Search

We study the results of a massive nationwide correspondence experiment sending more than 83,000 fictitious applications with randomized characteristics to geographically dispersed jobs posted by 108 of the largest U.S. employers. Distinctively Black names reduce the probability of employer contact by 2.1 percentage points relative to distinctively white names. The magnitude of this racial gap in contact rates differs substantially across firms, exhibiting a between-company standard deviation of 1.9 percentage points. Despite an insignificant average gap in contact rates between male and female applicants, we find a between-company standard deviation in gender contact gaps of 2.7 percentage points, revealing that some firms favor male applicants and others favor women. Company-specific racial contact gaps are temporally and spatially persistent, and negatively correlated with firm profitability, federal contractor status, and a measure of recruiting centralization. Discrimination exhibits little geographical dispersion, but two-digit industry explains roughly half of the cross-firm variation in both racial and gender contact gaps. Contact gaps are highly concentrated in particular companies, with firms in the top quintile of racial discrimination responsible for nearly half of lost contacts to Black applicants in the experiment. Controlling false discovery rates to the 5% level, 23 companies are found to discriminate against Black applicants. Our findings establish that discrimination against distinctively Black names is concentrated among a select set of large employers, many of which can be identified with high confidence using large-scale inference methods.

The E. coli aminoacyl transferase (AaT) can be used to transfer a variety of unnatural amino acids, including those with azide or alkyne groups, to the α-amine of a protein with an N-terminal Lys or Arg. Subsequent functionalization through either copper-catalyzed or strain-promoted click reactions can be used to label the protein with fluorophores or biotin. This method can be used to directly detect AaT substrates or in a two-step protocol to detect substrates of the mammalian ATE1 transferase.

Empirical researchers often combine multiple instrumental variables (IVs) for a single treatment using two-stage least squares (2SLS). When treatment effects are heterogeneous, a common justification for including multiple IVs is that the 2SLS estimand can be given a causal interpretation as a positively weighted average of local average treatment effects (LATEs). This justification requires the well-known monotonicity condition. However, we show that with more than one instrument, this condition can only be satisfied if choice behavior is effectively homogeneous. Based on this finding, we consider the use of multiple IVs under a weaker, partial monotonicity condition. We characterize empirically verifiable sufficient and necessary conditions for the 2SLS estimand to be a positively weighted average of LATEs under partial monotonicity. We apply these results to an empirical analysis of the returns to college with multiple instruments. We show that the standard monotonicity condition is at odds with the data. Nevertheless, our empirical checks reveal that the 2SLS estimate retains a causal interpretation as a positively weighted average of the effects of college attendance among complier groups. (JEL C26, I23, I26, J24, J31, R23)

Correspondence experiments probe for discrimination by manipulating employer perceptions of applicant characteristics. We consider the gains from dynamically adapting the number and quality of fictitious applications each employer receives to their prior callback decisions. Calibrating employer behavior to experimental data from Nunley et al. (2015), we find that it is possible to cut the number of applications required to detect a fixed number of discriminators roughly in half relative to a benchmark design with a fixed number of applications per job. These gains are achieved primarily from abandoning jobs with very low callback probabilities and those that call back Black applicants.

Can schools that boost student outcomes reproduce their success at new campuses? We study a policy reform that allowed effective charter schools in Boston, Massachusetts to replicate their school models at new locations. Estimates based on randomized admission lotteries show that replication charter schools generate large achievement gains on par with those produced by their parent campuses. The average effectiveness of Boston’s charter middle school sector increased after the reform despite a doubling of charter market share. An exploration of mechanisms shows that Boston charter schools compress the distribution of teacher effectiveness and may reduce the returns to teacher experience, suggesting the highly standardized practices in place at charter schools may facilitate replicability. (JEL H75, I21, I28)

This paper develops methods for detecting discrimination by individual employers using correspondence experiments that send fictitious resumes to real job openings. We establish identification of higher moments of the distribution of job‐level callback rates as a function of the number of resumes sent to each job and propose shape‐constrained estimators of these moments. Applying our methods to three experimental data sets, we find striking job‐level heterogeneity in the extent to which callback probabilities differ by race or sex. Estimates of higher moments reveal that while most jobs barely discriminate, a few discriminate heavily. These moment estimates are then used to bound the share of jobs that discriminate and the posterior probability that each individual job is engaged in discrimination. In a recent experiment manipulating racially distinctive names, we find that at least 85% of jobs that contact both of two white applications and neither of two black applications are engaged in discrimination. To assess the potential value of our methods for regulators, we consider the accuracy of decision rules for investigating suspicious callback behavior in various experimental designs under a simple two‐type model that rationalizes the experimental data. Though we estimate that only 17% of employers discriminate on the basis of race, we find that an experiment sending 10 applications to each job would enable detection of 7–10% of discriminatory jobs while yielding Type I error rates below 0.2%. A minimax decision rule acknowledging partial identification of the distribution of callback rates yields only slightly fewer investigations than a Bayes decision rule based on the two‐type model. These findings suggest illegal labor market discrimination can be reliably monitored with relatively small modifications to existing correspondence designs.

School choice may lead to improvements in school productivity if parents’ choices reward effective schools and punish ineffective ones. This mechanism requires parents to choose schools based on causal effectiveness rather than peer characteristics. We study relationships among parent preferences, peer quality, and causal effects on outcomes for applicants to New York City’s centralized high school assignment mechanism. We use applicants’ rank-ordered choice lists to measure preferences and to construct selection-corrected estimates of treatment effects on test scores, high school graduation, college attendance, and college quality. Parents prefer schools that enroll high-achieving peers, and these schools generate larger improvements in short- and long-run student outcomes. Preferences are unrelated to school effectiveness and academic match quality after controlling for peer quality. (JEL D12, H75, I21, I26, I28)

Site-selective incorporation of thioamides into peptides and proteins provides a useful tool for a wide range of applications. Current incorporation methods suffer from low yields as well as epimerization. Here, we describe how the use of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) rather than piperidine in fluorenylmethyloxycarbonyl (Fmoc) deprotection reduces epimerization and increases yields of thioamide-containing peptides. Furthermore, we demonstrate that the use of DBU avoids byproduct formation when synthesizing peptides containing side-chain thioamides.

Despite growing interest in improving early-childhood education in developing countries, there is little evidence on cost-effective ways of doing so at scale. We use a large-scale randomized experiment to study the impact of adding an extra worker focused on pre-school education (for children aged 3-5) in the world’s largest public early-childhood program: India’s Integrated Child Development Services (ICDS). Adding a worker doubled net instructional time and led to 0.29 and 0.46 standard deviation increases in math and language test scores after 18 months for children who remained enrolled in the program. Rates of stunting and severe malnutrition were also lower in the treatment group, likely reflecting the effect of freeing up time of the incumbent worker to focus more on nutrition-related tasks. A cost-benefit analysis suggests that the benefits of the program are likely to significantly exceed its costs even under conservative assumptions. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

We study the results of a massive nationwide correspondence experiment sending more than 83,000 fictitious applications with randomized characteristics to geographically dispersed jobs posted by 108 of the largest U.S. employers. Distinctively Black names reduce the probability of employer contact by 2.1 percentage points relative to distinctively white names. The magnitude of this racial gap in contact rates differs substantially across firms, exhibiting a between-company standard deviation of 1.9 percentage points. Despite an insignificant average gap in contact rates between male and female applicants, we find a between-company standard deviation in gender contact gaps of 2.7 percentage points, revealing that some firms favor male applicants while others favor women. Company-specific racial contact gaps are temporally and spatially persistent, and negatively correlated with firm profitability, federal contractor status, and a measure of recruiting centralization. Discrimination exhibits little geographical dispersion, but two digit industry explains roughly half of the cross-firm variation in both racial and gender contact gaps. Contact gaps are highly concentrated in particular companies, with firms in the top quintile of racial discrimination responsible for nearly half of lost contacts to Black applicants in the experiment. Controlling false discovery rates to the 5% level, 23 individual companies are found to discriminate against Black applicants. Our findings establish that systemic illegal discrimination is concentrated among a select set of large employers, many of which can be identified with high confidence using large scale inference methods.

We use admissions lotteries to estimate the effects of large-scale public preschool in Boston on college-going, college preparation, standardized test scores, and behavioral outcomes. Preschool enrollment boosts college attendance, as well as SAT test-taking and high school graduation. Preschool also decreases several disciplinary measures including juvenile incarceration, but has no detectable impact on state achievement test scores. An analysis of subgroups shows that effects on college enrollment, SAT-taking, and disciplinary outcomes are larger for boys than for girls. Our findings illustrate possibilities for large-scale modern, public preschool and highlight the importance of measuring long-term and non-test score outcomes in evaluating the effectiveness of education programs.

The thioamide is a versatile replacement of the peptide backbone with altered hydrogen bonding and conformational preferences, as well the ability participate in energy and electron transfer processes. Semi-synthetic incorporation of a thioamide into a protein can be used to study protein folding or protein/protein interactions using these properties. Semi-synthesis also provides the opportunity to study the role of thioamides in natural proteins. Here we outline the semi-synthesis of a model protein, the B1 domain of protein G (GB1) with a thioamide at the N-terminus or the C-terminus. The thioamide is synthetically incorporated into a fragment by solid-phase peptide synthesis, whereas the remainder of the protein is recombinantly expressed. Then, the two fragments are joined by native chemical ligation. The explicit protocol for GB1 synthesis is accompanied by examples of applications with GB1 and other proteins in structural biology and protein misfolding studies.

Many large urban school districts match students to schools using algorithms that incorporate an element of random assignment. We introduce two simple empirical strategies to harness this randomization for value-added models (VAMs) measuring the causal effects of individual schools. The first estimator controls for the probability of being offered admission to different schools, treating the take-up decision as independent of potential outcomes. Randomness in school assignments is used to test this key conditional independence assumption. The second estimator uses randomness in offers to generate instrumental variables (IVs) for school enrollment. This procedure uses a low-dimensional model of school quality mediators to solve the under-identification challenge arising from the fact that some schools are under-subscribed. Both approaches relax the assumptions of conventional value-added models while obviating the need for elaborate nonlinear estimators. In applications to data from Denver and New York City, we find that models controlling for both assignment risk and lagged achievement yield highly reliable VAM estimates. Estimates from models with fewer controls and older lagged score controls are improved markedly by IV. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Marginal treatment effect methods are widely used for causal inference and policy evaluation with instrumental variables. However, they fundamentally rely on the well-known monotonicity (threshold-crossing) condition on treatment choice behavior. Recent research has shown that this condition cannot hold with multiple instruments unless treatment choice is effectively homogeneous. Based on these findings, we develop a new marginal treatment effect framework under a weaker, partial monotonicity condition. The partial monotonicity condition is implied by standard choice theory and allows for rich heterogeneity even in the presence of multiple instruments. The new framework can be viewed as having multiple different choice models for the same observed treatment variable, all of which must be consistent with the data and with each other. Using this framework, we develop a methodology for partial identification of clearly stated, policy-relevant target parameters while allowing for a wide variety of nonparametric shape restrictions and parametric functional form assumptions. We show how the methodology can be used to combine multiple instruments together to yield more informative empirical conclusions than one would obtain by using each instrument separately. The methodology provides a blueprint for extracting and aggregating information about treatment effects from multiple controlled or natural experiments while still allowing for rich heterogeneity in both treatment effects and choice behavior.

Site-specific functionalization of unprotected native peptides and biomolecules remains a useful transformation in synthetic design and chemical biology, yet until recently, advancements in transition metal-catalyzed methods, which have prevailed in organic synthesis, have been relatively ineffective when applied to large and structurally complex biomolecules. Here, the mechanistically distinct, Ni/photoredox-catalyzed arylation of unprotected, native thiols (e.g., cysteine residues) is reported - a process initiated through a visible light-promoted, hydrogen atom transfer (HAT) event under ambient conditions. Sub-stoichiometric loadings of the dual-catalyst system (≤5 mol%) are employed, granting excellent site-specificity, broad substrate scope, and low chemical waste. Reaction scalability (from μg to grams) has been achieved through modest reagent adjustments, and high throughput experimentation (HTE) demonstrates the ease of reaction setup, enabling prompt screening of aryl halide coupling partners and conditions. Scores of thiol substrates and aryl entities were examined and effectively conjugated, suggesting further diverse, practical applications.

A central argument for school choice is that parents can choose schools wisely. This principle may underlie why lottery-based school evaluations have almost always reported positive or zero achievement effects. This paper reports on a striking counterexample to these results. We use randomized lotteries to evaluate the Louisiana Scholarship Program, a voucher plan that provides public funds for disadvantaged students to attend private schools. LSP participation lowers math scores by 0.4 standard deviations and also reduces achievement in reading, science, and social studies. These effects may be due in part to selection of low-quality private schools into the program. (JEL H75, I21, I22, I28)

Thioamides are single atom substitutions of the peptide bond that serve as versatile probes of protein structure. Effective use of thioamides requires a robust understanding of the impact that the substitution has on a protein of interest. However, the thermodynamic effects of thioamide incorporation have only been studied in small structural motifs, and their influence on secondary structure in the context of full-length proteins is not known. Here we describe a comprehensive survey of thioamide substitutions in three benchmark protein systems (calmodulin, the B1 domain of protein G, and collagen) featuring the most prevalent secondary structure motifs: α-helix, β-sheet, and polyproline type II helix. We find that in most cases, effects on thermostability can be understood in terms of the positioning and local environment of the thioamide relative to proximal structural elements and hydrogen bonding networks. These observations set the stage for the rational design of thioamide substituted proteins with predictable stabilities.

Site-specific fluorescence probes can be used to measure distances within proteins when used as part of a Förster resonance energy transfer (FRET) pair. Here we report the synthesis of a coumarin maleimide (Mcm-Mal) that is fluorogenic upon reaction with cysteine. We demonstrate that cysteine, acridonylalanine (Acd) double mutant proteins can be produced by unnatural amino acid mutagenesis and reacted with Mcm-Mal to generate Mcm/Acd labeled proteins for FRET studies. The Mcm/Acd FRET pair is minimally-perturbing, easy to install, and well-suited to studying protein distances in the 15-40 Å range. Furthermore, Mcm/Acd labeling can be combined with tryptophan fluorescence in three color FRET to monitor multiple interactions in one experiment.

We develop over-identification tests that use admissions lotteries to assess the predictive value of regression-based value-added models (VAMs). These tests have degrees of freedom equal to the number of quasi-experiments available to estimate school effects. By contrast, previously implemented VAM validation strategies look at a single restriction only, sometimes said to measure forecast bias. Tests of forecast bias may be misleading when the test statistic is constructed from many lotteries or quasi-experiments, some of which have weak first stage effects on school attendance. The theory developed here is applied to data from the Charlotte-Mecklenberg School district analyzed by Deming (2014).

We use admissions lotteries to estimate effects of attendance at Boston’s charter high schools on college preparation and enrollment. Charter schools increase pass rates on Massachusetts’ high-stakes exit exam, with large effects on the likelihood of qualifying for a state-sponsored scholarship. Charter attendance also boosts SAT scores sharply and increases the likelihood of taking an Advanced Placement (AP) exam, the number of AP exams taken, and AP scores. Charters induce a substantial shift from 2- to 4-year institutions, though the effect on overall college enrollment is modest. Charter effects on college-related outcomes are strongly correlated with charter effects on earlier tests.

Structural econometric methods are often criticized for being sensitive to functional form assumptions. We study parametric estimators of the local average treatment effect (LATE) derived from a widely used class of latent threshold crossing models and show they yield LATE estimates algebraically equivalent to the instrumental variables (IV) estimator. Our leading example is Heckman's (1979) two‐step (“Heckit”) control function estimator which, with two‐sided non‐compliance, can be used to compute estimates of a variety of causal parameters. Equivalence with IV is established for a semiparametric family of control function estimators and shown to hold at interior solutions for a class of maximum likelihood estimators. Our results suggest differences between structural and IV estimates often stem from disagreements about the target parameter rather than from functional form assumptions per se. In cases where equivalence fails, reporting structural estimates of LATE alongside IV provides a simple means of assessing the credibility of structural extrapolation exercises.

Thioamide substitutions in the backbones of proteins can modulate their structure and thermostability, or serve as spectroscopic probes in fluorescence quenching experiments. Using native chemical ligation, we have produced the first examples of a protein (calmodulin) containing two thioamides. Dithioamide variants were made to explore the effects of combining stabilizing, neutral, and destabilizing single thioamide substitutions. One of the dithioamide calmodulin variants exhibited stabilization greater than any monothioamide variant, although the effect could not easily be anticipated from the results of single substitutions. Each of the calmodulin variants retained the ability to bind a target peptide, and the dithioamide proteins exhibited an increase in fluorescence quenching of tryptophan relative to their single thioamide counterparts. These results show that multiply thioamidated proteins can be synthesized, and that properly placed thioamides can be used to increase protein thermostability or enhance fluorecsence quenching in peptide binding experiments.

This paper uses data from a randomized evaluation of Head Start to answer two questions: (i) How much do short-run causal effects vary across Head Start centers? and (ii) Do observed inputs explain this variation? I find that the cross-center standard deviation of cognitive effects is 0.18 test score standard deviations, which is larger than typical estimates of variation in teacher or school effectiveness. Centers offering full-day service and home visiting are more effective, while centers that draw more children from center-based preschool have smaller effects. Other key inputs, including the High/Scope curriculum, teacher education, and class size are not correlated with Head Start effectiveness. (JEL H75, I21, I28, J13, J24)

Increasing the postsecondary attainment rate of college-age youth is an important economic priority in the U.S. and in other developed countries. Yet little is known about whether different forms of public subsidy can increase degree completion. In this paper, we compare the impact of the marginal taxpayer dollar on postsecondary attainment when it is spent on lowering tuition prices versus increasing the quality of the college experience. We do so by estimating the causal impact of changes in tuition and spending on enrollment and degree completion in U.S. public postsecondary institutions between 1990 and 2013. We estimate these impacts using a newly assembled data set of legislative tuition caps and freezes, combined with variation in exposure to state budget shocks that is driven by differences in historical reliance on state appropriations. We find large impacts of spending on enrollment and degree completion. In contrast, we find no impact of price changes. Our estimates suggest that spending increases are more effective per-dollar than price cuts as a means of increasing postsecondary attainment.

Conventional value-added models (VAMs) compare average test scores across schools after regression-adjusting for students’ demographic characteristics and previous scores. This article tests for VAM bias using a procedure that asks whether VAM estimates accurately predict the achievement consequences of random assignment to specific schools. Test results from admissions lotteries in Boston suggest conventional VAM estimates are biased, a finding that motivates the development of a hierarchical model describing the joint distribution of school value-added, bias, and lottery compliance. We use this model to assess the substantive importance of bias in conventional VAM estimates and to construct hybrid value-added estimates that optimally combine ordinary least squares and lottery-based estimates of VAM parameters. The hybrid estimation strategy provides a general recipe for combining nonexperimental and quasi-experimental estimates. While still biased, hybrid school value-added estimates have lower mean squared error than conventional VAM estimates. Simulations calibrated to the Boston data show that, bias notwithstanding, policy decisions based on conventional VAMs that control for lagged achievement are likely to generate substantial achievement gains. Hybrid estimates that incorporate lotteries yield further gains.

Estimates using admissions lotteries suggest that urban charter schools boost student achievement, while charter schools in other settings do not. Using the largest available sample of lotteried applicants to charter schools, we explore student-level and school-level explanations for this difference in Massachusetts. In an econometric framework that isolates sources of charter effect heterogeneity, we show that urban charter schools boost achievement well beyond that of urban public school students, while non-urban charters reduce achievement from a higher baseline. Student demographics explain some of these gains since urban charters are most effective for non-whites and low-baseline achievers. At the same time, non-urban charter schools are uniformly ineffective. Our estimates also reveal important school-level heterogeneity within the urban charter sample. A non-lottery analysis suggests that urban charters with binding, well-documented admissions lotteries generate larger score gains than under-subscribed urban charter schools with poor lottery records. Finally, we link charter impacts to school characteristics such as peer composition, length of school day, and school philosophy. The relative effectiveness of urban lottery-sample charters is accounted for by these schools' embrace of the No Excuses approach to urban education.

The charter school landscape includes a variety of organizational models and a few national franchises. The nation’s largest network of charter schools is the Knowledge is Power Program (KIPP), with 80 schools operating or slated to open soon. KIPP schools target low income and minority students and subscribe to an approach some have called No Excuses (Abigail Thernstrom and Stephen Thernstrom 2003). No Excuses schools feature a long school day and year, selective teacher hiring, strict behavior norms, and encourage a strong student work ethic. KIPP schools have often been central in the debate over whether schools alone can substantially reduce racial achievement gaps. Descriptive accounts of KIPP suggest positive achievement effects (see, e.g., Jay Mathews 2009), but critics argue that the apparent KIPP advantage reflects differences between students who attend traditional public schools and students that choose to attend KIPP schools (see, e.g., Martin Carnoy, Rebecca Jacobsen, Lawrence Mishel, and Richard Rothstein 2005). There are few well-controlled studies of KIPP schools that might help sort out these competing claims, and none that focus on KIPP. This paper reports on a quasi-experimental evaluation of the only KIPP school in New England, KIPP Academy Lynn. KIPP Lynn opened in the fall of 2004 and is the only charter school in Lynn, Mass., a low income city north of

The economics literature provides considerable evidence on the performance of low-income housing programs. One issue that has not been adequately addressed by previous studies is the relative depreciation rates of subsidized and unsubsidized housing units. Simple economic theory suggests that subsidized projects should depreciate more quickly than unsubsidized units, as the rents received by owners of these projects do not depend directly on the condition of their units provided that they meet minimum housing standards. However, many government programs aimed at modernizing subsidized housing projects exist. Whether there is a difference between the depreciation rates of subsidized and unsubsidized housing units is therefore an empirical question. Using panel data from the American Housing Survey, I find that there is no significant difference in the depreciation rates of subsidized and unsubsidized housing units over the period from 1985 to 2005.

Conventional value-added models (VAMs) compare average test scores across schools after regression-adjusting for students’ demographic characteristics and previous scores. This article tests for VAM bias using a procedure that asks whether VAM estimates accurately predict the achievement consequences of random assignment to specific schools. Test results from admissions lotteries in Boston suggest conventional VAM estimates are biased, a finding that motivates the development of a hierarchical model describing the joint distribution of school value-added, bias, and lottery compliance. We use this model to assess the substantive importance of bias in conventional VAM estimates and to construct hybrid value-added estimates that optimally combine ordinary least squares and lottery-based estimates of VAM parameters. The hybrid estimation strategy provides a general recipe for combining nonexperimental and quasi-experimental estimates. While still biased, hybrid school value-added estimates have lower mean squared error than conventional VAM estimates. Simulations calibrated to the Boston data show that, bias notwithstanding, policy decisions based on conventional VAMs that control for lagged achievement are likely to generate substantial achievement gains. Hybrid estimates that incorporate lotteries yield further gains.

The E. coli aminoacyl transferase (AaT) can be used to transfer a variety of unnatural amino acids, including those with azide or alkyne groups, to the α-amine of a protein with an N-terminal Lys or Arg. Subsequent functionalization through either copper-catalyzed or strain-promoted click reactions can be used to label the protein with fluorophores or biotin. This method can be used to directly detect AaT substrates or in a two-step protocol to detect substrates of the mammalian ATE1 transferase.

The expression of proteins containing unnatural amino acids through suppression of a stop codon can be limited by truncation due to competition with release factors. When the site of incorporation is near the C-terminus, it may not be feasible to separate the full-length unnatural amino acid protein from the truncated form. We report a simple, traceless procedure that allows one to isolate the desired protein using a C-terminal intein fusion.

This paper empirically evaluates the cost-effectiveness of Head Start, the largest early- childhood education program in the United States. Using data from the randomized Head Start Impact Study (HSIS), we show that Head Start draws a substantial share of its participants from competing preschool programs that receive public funds. This both attenuates measured experimental impacts on test scores and reduces the program's net social costs. A cost-benefit analysis demonstrates that accounting for the public savings associated with reduced enrollment in other subsidized preschools can reverse negative assessments of the program's social rate of return. Estimates from a semi-parametric selection model indicate that Head Start is about as effective at raising test scores as competing preschools and that its impacts are greater on children from families unlikely to participate in the program. Efforts to expand Head Start to new populations are therefore likely to boost the program's social rate of return, provided that the proposed technology for increasing enrollment is not too costly.

This paper studies the demand for charter schools in Boston, Massachusetts, with an emphasis on comparative advantage in school choice. I model charter school application and attendance decisions in a generalized Roy selection framework that links students’ preferences to the achievement gains generated by charter attendance. I estimate the model using instruments based on randomized admission lotteries and distance to charter schools. The estimates show that students do not sort into charter schools on the basis of comparative advantage in academic achievement. Charter schools generate larger test score gains for disadvantaged, low-achieving students, but demand for charters is stronger among richer students and high achievers. Similarly, achievement benefits are larger for students with weaker unobserved preferences for charter schools. As a result, counterfactual simulations indicate that charter expansion is likely to be most effective when accompanied by efforts to target students who are currently unlikely to apply.

We use admissions lotteries to estimate the effects of attendance at Boston's charter high schools on college preparation, college attendance, and college choice. Charter attendance increases pass rates on the high-stakes exam required for high school graduation in Massachusetts, with especially large effects on the likelihood of qualifying for a state-sponsored college scholarship. Charter attendance has little effect on the likelihood of taking the SAT, but shifts the distribution of scores rightward, moving students into higher quartiles of the state SAT score distribution. Boston's charter high schools also increase the likelihood of taking an Advanced Placement (AP) exam, the number of AP exams taken, and scores on AP Calculus tests. Finally, charter attendance induces a substantial shift from two- to four-year institutions, though the effect on overall college enrollment is modest. The increase in four-year enrollment is concentrated among four-year public institutions in Massachusetts. The large gains generated by Boston's charter high schools are unlikely to be generated by changes in peer composition or other peer effects.

Estimates using admissions lotteries suggest that urban charter schools boost student achievement, while charter schools in other settings do not. Using the largest available sample of lotteried applicants to charter schools, we explore student-level and school-level explanations for this difference in Massachusetts. In an econometric framework that isolates sources of charter effect heterogeneity, we show that urban charter schools boost achievement well beyond that of urban public school students, while non-urban charters reduce achievement from a higher baseline. Student demographics explain some of these gains since urban charters are most effective for non-whites and low-baseline achievers. At the same time, non-urban charter schools are uniformly ineffective. Our estimates also reveal important school-level heterogeneity within the urban charter sample. A non-lottery analysis suggests that urban charters with binding, well-documented admissions lotteries generate larger score gains than under-subscribed urban charter schools with poor lottery records. Finally, we link charter impacts to school characteristics such as peer composition, length of school day, and school philosophy. The relative effectiveness of urban lottery-sample charters is accounted for by these schools' embrace of the No Excuses approach to urban education.

The nation’s largest charter management organization is the Knowledge is Power Program (KIPP). KIPP schools are emblematic of the No Excuses approach to public education, a highly standardized and widely replicated charter model that features a long school day, an extended school year, selective teacher hiring, strict behavior norms, and emphasizes traditional reading and math skills. No Excuses charter schools are sometimes said to target relatively motivated high achievers at the expense of students who are more difficult to teach, including limited English proficiency (LEP) and special education (SPED) students, as well as students with low baseline achievement levels. We use applicant lotteries to evaluate the impact of KIPP Academy Lynn, a KIPP school in Lynn, Massachusetts that typifies the KIPP approach. Our analysis focuses on special needs students that may be underserved. The results show average achievement gains of 0.36 standard deviations in math and 0.12 standard deviations in reading for each year spent at KIPP Lynn, with the largest gains coming from the LEP, SPED, and low-achievement groups. Average reading gains are driven almost entirely by SPED and LEP students, whose reading scores rise by roughly 0.35 standard deviations for each year spent at KIPP Lynn. C � 2012 by the Association for Public Policy Analysis and Management.