Introduction

Over 65% of defendants in the United States (US) are unable to afford their own counsel, and thus have an attorney given to them by the court (i.e., referred to as “indigent defendants”; Harlow, 2000).Footnote 1 March 18, 2023 marked the 60th anniversary of the milestone Supreme Court case Gideon v. Wainwright, which represented a critical turning point in the American criminal legal system and established the constitutional right to counsel for every defendant regardless of socioeconomic status. The percentage of defendants in need of court-appointed counsel provided by Harlow (2000) more than 20 years ago is, unfortunately, still the most recent comprehensive data available. The US Department of Justice, Office of Justice Programs, reported in 2007 that just under 1000 “public defender offices nationwide received nearly 6 million indigent defense cases” (OJP, 2011). Indeed, with rising economic inequality, it is unlikely the number of defendants requiring public legal assistance has decreased. Since the 1970s the distribution of wealth in the US has become increasingly more concentrated to top wealth-holders, while the proportion of wealth held by 90% of the US population has steadily decreased (Bor et al., 2017; Saez & Zucman, 2016). The increasing wealth disparity is even more pronounced for individuals involved in the criminal legal system, who tend to have significantly lower annual income compared to the general population (Rabuy & Kopf, 2016). As wealth disparity continues to rise, public defenders also continue to face caseloads that are well above the national caseload guidelines published in 1973 (Beeman & Buetow, 2023). Despite the high percentage of indigent defendants in the US criminal legal system, and the six decades that have now passed since Gideon v. Wainwright, there is still much that is not widely known about their experiences compared to defendants who do not have indigent status. This meta-analysis is the first attempt at bridging gaps in the empirical knowledge regarding criminal legal outcomes for indigent defendants in criminal cases as a function of attorney type.

Among the existing empirical literature on outcomes for indigent defendants in criminal court cases, some research suggests indigent defendants experience more negative criminal legal outcomes than their non-indigent counterparts. To demonstrate, some studies show that indigent defendants have higher odds of being held pretrial, pleading guilty, conviction, longer custodial sentences, and very long delays in their legal proceedings compared to defendants who can afford their own counsel (Heaton et al., 2017; Linhorst et al., 2017; Roach, 2017). Conversely, other studies find no differences in bail, plea bargaining, and sentencing decisions between indigent defendants and defendants with a retained (i.e., privately hired) attorney (Hartley et al., 2010; Harlow, 2000). Studies’ incongruency in results and interpretation of findings is reflective of research that was isolated to specific jurisdictions, populations, and court processing outcomes. The disjointed nature of individual research studies examining individuals with indigent defense status currently limits our overall understanding of outcomes experienced by indigent defendants and underscores the need for a systematic evaluation of the current empirical literature. The purpose of the current study was to synthesize the research that has been conducted on outcomes throughout the legal process (e.g., plea bargaining, conviction, sentencing) for individuals with public defenders, assigned counsel, or retained counsel to better understand whether discrepancies exist in criminal legal system outcomes as a function of indigent defense status and/or attorney type.

Indigent Defense Systems in the United States and Attorney Types

Defendants’ right to an attorney in criminal cases is covered under the Sixth Amendment of the US Constitution. In 1963, the Supreme Court Case Gideon v. Wainwright (1963) clarified defendants’ constitutional right to an attorney by stating that the right to counsel is fundamental to a fair trial, and thus is also protected by the Fourteenth Amendment (i.e., Due Process Clause) of the US Constitution. Prior to this landmark Supreme Court case, the history of indigent defense in the US was marked by significant shortcomings. Though the right to counsel had been recognized in principle, spurred by the 1932 case of Powell v. Alabama which provided counsel to indigent defendants in capital cases, representation was often inconsistent and inadequate. Because many states relied on volunteer lawyers or underfunded public defense systems, defendants often faced criminal charges without proper legal representation. As such, Gideon v Wainwright (1963) ruled that if a defendant in federal or state court is unable to afford their own attorney, they must have one given to them. This ruling revolutionized the American criminal legal system by affirming the constitutional right to counsel for all defendants, regardless of financial means. For most jurisdictions in the United States, this marked the establishment of indigent defense systems.

The two main types of publicly funded defenders within indigent defense systems include public defenders and assigned counsel. Public defenders are defense attorneys who are employed by the government full-time, and thus only take on indigent defense cases. Assigned counsel (also commonly referred to as “court-appointed attorneys”) are defense attorneys who work privately but also take on indigent defense cases when requested by the court. Usually, assigned counsel is granted if there are not enough public defenders in the jurisdiction to handle the current caseload or if the jurisdiction does not have a public defenders office. For assigned counsel, most jurisdictions will use either a contract system or a list system for assigned counsel. That is, courts will either contract with specific attorneys after completing a bidding process in which compensation is negotiated, or courts will have a list of defense attorneys who work privately that they can request indigent defense services from periodically. Public defenders, conversely, do not engage in a bidding or compensation negotiation process. Jurisdictions vary in their exact indigent defense practices. For instance, some jurisdictions rely solely on either public defenders or assigned counsel, while other jurisdictions include both attorney types in their indigent defense systems. Compensation for public defenders and assigned counsel also vary across jurisdictions (OJP, 2011; Primus, 2017). Further, some jurisdictions provide holistic indigent defense services, which combine legal representation, legal advocacy, and connection to social services as a way to address circumstances that lead to criminal legal system involvement (Anderson et al., 2019).

Pragmatic and Theoretical Mechanisms for Outcomes by Attorney Type

In the absence of synthesized data on outcomes for indigent defendants, legal scholars commonly cite the pragmatic and theoretical mechanisms for publicly funded attorneys’ limitations or strengths to support their arguments about the effectiveness of indigent defense counsel. Much of the legal literature asserts that indigent defendants are disadvantaged and experience negative outcomes in the criminal legal system compared to those with retained attorneys, including reports from the American Bar Association (e.g., ABA, 2004; Eldred, 2013; Posner & Yoon, 2011). These assertions largely stem from the wide breadth of research examining the funding, compensation, and caseload of public defenders. National studies on funding for indigent defense systems have found that indigent defense systems, mainly public defender systems, are consistently under-funded (Jaffe, 2018; Lefstein & Spangenberg, 2009; Stevens et al., 2010), especially in contrast to funding for prosecutor offices (Lefstein & Spangenberg, 2009) and compensation for retained attorneys (NALP, 2010). Additionally, over a decade of literature indicates that public defenders tend to have high caseloads (Benner, 2011; Brink, 2018, 2019; Farole & Langton, 2010; Lefstein & Spangenberg, 2009). With this evidence in mind, scholars and legal professionals argue that, pragmatically, public defenders are overworked and underpaid, and thus unable to give their clients the time and attention necessary to ensure the fairest case outcomes (ABA, 2006; Benner, 2011; Brink, 2019; Jaffe, 2018; Joy, 2010).

Legal scholars also cite the theoretical implications of attorneys’ payment on their motivations to obtain the most favorable outcomes possible for their clients. Specifically, type and amount of payment act as positive reinforcement for work-related behavior, with ratio schedules (payment that occurs after a certain number of actions or responses) providing stronger reinforcement for behaviors than interval schedules (payment that occurs after a certain amount of time has passed; Ferster & Skinner, 1957; Morgan, 2010; Skinner, 1948). Public defenders are consistently paid on an interval schedule, regardless of if they have high caseloads or not. Meanwhile, privately retained attorneys (and sometimes assigned counsel) are typically paid on a ratio schedule. Further, assigned counsel and privately retained attorneys are often able to negotiate payment amounts or accept additional private cases with higher payments, while public defenders only take on indigent defense cases with defined payment amounts. Consequently, public defenders are provided weaker reinforcements that would motivate them to consistently ensure the best possible outcomes for their clients. Indeed, a small body of research examining the impact of financial incentives on defense attorneys’ behavior supports these theoretically based arguments, showing that the payment amount and type of payment (i.e., fixed rate versus hourly) can influence defense attorneys’ decisions to negotiate, accept, or reject plea deals offered by the prosecution (Roach, 2017; Stephen et al., 2008).

Literature on the effectiveness of indigent defense systems has also cited public defenders’ full-time role in the courts (referred to as the “courtroom workgroup”) as a mechanism for outcomes experienced by indigent defendants. Some legal scholars argue that public defenders, compared to assigned or retained counsel, have a unique working relationship with judges and prosecutors which results in more favorable outcomes for their clients. Specifically, this relationship allows the public defenders to have a more substantial say in the case decisions (Gertz, 1980; Walker et al., 1996) and results in judges making more favorable decisions for them compared to retained or assigned defense attorneys who they are less familiar with (Tversky & Kahneman, 1974). Conversely, other legal scholars argue that the courtroom workgroup instead results in less favorable outcomes for their clients, asserting that, because of public defenders’ relationship with other members of the court, their allegiance is actually with the court and not with their client (Blumberg, 1967; Eisenstein et al., 1987; Uphoff, 1992). Thus, the current state of literature identifies multiple different mechanisms that may either strengthen or limit the effectiveness of publicly funded defenders within indigent defense systems.

Gaps in Current Knowledge

Despite the wide range of research and attention that has been given to indigent defense systems, gaps in knowledge still remain. First, the research on outcomes for indigent defendants is primarily limited to studies conducted in specific jurisdictions. While this is important for informing local- and state-level decisions, it does not provide a comprehensive understanding of outcomes as a function of indigent defense status or attorney type. Second, studies that include information about indigent defendants focus on outcomes for specific stages in the court process. Thus, our understanding of indigent defendants’ experiences throughout the entire court process is disconnected and isolated to certain stages. Third, given the disjointed nature of the research to-date on indigent defendants, there is disagreement among legal scholars and professionals as to whether indigent defendants are disadvantaged or privileged within criminal courts. For these reasons, a systematic review and meta-analysis on the empirical evidence regarding criminal legal outcomes for indigent defendants is needed; elucidating if and when individuals experience unequal criminal legal outcomes as a function of attorney type or indigent defense status is necessary to understand potential mechanisms for unfair treatment in criminal legal processes.

The Current Study

The purpose of the current study was to conduct a systematic literature review and meta-analysis of outcomes for defendants with public defenders, defendants with assigned counsel, and defendants with retained attorneys to better understand what (if any) discrepancies exist in criminal legal outcomes as a function of indigent defense status and type of representation. Specifically, to understand the role of indigent defense status in criminal legal outcomes, we examined the current empirical literature on pretrial outcomes, case outcomes, sentencing outcomes, and post-case outcomes for indigentFootnote 2 defendants compared to defendants with retained attorneys. To further understand the role of attorney type in criminal legal outcomes for people with indigent defense status, we also examined outcomes for people with public defenders compared to those with assigned counsel.

Methods

To ensure that this study was conducted and reported in a thorough and rigorous manner, we followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Liberati et al., 2009; Moher et al., 2009, 2015). The PRISMA statement is a guideline created to improve the reporting, reproducibility, and quality of systematic reviews and meta-analyses (Liberati et al., 2009; Moher et al., 2009).Footnote 3

Systematic Literature Review

To identify relevant studies for this meta-analysis, we searched six databases (PsychInfo, ProQuest, Web of Science, Google Scholar, Google, and NCJRS Abstracts) using the following search terms: “indigent defendant”, “indigent defense”, “public defender”, “attorney type”, “public defense”, “court appointed counsel”, “court appointed attorney”, “legal defense”, “unable to afford counsel”, “private vs court appointed counsel”, and “offenders type of attorney”. We conducted all searches through July of 2023. We identified additional sources by reviewing the reference lists of included studies.

We included studies if they met the following criteria: (1) included quantitative empirical data on outcome(s) for indigent defendants in criminal courts; (2) were conducted in the United States; and (3) compared outcomes across attorney types. “Indigent defendants” are defined as any defendant who was unable to retain their own attorney and had one given to them by the court; specific definitions of different attorney types are described in the section below. We excluded studies if they did not include quantitative empirical data on outcome(s) for indigent defendants, were conducted in non-criminal courts, were conducted outside of the United States, only included outcomes for one attorney type, or were duplicates. Studies that only included qualitative data were excluded because they could not be used in meta-analyses and were beyond the scope of the current study (please see the Limitations section for additional discussion on the exclusion of qualitative studies). We first screened studies by title, then by abstract, then by full text. We conducted all study screenings in Covidence, an online systematic review software. At each screening stage, a minimum of two members of the research team screened each study; disagreements in screening decisions were resolved by the primary investigator.

Database searches and screening by title yielded 661 records, 124 of which were excluded as duplicates. We then screened 538 records by their abstracts and excluded 282 records that did not meet inclusion criteria. We attempted to retrieve 256 records for full text screening and were unable to obtain three records, resulting in 253 records for full text screening. During the full text screening, 47 records were excluded because they did not include any quantitative empirical data, 114 records were excluded because their data did not include outcomes for indigent defendants, 17 records were excluded because they did not compare outcomes across attorney type (i.e., only reported outcomes for defendants with one type of attorney), one record was excluded because it was conducted outside the United States, and one record was excluded because it used a duplicate sample to another included source. As a result, 73 records met inclusion criteria and were included in the review for full data extraction coding. (See Fig. 1 for the PRISMA flowchart depicting these numbers).

Fig. 1
figure 1

PRSMA flow diagram of studies identified in systematic review

Full size image

Data Extraction

Three coders extracted the following data from the included studies: study information; study jurisdiction(s); data collection years and source of information; sample and demographic information; comparison groups; outcomes; and reported statistical results. We completed coding together for the first three studies. Then, we independently coded the same 10 studies to determine inter-rater reliability (Yeaton & Wortman, 1993). Overall, we reached an intercoder agreement score of 91%. We then divided the rest of the studies to code independently, with 20 randomly selected studies reviewed by a second coder to ensure accuracy.

Comparison Groups

For each study included in the analysis, we extracted data for people with public defenders (i.e., attorneys fully employed by the state in a public defender office), assigned counsel (attorneys employed in private practices who accept assignments from the court in addition to private clients), and retained counsel (attorneys privately secured by the defendant without being appointed by the court). If studies reported outcomes for multiple groups with the same attorney type, we combined the separate groups into their overarching attorney type (e.g., holistic public defenders and traditional public defenders were combined into public defenders).

Meta-analyses included two types of comparisons: indigent defense counsel (i.e., public defenders and assigned counsel) compared to retained counsel, and public defenders compared to assigned counsel. If studies reported outcomes for all three attorney groups (i.e., public defenders, assigned counsel, and retained counsel), we combined public defenders and assigned counsel into the indigent defense counsel group when conducting analyses compared to retained counsel. If studies aggregated all indigent defense counsel types into one group or did not specify the type of indigent defense counsel, we coded the attorney type as indigent defense counsel and excluded them from analyses comparing public defenders and assigned counsel. (See Table 1 for the attorney types coded in each study included in meta-analyses).

Table 1 Studies included in meta-analyses
Full size table

Outcomes

We extracted data on four main types of outcomes: pretrial outcomes, case outcomes, sentencing outcomes, and post-case outcomes. We extracted the relevant statistics and analytic information for each defendant outcome included in each study. This included sample size (n); which comparison groups (i.e., attorney types) were included; type of statistical test (e.g., mean difference, correlation, odds ratio); statistical test results, including measures of variance (e.g., standard deviation, confidence intervals) and significance (e.g., p-value); and any other relevant information, such as the inclusion of covariates or transformation of raw data.

If a study’s outcome(s) did not contain enough statistical information to be included in the meta-analyses (e.g., odds ratios provided without confidence intervals, or means provided without standard deviations), we reached out (if possible) to the author(s) to request their data or the information needed to convert their findings; if no data were provided, we excluded the study’s incomplete outcome(s) from the meta-analysis. Ultimately, over 100 outcome entries and 18 complete studies were excluded because we were unable to obtain complete statistical information for their reported outcomes. Further, 14 studies were excluded because they reported on outcomes that were not reported across enough studies to conduct a meta-analysis.Footnote 4 Thus, 41 studies were included in the final analytic sample (see Fig. 1 for PRISMA flowchart). The studies included in meta-analyses, their outcomes, and study information are depicted in Table 1.

Moderators

We included two moderator variables: data collection start year and geographic region (national, Northeast, Southeast, Midwest, Southwest, or West). The reference category for the categorical geographic region moderator was national. These two moderators were included as indigent defense counsel practices were expected to vary across time, place, and counsel type. Further, the consistency in the reporting of these variables across coded studies allowed for their inclusion in analyses; other potential moderators (e.g., criminal offense type or severity) could not be included as they were not consistently reported across coded studies.

Data Analyses

Calculation of Effect Sizes

The indicator(s) of effect size used for this study were chosen based on the statistical results reported for each outcome in the included studies, with the goal of transforming the smallest number of effect sizes possible (Polanin & Snilstveit, 2016). Outcome measurements were also assessed for similarity across studies; if some studies contained full statistical information but did not measure the outcome in a way that was comparable to the other included studies, it was ultimately excluded from analyses. All outcomes were primarily reported dichotomously, and thus odds ratios were used as the indicator of effect size for each analysis. We calculated all effect sizes using Comprehensive Meta-Analysis, version 3 (Borenstein et al., 2009).

Statistical Analyses

For each outcome we compared attorney types in two ways: (1) any indigent defense counsel compared to retained counsel and (2) public defenders compared to assigned counsel. Effect size values for each outcome were weighted using the inverse-variance method and then combined. We conducted analyses using a random-effects model because of the known variability in indigent defense system practices across jurisdictions and across time. The random-effects model accounts for variability in indigent defense system characteristics, study-level characteristics, and sampling methods (Lipsey & Wilson, 2001). For each outcome, we assessed statistical heterogeneity using Cochran’s Q statistic, the distribution of true effects, and I2. When high heterogeneity was present, we conducted “one study removed” sensitivity analyses to determine each study’s impact on the overall effect size (Borensteinet al., 2009). We assessed sources of variance in effect sizes through meta-regressionFootnote 5 (Borenstein et al., 2009; Richardson et al., 2019). To assess publication bias, we examined funnel plots of standard errors from random effects; asymmetrical funnel plots were assessed for potential non-reporting bias, though it should be noted that funnel plots with less than ten studies are difficult to assess (Page et al., 2021; Polanin & Tanner-Smith, 2014). We conducted all analyses using Comprehensive Meta-Analysis software, Version 3 (Borenstein et al., 2009).

Results

Overall, we included 41 studies with 212 unique independent effect sizes in these meta-analyses. The included studies covered data collection years 1962 to 2016: 9.8% of studies began data collection in the 1960’s, 26.8% began data collection in the 1970’s, 7.3% began data collection in the 1980’s, 34.1% began data collection in the 1990’s, 17.1% began data collection in the early 2000’s, and 4.9% began data collection in the 2010’s. Almost half (43.9%) of studies were from peer-review journal publications, 22% were from government reports, 19.5% were from law review publications, 12.2% were from dissertations, and 2.4% were from book chapters. All 50 states were represented in the included studies; in terms of geographic region, included studies were most commonly conducted in the Southeast (24.4%) and least commonly in the Southwest (7.3%). Only one study indicated that its sample included defendants in federal courts. (See Table 1 for included studies’ information). These studies included more than 2,400,000 cases represented by public defenders, assigned counsel, or retained counsel.

The pooled odds ratios, “one study removed” pooled odds ratios, Cochran’s Q, and I2 for each meta-analysis can be seen in Table 2. All meta-analyses indicated high levels of heterogeneity in effect sizes. Results of meta-analyses and meta-regressions are presented in Table 2 and the sections that follow.

Table 2 Meta-analyses pooled odds ratios
Full size table

Pretrial Outcomes

One pretrial outcome had sufficient statistical information to be included in meta-analysis: pretrial release. Other pretrial outcomes reported in coded studies were bail amount, transfer to adult court in juvenile cases, decision to indict, and case outcomes; however, each of these outcomes contained less than five unique effect sizes, i.e., not enough studies contained complete statistical information on these outcomes to include them in meta-analyses.

Pretrial release

Indigent defendants were about 75% less likely to be released pretrial than defendants with retained attorneys (see Table 2). Meta-regression results indicated that pretrial release effect sizes did not significantly vary by year, but effect sizes did vary across geographical regions; studies conducted in the Southwest (regression coefficient = -8.1217, 95% CI = -11.9702–-4.2732, p < 0.0001) and Midwest (regression coefficient = -2.0801, 95% CI = -3.9107 – -0.2494, p = 0.0259) had significantly less differences in pretrial release rates between counsel type than studies conducted nationally. Data collection year and geographic region accounted for 46% of the variance in effect sizes. Examination of the funnel plot of standard errors showed potential reporting bias and further indicated high heterogeneity, in that almost all studies were gathered around the top of the funnel plot.

We found no significant differences in pretrial release status between defendants with public defenders and those with assigned counsel (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. We did not find strong evidence of reporting bias in our examination of the funnel plot of standard errors, though this may also be due to the low number of studies included in the analysis.

Case Resolution Methods and Disposition

Six court case outcomes had sufficient statistical information to be included in meta-analysis: case dismissal, case resolved by plea, case resolved by trial, acquittal, conviction, and conviction severity. Case length was also reported by coded studies, but we were unable to extract enough statistical information to include this outcome in meta-analyses.

Case Dismissal

Indigent defendants were about 25% less likely to have their cases dismissed than defendants with retained attorneys (see Table 2). Meta-regression results indicated that effect sizes for case dismissal did not differ as a function of data collection years. Effect sizes did vary by geographic region. Specifically, studies conducted in the Southeast (regression coefficient = -0.6571, 95% CI = -1.0554–-0.2588, p = 0.0012) and the Southwest (regression coefficient = -0.8329, 95% CI = -1.3056–-0.3601, p = 0.0006) had significantly lower differences in case dismissal rates. Year and geographic region accounted for 49% of the variance in effect sizes. Examination of the funnel plot indicated potential reporting bias as a result of insignificant findings being unreported, even though the studies in this analysis came from a variety of sources.

We found no significant differences in case dismissals between defendants with a public defender and those with assigned counsel (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. We did not find strong evidence of reporting bias in our examination of the funnel plot of standard errors, though this may also be due to the low number of studies included in the analysis.

Case resolved by Guilty Plea

Indigent defendants were 50% more likely to have their case resolved by entering a guilty plea than defendants with retained attorneys (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. Examination of the funnel plot of standard errors did not strongly indicate presence of publication bias, though this may be due to the low number of studies included in the analysis.

Defendants with public defenders were 35% more likely to resolve their case by entering a guilty plea than defendants with assigned counsel (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. Examination of the funnel plot of standard errors did not strongly indicate presence of publication bias, though this may be due to the low number of studies included in the analysis.

Case Resolved by Trial

We found no significant differences in rates of cases going to trial between indigent defense counsel and retained counsel (see Table 2). We were unable to examine sources of heterogeneity using meta-regression due to issues of collinearity among moderator variables. Examination of the funnel plot of standard errors only slightly indicated the presence of publication bias, in that there were slightly more studies with significant results reported than studies with nonsignificant findings.

We found no significant differences in rates of cases going to trial between public defenders and assigned counsel (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. Examination of the funnel plot of standard errors did not strongly indicate presence of publication bias, though this may be due to the low number of studies included in the analysis.

Acquittal

Indigent defendants were over 50% less likely to be acquitted than defendants with retained counsel (see Table 2). Meta-regression results indicated that effect sizes did not significantly vary by data collection start year or geographic region. Examination of the funnel plot of standard errors did not strongly indicate presence of publication bias.

There was not a sufficient number of studies examining case acquittals between public defenders and assigned counsel to conduct a meta-analysis.

Conviction

Indigent defendants were almost 40% more likely to be convicted than defendants with retained counsel (see Table 2). Meta-regression analyses indicated that the effect sizes did not significantly vary by data collection year. Effect sizes did vary by geographic region; specifically, studies in the Southeast (regression coefficient = 0.523, 95% CI = 0.1923–0.8528, p = 0.0019), Southwest (regression coefficient = 1.609, 95% CI = 0.9603–2.2577, p < 0.0001), and West (regression coefficient = 0.3449, 95% CI = 0.0482–0.6415, p = 0.0227) had significantly higher differences in conviction by attorney type. Year and geographic region accounted for 86% of the variance in effect sizes. The funnel plot of standard errors was slightly skewed to the left, indicating presence of potential publication bias.

Defendants with public defenders were less likely to be convicted than defendants with assigned counsel (see Table 2). Meta-regression results indicated that data collection years and geographic region did not significantly account for variance in effect sizes. Examination of the funnel plot of standard errors showed strong evidence of publication bias, in that the majority of studies clustered around the top of the funnel plot. This indicates that most of the studies included in this analysis were larger, which may mean that findings of smaller studies were not published or did not contain enough statistical information to be included in this analysis.

Conviction Severity (convicted of lesser charges)

We found no significant differences in charge reductions between defendants with indigent defense counsel and those with retained counsel (see Table 2). Meta-regression analyses indicated that effect sizes did not significantly vary by data collection year or geographical region. Examination of the funnel plot of standard errors showed slight potential publication bias; specifically, more studies gathered at the top of the funnel plot, indicating that most studies included were larger.

We found no significant differences in charge reductions between defendants with public defenders and those with assigned counsel (see Table 2). Further assessments of variance in effect sizes were not possible due to an insufficient number of studies. Examination of the funnel plot of standard errors indicated presence of publication bias, though this may be due to the low number of studies included in the analysis.

Sentencing Outcomes

One sentencing outcome had sufficient statistical information to be included in meta-analysis: sentence type. Sentence length and harshness of sentence (relative to the maximum punishment possible for charges received) were also reported by coded studies, but we were unable to extract sufficient statistical information to include this outcome in meta-analyses. Specifically, sentence length was inconsistently measured and reported across studies and, thus, could not be aggregated into a meta-analysis. Harshness of sentence was only fully reported in one study and as such, we did not have enough independent effect sizes for meta-analysis.

Sentence Type (incarceration vs community)

Defendants with indigent defense counsel were significantly more likely to receive carceral sentences than defendants with retained counsel (see Table 2). Meta-regression analyses indicated that the effect sizes did not significantly vary by data collection year. Effect sizes did vary by geographic region; specifically, studies conducted in the Southwest had higher differences in sentences of incarceration between counsel types (regression coefficient = 1.2281, 95% CI = 0.5169–1.9393, p = 0.0187). Data collection start year and geographic region accounted for 23% of the variance in effect sizes. Examination of the funnel plot of standard errors showed some potential publication bias, primarily due to the outlier studies listed above falling farther outside the 95% confidence interval.

We found no significant differences in carceral sentences between defendants with public defenders and those with assigned counsel (see Table 2). Meta-regression analyses indicated that effect sizes varied across data collection years and geographic regions. More recent years had lower differences in incarceration by counsel type (regression coefficient = -0.0249, 95% CI = -0.0397–-0.0100, p = 0.0011), and studies conducted in the Southwest had lower differences between counsel type (regression coefficient = -2.7935, 95% CI = -3.6965–-1.8905, p < 0.0001). Year and geographic region accounted for 61% of the variance in effect sizes. Examination of the funnel plot of standard errors showed strong evidence of publication bias, in that the majority of studies clustered around the top of the funnel plot. This indicates that most of the studies included in this analysis were larger, which may mean that findings of smaller studies were not published or did not contain enough statistical information to be included in this analysis.

Post-Case Outcomes

For post-case outcomes, appeals decision and recidivism were reported by some coded studies; however, we were unable to extract enough statistical information to include these outcomes in meta-analyses.

Discussion

This meta-analysis covered over five decades of research across all 50 states on criminal legal outcomes for people with indigent defense counsel. Overall, findings showed that indigent defendants face worse outcomes than defendants with retained counsel in six court processing stages. Specifically, people with indigent defense counsel were more likely to resolve their case with a guilty plea, be convicted, and receive a sentence of incarceration, and were also less likely to be released pretrial, have their case dismissed, and have their case acquitted than compared to people with retained counsel. Results suggest that there were fewer discrepancies between types of indigent defense counsel – individuals with public defenders were more likely to resolve their case with a guilty plea than defendants with assigned counsel, though defendants with assigned counsel were more likely to be convicted than those with public defenders. Findings of this meta-analysis also suggest that effect sizes frequently varied by geographic region and sometimes varied by data collection years, demonstrating a potential lack of consistency in indigent defense system practices and judicial decision-making across jurisdiction and time. These results and their implications are discussed in further detail in the paragraphs that follow.

Findings indicate more discrepancies in criminal legal outcomes in the comparison between indigent defendants and those with retained counsel than in the comparison between indigent defense counsel types (public defenders and assigned counsel). This difference suggests that indigent defendants are likely disadvantaged as a result of biases in court processes and judicial decision-making rather than by attorney performance. As stated in the introduction of this manuscript, many scholars use attorney caseload, compensation, and the courtroom workgroup as explanations for discrepancies in outcomes for indigent defendants (e.g., ABA, 2006; Benner, 2011; Uphoff, 1992). While it is true that public defenders are often overworked and underpaid (Benner, 2011; Brink, 2020, 2019), assigned counsel attorneys do not necessarily face the same issues of heavy caseloads and low attorney fees, as they are able to negotiate these aspects for their indigent defense work and may choose to supplement their earnings by taking on private cases. If caseload and fee issues were truly the primary source of differential outcomes between defendants, then we would expect to see more discrepancies between public defenders and assigned counsel in meta-analysis results. Additionally, many legal scholars have cited the courtroom workgroup as a potential source of defendant outcome discrepancies (e.g., Uphoff, 1992; Walker et al., 1996). Some scholars claim that public defenders are able to secure more positive outcomes as they are more familiar to judicial decision-makers (Gertz, 1980; Walker et al., 1996), while others assert that their relationship with other members of the court causes them to feel more allegiance to the court than to their clients (Blumberg, 1967; Eisenstein et al., 1987; Uphoff, 1992). However, assigned counsel attorneys typically do not have the same consistent role in the courtroom workgroup. Thus, yet again, if the courtroom workgroup was indeed a primary source of differential outcomes between defendants, then we would expect to see more discrepancies between public defenders and assigned counsel in this meta-analysis.

Indigent defendants experiencing worse outcomes than defendants with retained counsel is consistent with the larger goals of the criminal legal system. Over the course of the twentieth century, the criminal legal system has increasingly acted as a method of poverty management (Soss et al., 2011; Waquant, 2009; Beckett & Western, 2001; Shelden & Vasiliev, 2017). To manage the population experiencing poverty, the criminal legal system relies on both criminalization of poverty and over-surveillance of poor communities (Brayne, 2020; Jones, 2016; Smith, 1986). Criminal laws that have been enacted explicitly punish the conditions and harms associated with poverty, solidifying the connection between guilt and economic status (Robinson, 2019; Reiman & Leighton, 2023; Beckett & Herbert, 2011; Herring, 2019). This increased surveillance contributes to higher arrest rates among people with lower socioeconomic status compared to people with higher socioeconomic status (Smith, 1986). In misdemeanor courts, adjudication of guilt is deprioritized and replaced with “managerial justice”, where individuals are kept in perpetual supervision through a growing body of legal requirements (Kohler-Hausmann, 2018). It is unsurprising then that individuals receiving indigent defense services receive worse criminal legal outcomes, as the criminal legal system is purposefully operating to capture and control those most likely to receive indigent defense services. Put another way, the promise of Gideon is concerned with what happens in a courtroom; the vast majority of harm that creates unequal criminal legal outcomes occurs before getting to the courtroom (Butler, 2013).

The creation of financial barriers for people in the criminal legal system enables differences in outcomes between indigent defendants and those with retained counsel. Indigent defendants in the current study were significantly less likely to obtain pretrial release than defendants with retained counsel. In most jurisdictions, bail practices are largely dependent on a person’s financial resources. The median assigned bail amount in the United States for felonies is $10,000 (Rabuy & Kopf, 2016) and for misdemeanors is $2,000 (Wykstra, 2018); however, almost half of people in the United States struggle to cover unexpected payments of $400 or more (Federal Reserve Board of Governors, 2019) making them unable to afford even a fraction of the median assigned bail amounts. Similarly, statutorily mandated legal financial obligations (LFOs) typically exceed the capacity of low-income defendants to pay (Harris, 2016; Fernandes et al., 2019). Unable to pay, people are forced into “court ordered payment schedule[s]” extending both the length and type of control that the criminal-legal system holds (Pattillo & Kirk, 2021, p. 911). Unpaid LFOs can also deepen criminal-legal involvement through additional charges or revocation of community release (Ruhland et al., 2020; Friedman & Pattillo, 2019). While bail and LFOs are explicit criminal legal costs, there are other costs associated with full participation in the criminal-legal process. Defendants may have to obtain reliable transportation to and from the courthouse (which often requires paying for parking or public transportation) and be able to take time off work or pay for childcare (Bornstein et al., 2012). The ability to weather all of these costs shapes the favorability of criminal legal outcomes, leading to indigent clients receiving worse outcomes.

There are structural reasons that indigent defense services cannot reach parity with retained defense attorneys. Indigent defense services are often financed in a way to serve as a managed opposition to the carceral state. As discussed above, indigent defense services are largely underfunded. This has been consistently true since Gideon, with the state of indigent defense funding being described as in a “permanent crisis” (Mayeux 2020, p. 117). This lack of funding limits the amount of time that defense lawyers can work on specific cases, leading to prioritization of more serious accusations at the cost of time spent working with people facing less severe charges (Clair, 2020; Natapoff, 2014; Wright & Roberts, 2023). Even in more serious cases, state-funded lawyers face funding barriers that limit the defenses available. While state legislatures refuse to adequately fund indigent defense services, they also point to indigent defense services as proof that defendants were able to have their day in court (Mayeux, 2020). In this way, the existence of any indigent defense service, regardless of quality, contributes to an illusion of fairness in the criminal-legal system (Butler, 2013; Reiman & Leighton, 2020). This is particularly harmful when the assistance of indigent defense is used to hide situations where an innocent person was found guilty because they went into trial with poor defense, or because of race or class bias. The findings of this meta-analysis, that people with indigent defense counsel experience worse outcomes, cannot weigh on whether the assumption of guilt is true, but they do raise concern about the quality of the indigent defense system. This is not to suggest that indigent defense attorneys are complicit partners; lawyers working indigent defense actively advocate against policies that harm poor communities, including the severe underfunding of defense services (Wright & Roberts, 2023; Waquant, 2009). Rather, it suggests that current indigent defense services are structured in a way that limits their ability to challenge larger goals of the criminal legal system.

Findings observed in this meta-analysis suggest that discrepancies in criminal legal outcomes for indigent defendants may also be the result of biases in judicial decision-making. Judicial decisions are often influenced by extralegal and even subjective factors (Bobo & Thompson, 2006; Eckhouse et al., 2018; Forlini, 2018; Skeem & Lowenkamp, 2016), and research has shown that defendants of lower socioeconomic status experience harsher criminal legal sanctioning (Lofstrom & Raphael, 2016; Neitz, 2013; van Eijk, 2017). Previous literature concerning role adaptation of decision-makers along the criminal-welfare continuum is suggestive here. Soss and colleagues (2011, p. 25) describe the concept of “neoliberal paternalism” as a governing logic that rejects poor people’s capacity to manage their own lives and insist on a disciplinary role “for the poor’s own good.” This logic spreads across government decision-makers, including welfare case managers (Soss et al., 2011), prosecutors (Packard, 2023), and police (Stuart, 2013; Stuart, 2016). If the criminal legal system is defined by its goal of poverty management, then class-based bias in judicial decision-making would be an expected outcome. To combat these discrepancies in legal decisions, the American Bar Association established a Model Code of Judicial Conduct that includes an assessment of risk of socioeconomic bias (ABA, 2020). As such, evidence of discrepancies in outcomes between indigent defendants and defendants with retained attorneys, but not between indigent defense counsel types, suggest that the defendant’s indigent status may unfairly prejudice defendants’ outcomes.

The overall finding that indigent defendants are more likely to experience negative criminal legal outcomes is concerning for a number of reasons. According to this meta-analysis, indigent defendants were more likely to face periods of incarceration both during the pretrial stage and in their sentencing. Incarceration serves to remove the right to freedom–a right so important and central to a person’s being that it is labeled “unalienable” in the second sentence of the United States Declaration of Independence. Moreover, decades of research demonstrate that people who experience incarceration are more likely to also experience long-term consequences of poor physical and mental health, economic strain, damaged personal relationships, and continued re-involvement in the criminal legal system (Mille et al., 2001; Mueller-Smith, 2015; Willmott & van Olphen, 2005). Many of these long-term consequences of incarceration are especially concerning during times of crisis, such as the COVID-19 pandemic. At the time of writing this manuscript, almost 400,000 people housed in prisons had reported contracting COVID-19 and less than two-thirds of them had recovered (The Marshall Project, 2021). Thousands of people housed in jails and prisons died from COVID-19 during the pandemic as a result of their incarceration (Equal Justice Initiative, 2021), many of whom were detained pretrial because they were unable to afford bail (Deitch et al., 2020). This meta-analysis indicates that indigent defendants are at increased risk for these negative consequences resulting from incarceration.

The findings that indigent defendants experience worse outcomes in their case dispositions or verdicts are also cause for concern. This meta-analysis shows that indigent defendants are less likely to have their case dismissed or be acquitted and are more likely to plead guilty and be convicted than defendants who retain their defense attorney. This means that indigent defendants are more likely to receive punitive actions and a criminal record by the end of their case. The stigmatization of criminality, especially for individuals belonging to intersecting marginalized groups, can lead to strained social relationships, economic adversity, and poor health outcomes for individuals convicted in criminal courts (Laredo, 2012; West, 2015). For instance, individuals with criminal convictions often face barriers to stable employment and housing (Laredo, 2012). According to strain theory of criminal behavior, these economic strains directly contribute to a person’s engagement in criminal activity (Agnew, 1985, 1992; Merton, 1938) which, in turn, contributes to high rates of recidivism and subsequent re-involvement in the criminal legal system (Alper et al., 2018). Stigmatization of criminality can also directly impact individuals’ self-identity, well-being, and future behavior (Rasmusen, 1996; West, 2015). By being more likely to be convicted of the crimes they are charged with, indigent defendants are more likely to experience damages to their economic, social, and mental well-being, even after their involvement in the criminal legal system is over.

Findings from this meta-analysis also highlight potential ways to improve outcomes for indigent defendants. Results show that indigent defendants begin experiencing disadvantages in their court processing as early as the pretrial stage, in which indigent defendants are less likely to be released pretrial than individuals with retained attorneys. Legal scholars and researchers widely recognize the concept of cumulative disadvantage in criminal legal system processing; that is, the disadvantages people experience in the criminal legal system are not isolated to each stage, but rather accumulate into more disadvantages as they move through the system (Kurlycheck & Johnson, 2019). In this way, the negative outcomes indigent defendants experience in later stages of court processing may be improved by addressing issues in pretrial release status. Indeed, research on the impacts of pretrial detention has consistently demonstrated that individuals detained pretrial are more likely to enter a guilty plea, be convicted of the charges against them, and receive sentences of incarceration (Austin, 2017; Cohen & Reaves, 2007; Dobbie et al., 2018; Heaton, et al., 2017; Kellough & Wortley, 2002; Ottone & Scott-Hayward, 2018; Philips, 2012; Sacks, et al., 2015; Stevenson, 2018), all of which are outcomes more likely to be experienced by indigent defendants in this meta-analysis. One potential way to alleviate indigent defendants’ likelihood of pretrial detention is to reform traditional bail practices. As stated above, indigent defendants are, by definition, more financially disadvantaged than their non-indigent counterparts, and thus more likely to be detained with even low bail amounts. Beyond bail practices, jurisdictions should also consider revising their pretrial release guidelines to allow for more defendants to be released pretrial, especially defendants who are charged with lower-level crimes that pose no threat to public safety.

Another potential way to improve outcomes for indigent defendants highlighted in this meta-analysis is by establishing evidence-based standards for indigent defense systems, financial-based court practices, and judicial decision-making across jurisdictions. Meta-regression results in this study consistently showed that effect sizes significantly varied across geographic regions. After Gideon v Wainwright (1963), each jurisdiction across the United States was largely left to decide how they implemented indigent defense systems and the policies and practices that guide them. As highlighted previously, the lack of guidance or standards for indigent defense systems has resulted in a lot of variances in indigent defense system practices. For example, some jurisdictions provide indigent defense for misdemeanor cases, while others only provide indigent defense for felony cases or cases that could result in sentences of incarceration. Additionally, financial-based practices also vary widely across jurisdiction. For example, some jurisdictions rely more heavily on cash bail, while others provide alternatives to cash bail practices. Additionally, court fees vary widely across jurisdictions. The differences between indigent defense policies and practices, as well as differences in financial-based policies and practices, across jurisdictions may explain the high levels of heterogeneity in the analyses conducted in this study. Thus, creating evidence-based standards that are applicable to all jurisdictions may help to alleviate the discrepancies in negative outcomes experienced by indigent defendants. Additionally, the variance by geographic region also suggests discrepancies in judicial decision-making across jurisdictions. This further highlights the impact of biases in judicial decisions. Results of this meta-analysis suggest some jurisdictions may demonstrate more bias as a function of indigent defense status through larger discrepancies in criminal legal outcomes between indigent defendants and defendants with retained counsel. However, these reforms, if siloed, are likely insufficient to truly address, and eliminate, inequities in criminal legal outcomes experienced by indigent defendants. Efforts to reforming court practices, indigent defense systems, or judicial decision-making will not address, for example, the over-policing and criminalization of people experiencing poverty. Whereas these individual reform strategies may reduce inequities and improve criminal legal outcomes for people with indigent defense status compared to those with retained counsel, the broader systemic issues that lead to their criminal legal system involvement will remain. Thus, meaningful system transformation will require multiple, coordinated efforts across criminal legal system entities (Desmarais et al., 2021).

Another key finding of the current study is the demonstrated issues with research done on criminal legal outcomes for indigent defendants thus far. As stated previously, many studies were excluded because they did not include enough statistical information for meta-analyses or because the outcomes examined were not measured consistently across studies. This shows a lack of cohesion across the field of study, making it difficult to identify patterns and mechanisms for outcomes experienced by indigent defendants. Additionally, we were unable to control for many variables that contribute to examined outcomes, such as charge type and severity, due to a lack of inclusion of this information in the original studies. As such, it is difficult to distinguish between the unique impact of indigent defense status and other factors that may be associated, such as severity of charges. Studies also did not consistently report defendants’ race, jurisdictions’ urbanicity, and other socio-demographic characteristics that are known to contribute to criminal legal-related outcomes (e.g., Arnold, et al., 2018; Ball & Bostaph, 2009; Donnelly & Asiedu, 2020; Mitchell, 2005); thus, we were unable to control for these factors as well. Indeed, various studies demonstrate that people of color are more likely to have indigent defense status (Lyon, 2012; Myrick et al., 2012), that racial bias results in worse criminal legal-related outcomes for people of color (Johnson & DiPietro, 2012; Mitchell, 2005; Spohn, 2001; Ulmer et al., 2016), and that attorneys’ decisions are influenced by racial bias, especially when working with high caseloads (Blakemore, 2016; Clair, 2021; Richardson & Goff, 2013). Therefore, race may be acting as a confounding variable, though this is impossible to determine without controlling for race. This issue underscores the need for future studies to include comprehensive information on defendant demographics as well as legal factors.

Finally, the current study demonstrated a lack in more recent evidence for outcomes experienced by indigent defendants, in that very few studies included in meta-analyses used a sample from the 2010’s. This pattern aligns with research conducted on other criminal legal system practices (e.g., bail reform), in which more research was conducted prior to the 1980’s when reform of the criminal legal system was a higher priority for government entities; indeed, Gideon v. Wainwright was decided in 1963, and President Johnson’s taskforce, “The President’s Commission on Law Enforcement and the Administration of Justice,” was formed in 1965. When the “tough on crime” era began in the 1980’s, research on criminal legal system practices subsequently began to dwindle (Reiman & Leighton, 2023). This study underscores the need for more modern and rigorous examinations of outcomes experienced by individuals with indigent defense status compared to those with retained counsel.

Limitations

One of the largest limitations of this study was the high levels of heterogeneity not explained by subgroup analyses and meta-regressions. Because of the largely unexplained variance in effect sizes, results from this meta-analysis should be interpreted with caution. However, high levels of heterogeneity are not uncommon in meta-analyses on observational studies (Mueller et al., 2018; Stroup et al., 2000), especially in meta-analyses of outcomes in the US criminal legal system given the high variability of criminal legal system practices across jurisdictions (e.g., Lowder et al., 2018; Viljoen et al., 2019). In fact, findings of high heterogeneity and the sources of variances in effect sizes can be just as informative as the meta-analysis results (Dwyer et al., 2001). Additionally, given the scarcity of information on outcomes for indigent defendants in the current literature, the decisions we made to address high levels of heterogeneity were considered alongside the importance of reporting findings of this meta-analysis (Imrey, 2020; Lowder et al., 2018; Sabitova et al., 2020; Viljoen et al., 2019).

Another limitation of this meta-analysis relates to the inconsistency in measurement and reporting of outcomes across studies. While we were able to find over 70 empirical studies examining outcomes for indigent defendants, we were ultimately only able to include 41 in the final meta-analysis. We excluded many studies from the meta-analysis due to differences in definition and measurement of outcome variables, as well as incomplete statistical information. For example, although the outcome of sentence length was reported for more than 10 studies, operationalized definitions of sentence length as well as statistics reported differed so widely that there was not enough overlap to comparison group analyses. Inclusion of these studies may have led to different results of meta-analyses and meta-regression, though there is currently no way of knowing this. Additionally, many outcomes that are included in this meta-analysis contain less than ten studies. This low number of studies in certain outcomes made it difficult to interpret sources of variance and publication bias.

Additionally, the exclusion of qualitative studies from this systematic review is another limitation. Qualitative studies are useful for understanding complex or emerging issues, and the issue of individuals with indigent defense status experiencing worse criminal legal system outcomes qualifies as a complex issue that is not yet fully understood. While the examination of qualitative studies was outside the scope of the current study, future studies should include and closely examine findings from qualitative studies on outcomes experienced by indigent defendants. Future qualitative studies should also be sure to include information from multiple sources, especially from the individuals experiencing these criminal legal system outcomes as indigent defendants; it is imperative that people with lived (i.e., prior) and living (i.e., current) experience of criminal legal system involvement are consulted as subject matter experts in understanding outcomes they experience and informing reform efforts to address discrepancies in those outcomes. A recent environmental scan emphasized the need for more studies that include perspectives and insights directly from indigent defendants to better understand the quality of indigent defense representation, what quality representation means to people involved in the criminal legal system, and effective strategies for improving indigent defense (Banks et al., 2023). Previous qualitative studies with individuals impacted by the criminal legal system have highlighted the importance of qualitative research focused on this population. For instance, qualitative studies with people with substance use disorders currently housed in jails or prisons highlight barriers to accessing treatment while incarcerated and offer suggestions for better facilitating access to treatment among individuals experiencing incarceration, such as providing holistic services that include treatment, rather than providing treatment as a standalone service (Barenie et al., 2022). Additionally, qualitative and mixed-methods studies examining the inclusion of individuals with felony convictions in jury systems have effectively demonstrated how “felon-jurors” can improve jury deliberations by teaching other jury members and providing insights that are otherwise missing from jury decision-making (Binnall, 2021). Thus, the current study is limited in its interpretations of findings without the qualitative perspectives of individuals with lived and living experience.

Finally, as previously mentioned, interpretation of findings of this meta-analysis is limited by our inability to control for variables that likely contributed to criminal legal outcomes, such as charge type and severity, defendants’ race, jurisdictions’ urbanicity, and other socio-demographic characteristics, because these variables were not reported on in the original studies. While our inability to control for these factors introduces potential bias in our findings, it also elucidates a major gap that future research can help address by consistently and comprehensively reporting such information. Further, most studies only included complete statistical information for frequencies of events. If more studies had complete statistical information for more robust analyses that adjust and control for other variables (e.g., odds ratios), the results of this meta-analysis would account for more of the variability in outcomes experienced by defendants. However, despite this limitation and the ones listed above, this meta-analysis presents novel findings that contribute to what is known about outcomes experienced by indigent defendants throughout court processing.

Future Directions

This meta-analysis highlights the need for more research on criminal legal system outcomes experienced by indigent defendants. For example, most studies included in this meta-analysis focused on outcomes related to conviction and sentences of incarceration. Yet, to understand high rates of conviction and incarceration among indigent defendants, we also need to better understand outcomes related to pretrial decisions (e.g., pretrial release vs detention, conditions of release, bail assignments) and case resolution methods (e.g., dismissal, plea deal). Additionally, we were unable to examine post-sentencing outcomes related to reentry, recidivism, and long-term personal success. Conviction and incarceration, which indigent defendants are more likely to experience, can have long-term impacts. More research is needed to better understand outcomes associated with negative long-term consequences that may be experienced by indigent defendants after their case’s resolution.

In addition to needing more research on outcomes, more research is also needed to determine the mechanisms for the discrepancies experienced by indigent defendants. Currently, we are only able to speculate on the mechanisms that lead to more negative outcomes for indigent defendants. While we posit that these discrepancies are primarily the result of biases embedded in criminal legal system processes, we cannot definitively explain the differences in outcomes experienced by indigent defendants compared to defendants with retained counsel. Research that examines the causes and solutions for negative outcomes experienced by indigent defendants is needed. Specifically, given the variability in effect sizes across geographic regions in this meta-analysis, more research is needed to examine the impacts of jurisdictional differences in outcomes experienced by indigent defendants. Additionally, future research should examine differences in criminal legal outcomes resulting from variation in indigent defense service practices. For example, it was not within the current study’s scope to examine differences between people represented by holistic public defenders and those by traditional public defenders, but future research should examine the impact of holistic indigent defense services on outcomes for indigent defendants. Finally, future research conducted through an intersectional lens that examines and reports on extralegal factors (i.e., race and other socio-demographic information) is needed to better understand differences in criminal legal outcomes experienced by indigent defendants who belong to multiple intersecting marginalized groups.

Conclusion

This was the first meta-analysis conducted on outcomes across court processing stages for indigent defendants compared to defendants with retained counsel and across indigent defense counsel types. In this meta-analysis, we screened over 600 sources, read over 250 studies, coded 73 studies, and analyzed more than 200 unique effect sizes within 41 studies. Findings of this meta-analysis indicate that indigent defendants experience worse outcomes in pretrial release, case disposition, and sentencing decisions compared to defendants with retained counsel. In comparisons across types of indigent defense counsel, individuals with public defenders were more likely to resolve their case with a guilty plea than those with assigned counsel, while individuals with assigned counsel were more likely to be convicted than those with public defenders; all other analyses comparing type of indigent defense counsel showed no significant differences in outcomes. All analyses showed considerable heterogeneity in effect sizes, some of which was explained by variances in effect sizes across geographic regions or data collection years. These results suggest that the disadvantages indigent defendants experience in criminal legal outcomes are likely not the result of ineffective counsel, but rather the result of systemic and individual biases. Further research is needed to understand additional outcomes experienced by indigent defendants within pretrial and post-case stages, and to understand the mechanisms through which indigent defendants experience more negative criminal legal outcomes than their privately represented counterparts.