices; the others described salaries as adequate or even, in one case, “generous.” One noted that office staffing levels had been developed in consultation with national caseload standards, in contrast to the many agencies across the state with caseloads that far exceed such standards (Office of Indigent Legal Services 2013). One provider reported that county government had “been good to us,” and another that it had been “very supportive,” when considering budget requests. The treatment providers received could be changeable of course: one defender noted that the county had also engaged in budget strategizing, allocating new charges for rent, utilities, security, and IT services to the program in order to create an appearance of continually rising appropriations over the years. Despite challenges these providers faced, their programs appear to have benefited from active involvement in budget decisionmaking and from courthouse policies premised on equal standing.

The contrast between these counties and those that spent less than predicted was stark. In the programs that contracted to provide services, administrators reported that because they had to submit bids to keep their contracts, they were not on the inside of the budgeting process, and recent bidding contests were fresh in their memories. The county with a public defender office, had recently cancelled the agency head's employment benefits and had ceased to pay vouchers for services such as investigators and expert witnesses. In two assigned counsel programs, administrators expressed continuing fear that county officials would adopt competitively bid contract programs in order to save money.

Rather than negotiating with county governments, therefore, these agencies competed for funding by strategizing to reduce costs. Two such programs were established specifically to cut costs by replacing assigned counsel programs (both in 2004, when, not coincidentally, the state raised assigned counsel fees). The fear of being undercut, and the need to battle for contract renewals, was present across these programs. Meanwhile, agency directors believed that the counties themselves preferred contracting with providers precisely because of the predictable, fixed fees that contracts paid, and the absolution they offered from fringe benefit expenses.

Infrastructure

Infrastructure encompasses staffing, specialization, oversight, and capacity to meet unpredictable needs. We noted important differences in infrastructural development between agencies that spent more, rather than less, than predicted. Those in the former group reported having sizeable support staffs including investigators, administrators, secretaries, and paralegals. One reported the county maintained a discretionary fund the defender could access for cases requiring expert testimony, and that the agency's favorable relationship with the county executive guaranteed access to the fund. In the one assigned counsel county among these six programs, defenders reported that attorneys faced few obstacles when requesting financial support for expert witnesses and other necessary tools from local judges, and they attributed that ease of access to the “progressive politics” and “intellectually liberal” judges and attorneys emerging from the local university law school. In short, the political environment in these counties, while not always generous toward defense, provided openings for defenders to make such requests, and extended trust to those defenders to use those resources wisely.

Structurally, all five institutional defenders had full-time attorney staffs; one administrator opined that full-time contracts were necessary for professional commitment to the work. Three of those offices had dedicated appellate attorneys on staff, available for consultation with trial attorneys—a rarity in New York

By contrast, all programs in the counties spending less than expected reported that they lacked basic infrastructure. Three of the six programs had no central office at all, but instead ran on a part-time basis out of program directors' private law offices. These “agencies” existed on paper only, as decentralized service providers operating as groups of “independent contractors.” In two counties, the agencies' practice was limited to criminal representation, leaving family court representation without any organizational oversight and entirely at the discretion of local judges. Others reported their programs provided no oversight or supervision for attorneys, did not provide or require any training for attorneys, and, in some cases, operated within the context of known conflicts of interest. Agencies in three counties reported that they had no support staff at all.

Three of the heads of the four public defender agencies in this group reported that their attorneys were part-time (which allowed them to take cases on private retainer in addition to their publicly funded work). Two program administrators defended the use of part-time attorney staff specifically on the grounds the salaries paid by the county were so low that no attorney could take a full-time position with such meager pay and the caseloads would be impossibly high. While part-time attorneys were pulled between providing services to their paying clients and their indigent ones, program administrators explained the advantage of employing them was that they brought their experience of private practice to the table, kept the caseloads of each attorney at a more manageable level, and reduced salary bills to levels the county would accept.

More rarely, administrators claimed that despite the many disadvantages their program faced, they provided high quality services, by demanding high standards from program attorneys and maintaining a credible threat of trial against prosecutors (whom, several acknowledged, were as underfunded as defense), all in spite of their poor funding. One stated that the program maintained quality services through reliance on attorneys willing to work more hours than they were paid, and without reimbursement for overhead—in short, through reliance on a pro bono ethic, anachronistic as that might be in the context of the 6th Amendment.

Ideas

Indigent defense can be managed as a routine case-by-case function, but it can also be imagined as a broader opportunity for continuous improvement and experimentation, a dimension we term “ideas.” All the leaders in the agencies which spent more than expected focused on seeking ways to improve the quality of representation they provided through policy change, systemic reform, and case-by-case advocacy. Several discussed strategies they had used to address judicial or district attorney practices which systematically disadvantaged their clients. For example, one provider reported that the District Attorney had adopted an “open file discovery” policy—a practice which guarantees timely disclosure of evidence to the defense, and yet which is not required under New York law. Others had adopted practices directly aimed at improving the quality of representation, such as “vertical representation” whereby a client is represented throughout her case by the same attorney (a practice commended for its positive effect on attorney-client relationships); regular training seminars (including, in one case, training which took place out-of-state); and consistent supervision of attorneys, albeit with varying levels of formality.

Administrators in these counties spoke about acting as guardians of the integrity of criminal justice systems, as well as advocates for clients. Each expressed concern about clients who “fell through the cracks,” failing to meet with their attorney or missing out on programs for which they should have been eligible, and sought to put in place procedures to prevent that happening. All expressed interest in improving capacity to collect and analyze data in order to assess and address systematic weaknesses. One dedicated a staff member to implement and utilize a high-quality case-management system; another considered commissioning a study of a specific program; and a third was committed to developing a data system despite, she confessed, “getting a C in undergrad stats.”

In the counties spending less than expected, however, administrators adopted policies that minimized the services they provided. One program leader accused another of failing to visit clients housed at the county jail, and also of failing to take any cases to trial. Another, responsible for an assigned counsel program, refused to provide attorney training because he believed that level of supervision might oblige the county to provide employee benefits to them (which it was unwilling to provide). Administrators also reported that they informally capped assigned counsel fees, refused to honor vouchers for ancillary services such as investigators or expert witnesses, and denied counsel to applicants whose income and assets rendered them marginally indigent. One program leader declined to collaborate with service providers to improve the quality of sentencing advocacy out of fear it would ultimately complicate and drive up the cost of representation.

Interviewers described two agency heads as demoralized and isolated; they acted less like leaders than like embattled bureaucrats maintaining—minimum levels of functionality. The other four, however, adopted leadership strategies to cope with the chronic shortage of, and threats to, agency funding. One, described as a “capable administrator,” noted that in over 10 percent of cases assigned attorneys did not submit vouchers for payment, which she considered a cost-saving windfall rather than a reason for concern. Another agency head used his involvement in local politics to protect what small funding his program received, while yet another aggressively pursued additional contracts for the office on the argument that this would save the county money in other ways.

Discussion

Limitations

Below, we draw conclusions from this study for future research, theory, and policy. But we do so cautiously, because we faced limitations in the design and execution of the project. First, we have implicitly equated, at least as a beginning premise, higher quality defense with higher expenditures. But while public defense programs are widely understood to operate well below the costs of quality representation from the private sector, the relationship between cost and quality merits further empirical scrutiny. Spending more might indicate more generous funding—using slack resources to more faithfully adhere to the spirit of adequate and effective defense. At the same time, spending more might reflect a lack of economy, and a reliance on a reimbursement system that does not best serve indigent clients. The analyses presented here cannot resolve this question.

Second, the study is limited to a single state. However, New York is representative of many states insofar as it delegates the organization and funding of this due process right to county governments. Further, New York's upstate counties are diverse economically, demographically, geographically, and politically.

Third, we observe that our data reflect local conditions at a particular point in time. In 2012, New York counties had spent 5 years besieged by economic downturns (which did not affect all counties equally), but had also enjoyed falling crime rates, the reversal of the state's Rockefeller drug laws, and the closing of several prisons. They had also been the target of a significant lawsuit which had shed light on service deficiencies in five counties in particular. In a future study, we hope to analyze the impact of temporal shifts on local funding of defense programs.

Fourth, our interview data came from program administrators. Assessing the full complexity of budget decisionmaking would require additional data on the perceptions of other stakeholders, such as local government executives, criminal justice actors, and even the general public.

Summary of Findings

In this study, we attempted to model the determinants of county funding of indigent defense in 57 upstate New York counties. Consistent with Peterson's thesis, public defense “behaved” in many ways like a redistributive policy, insofar as variation in funding was significantly influenced by counties' tax revenues. Furthermore, counties with more disadvantaged residents spend less per capita on public defense, all else equal—possibly because in those counties defense services must compete with the higher demands those residents place on other types of redistributive programs.²³

We also observe, however, that the greater the need for public defense services, the less counties spent per case. This suggests that equally situated counties provide less generous services (or attorneys work for lower reimbursement per case) when caseloads are higher.

Although Peterson predicted that variation in the costs of providing services would play little role in redistributive funding, we found otherwise. This may be because for this policy, the costs are borne by the counties themselves.

The results of our exploratory analysis of public defender interviews suggest that while stable county characteristics may place some floors and ceilings on budgets, anomalies in expenditures occur in conjunction with observable variation in the local dynamics of institutional arrangements and program leadership. Defender agencies that enjoy more resources than their peers also appear to have secured greater influence in local political circles, built up stronger and more professional infrastructures, and be led by administrators who are willing (and able) to adopt innovations and best practices. This may permit these programs to establish and maintain a culture that supports continuous improvement in the quality of the services they provide to clients as well a strong professional position within the court community. Programs that are significantly under-funded (compared with peers) suffer from infrastructure deficits, constrained and even demoralized leadership, and attenuated relationships with key actors in their counties—conditions that provide little purchase for organizational growth and professionalization. These programs seem to face continual pressure to cut costs. It is important to note that these observations are consistent with Peterson's description of the real politics of local redistributive policymaking: “Where [partisan] politics are at a low pressure, informal channels of communication substitute for formal ones. In these channels the political resources that count are technical expertise, the power of persuasion and the capacity to reason soundly” (Peterson, 1981: 129).

Directions for Future Theory and Research

These analyses raise interesting questions about the budget process, and the implications for quality representation in indigent defense programs. First, budgeting is inherently political—but what kind of politics are at work here, and whose interests might matter? Second, budgeting is incremental—1 year's allocation always begins at the previous year's number—so what can we learn about the opportunities and constraints imposed by that incrementalism? Third, leadership matters—but what selects for, and sustains, leaders and administrators who can successfully navigate local challenges to maintain creditable programs? And finally, at the broader level of accountability, how should society allocate responsibility—especially financial responsibility—for critical due process protections in a federalist system of government?

We observe that county level measures of ideology (and the oft-cited racial threat hypothesis), while frequently invoked in studies of criminal justice policy, offer no real explanatory power in these analyses. Upon reflection, this makes sense: the public knows little about public defense, and is seldom involved directly in budget negotiations; and elected representatives perhaps see indigent defense as less an ideological than a practical problem to resolve. But that does not mean these decisions are not political. It seems likely that programs are responsive to the expectations, constraints, and incentives of courthouse actors, as is more generally true in local budget negotiations (Duncombe et al. 1992; Sokolow & Honadle 1984). Hence, we might ask whether or not the people who develop and apply financial eligibility standards (judges and administrators) are under pressure to minimize costs. What role does the local bar play in advocacy, program design, and oversight (Butcher & Moore 2000; Eisenstein et al. 1987: Chapter 6; Gilboy 1981; Kessler 1986; Stumpf et al. 1971; Worden & Worden 1989)? Particularly where assigned counsel programs exist or are contemplated, how important are indigent defense assignments in the livelihoods of participating lawyers in private practice? And to what extent are administrators motivated, or pressured (by advocacy groups, state agencies, or lawsuits), to adhere to best practices such as representation at first appearances or advocacy for greater use of diversion and rehabilitative programs?

We also note that we have scratched the surface of an historical narration of the development of local policies, but we could not directly investigate this line of inquiry with the data available. The budget process always begins with last year's budget, and that can be either insurance or burden. The association, in our analysis of outliers, between resources and infrastructure is probably best explained by historical processes that allow organizations to lock in resources and create new plateaux for funding. But understanding this sort of institution-building would require detailed understanding of historical opportunities (such as lawsuits, media interest, or high-profile cases) by which administrators might “lock in” (or lose) established gains.

The vast literature on leadership and management has scarcely been applied to public defense (but see Pogorzelski 2003, & Moore 2002). Lawyers are not trained to be administrators, and it may be that in the minds of many, “successful” public defense is identified with aggressive advocacy and trial work and an adversarial public posture. Yet the ability of an administrator to maintain credibility and negotiate relationships may be critical to attracting sustained support from county governments.

Implications for Policy and Practice

The challenges of funding and organizing indigent defense are complex. It is a service that requires a high level of expertise and credentialing, and can be provided only by professionals who are well paid in the private sector. Public defense clients often have myriad challenges beyond their immediate legal problems. Neither these clients, nor their lawyers, are generally respected or understood by taxpayers. It is not easy to build a public defense program that can overcome these challenges, and opinions—but not much data—abound on how best to do that. To the extent that our findings illuminate these questions, we address some of those challenges here.

A key decision, made in New York as in many states at the county level, is how to organize service delivery, and the most common distinction made is between institutionalized (public defender or Legal Aid) programs and traditional assigned counsel programs. While public defender programs are often touted as superior, this judgement may be simplistic. Our results suggest that where public defender programs operate with at least average levels of funding, they have full-time staff, well-established administrators. But where funding is meager, the “public defender” label may disguise nothing more than an assortment of county-contracted attorneys operating largely as independent agents. Our findings therefore suggest a need for new metrics which capture both the soundness of organizations and their capacity to provide high quality representation that includes not only simply processing cases but also communicating with clients, investigating allegations, and strategic advocacy planning.

The disparities that exist across counties, which also exist across states, illustrate a problem that is not just theoretically intriguing but also deeply troubling. In many states, like New York, public defense encompasses a wide array of formal structures as well as informal practices. This diversity can be beneficial, if it permits counties to fashion programs that best suit local circumstances, attorney populations, and caseloads. But these same factors might work to create programs convenient for their providers but not optimal for their clients. To the extent that counties' economic robustness, and the needs of their underprivileged populations, constrain expenditures, as these results indicate, we conclude that disparities in resources is closely tied to local economies. A federalist funding mechanism that almost ensures that some counties will not—or perhaps equally often, cannot—fund programs at adequate levels ensures that the caliber of justice will vary across those counties as well. We note as well that so long as administrators look within their counties for assessment of the quality and health of their programs, they may be unaware of their deficiencies in comparison with other counties' efforts.

These observations highlight a longstanding premise among advocates, policy experts, and even clients themselves: that quality of defense services is, in many places, so deficient as to “make a mockery of the great promise of the Gideon decision” (Constitution Project 2009: 2). Yet in a federalist system, funding for this low-visibility, sometimes politically unpopular constitutional right is, in many states, delegated to local governments, which have limited capacity and authority to raise tax revenues, are constrained by competing demands for social services, and must respond to needy populations in these constrained contexts. While few public officials would doubt the legitimacy of the right to counsel, many correctly observe that in many places it is an unfunded (or underfunded) mandate (Ban 2016). At this localized level, the variability in resources and demands, coupled with potential political pressures and the practical obstacles to timely and consistent service delivery, present a direct challenge to the mandate to provide competent counsel to indigent defendants. The diversity of county defense systems may represent something of the ingenuity of local policy laboratories, but that may come at the price of consistently adequately funded representation in the criminal courts.