Digital Commons Network^™

Cross Section Curve Autoregression: The Unit Root Case, Peter C.B. Phillips, Liang Jiang

Cowles Foundation Discussion Papers

This paper is part of a joint study of parametric autoregression with cross section curve time series, focussing on unit root (UR) nonstationary curve data autoregression. The Hilbert space setting extends scalar UR and local UR models to accommodate high dimensional cross section dependent data under very general conditions. New limit theory is introduced that involves two parameter Gaussian processes that generalize the standard UR and local UR asymptotics. Bias expansions provide extensions of the well-known results in scalar autoregression and fixed effect dynamic panels to functional dynamic regressions. Semiparametric and ADF-type UR tests are developed with corresponding limit theory …

Selective Turnout, Voting Policy, And Partisan Bias: Evidence From Multi-Level Data, Steven T. Berry, Christian Cox, Philip A. Haile

Cowles Foundation Discussion Papers

We study voting in general elections for the U.S. House of Representatives. Our data set includes demographics and turnout of all registered voters for the years 2016–2020, as well as vote shares at the precinct and contest level. We estimate a Downsian voting model incorporating rich observed and unobserved heterogeneity at the voter and contest level. We find that voters with high perceived voting costs tend to favor Democrats, as do marginal voters in most districts. Variation in state voting policies accounts for a modest share of overall estimated voting costs but is sufficient to determine the majority party in …

Outsourcing, Labor Market Frictions, And Employment, Mayara Felix, Michael B. Wong

Cowles Foundation Discussion Papers

We estimate the labor market impacts of Brazil’s 1993 outsourcing legalization us-ing North-South variation in pre-legalization court permissiveness, and comparing security guards to less-affected occupations. We find that outsourcing legalization persistently reallocated jobs from older incumbent guards to younger entrants. Total employ-ment of guards and their entry from informality persistently increased, while average demographic-adjusted wages remained constant. Meanwhile, a wave of occupational layoffs displaced some incumbent guards from high-wage firms. The evidence suggests that the rise of non-core activity outsourcing reduced labor market frictions, facilitated by firm-level economies of scale in human resources and spillovers to non-adopting firms.

College Application Mistakes And The Design Of Information Policies At Scale, Tomás Larroucau, Ignacio A. Rios, Anaïs Fabre, Christopher Neilson

Cowles Foundation Discussion Papers

We examine whether large-scale information interventions can improve college application outcomes in a centralized admissions system. Using nationwide surveys from Chile, we document widespread information frictions and frequent application mistakes, such as omitting attainable preferred programs or failing to include safety options. To address these frictions, we partnered with the Ministry of Education to implement a large-scale field experiment that provided applicants with personalized information on admission probabilities and program characteristics through customized online platforms. The intervention increased the probability that previously unmatched students received an assignment by 44% and improved placement into higher-ranked programs by 20%. Building on these …

Selection In Surveys: Using Randomized Incentives To Detect And Account For Nonresponse Bias, Deniz Dutz, Ingrid Huitfeldt, Santiago Lacouture, Magne Mogstad, Alexander Torgovitsky, Winnie Van Dijk

Cowles Foundation Discussion Papers

We show how to use randomized participation incentives to test and account for nonresponse bias in surveys. We first use data from a survey about labor market conditions, linked to full-population administrative data, to provide evidence of large differences in labor market outcomes between survey participants and nonparticipants, differences which would not be observable to an analyst who only has access to the survey data. These differences persist even after correcting for observable characteristics. We then use the randomized incentives in our survey to directly test for nonresponse bias, and find evidence of substantial bias. Next, we apply a range …

Lessons Learned: Luis Jácome, Mercedes Cardona

Journal of Financial Crises

Luis Jácome was appointed president of the board of Ecuador’s central bank in 1998 by newly elected President Jamil Mahuad. He and other members of the board resigned in 1999 in protest against a number of crisis-intervention measures they saw as threatening the bank’s independence to set monetary policy. Since the 1970s, Ecuador’s economy had experienced a period of growth fueled by oil exports, but by the mid-1990s the economy was reeling from a series of shocks, among them: a sharp drop in the price of oil, the effects of severe flooding on the country’s agricultural production, and the cost …

Lessons Learned: Benoît Cœuré, Mercedes Cardona

Journal of Financial Crises

Benoît Cœuré held several positions in the French Treasury in the years leading to the Global Financial Crisis (GFC). He was an economic adviser to the director general of the French Treasury from 1997–2002, deputy chief executive and chief executive of the French debt management office from 2002–2007, and assistant secretary for multilateral affairs, trade, and development from 2007–2009. He served as chief economist and deputy director general in 2009–2011. He joined the European Central Bank (ECB) during the European Sovereign debt Crisis and was responsible for market operations, market infrastructure supervision and European and international relations as a member …

Lessons Learned: Miguel Carcaño, Mercedes Cardona

Journal of Financial Crises

During the Global Financial Crisis (GFC), Miguel Carcaño served as head of the Spanish Treasury’s Fund for Orderly Bank Restructuring, the authority in charge of managing the restructuring process of the country’s credit institutions. The fund, known today as the Spanish Executive Resolution Authority, is integrated into the European network led by the Single Resolution Board (SRB) of the European Union’s banking union. Carcaño has held a number of posts within the SRB and in 2022 became head of the Single Resolution Fund, the SRB’s emergency fund, which serves as backstop for institutions across the banking union’s 21 countries.

Lessons Learned: Mark Branson, Mercedes Cardona

Journal of Financial Crises

Mark Branson joined the Swiss Financial Market Supervisory Authority (FINMA) as head of the banking division in 2010, during the European Sovereign Debt Crisis. He became deputy director of FINMA in 2013 and was named director a year later. Although Switzerland is not a member of the European Union (EU) or its banking union, the nation participates in bilateral agreements that govern trade with the EU, its largest trading partner. In the wake of the Global Financial Crisis (GFC), it enacted a number of regulations to improve oversight of the financial sector. Branson left FINMA in 2021 to become head …

United States: Rhode Island Limited Bank Holiday, 1991, Ayodeji George, Sophia Alden

Journal of Financial Crises

In 1990, the Rhode Island Share and Deposit Indemnity Corporation (RISDIC) was a private mutual deposit insurance corporation funded by member institutions. Late that year, after the failures of two of its insured institutions in July and October, other RISDIC member institutions faced large depositor withdrawals, as concerns began to focus on the financial health of RISDIC itself. RISDIC had maintained inadequate reserves, and on December 31, 1990, it found itself lacking the resources to cover depositor withdrawals from member institutions. RISDIC leadership requested a state-appointed conservator, which meant that all its member institutions no longer had the deposit insurance …

Survey Of Bank Holidays And Fund Suspensions, Rosalind Z. Wiggins, Owen Heaphy, Anmol Makhija, Stella Schaefer-Brown, Greg Feldberg, Andrew Metrick

Journal of Financial Crises

In this paper, we analyze seven case studies involving bank holidays and two involving mutual fund suspensions produced by the Yale Program on Financial Stability. Our main purpose is to assist policymakers who are considering utilizing a bank holiday in designing the most effective program as efficiently as possible. We find that a bank holiday may be most useful when designing and implementing a comprehensive remedy to an underlying problem distressing banks, particularly when an exogenous shock rather than balance sheet weaknesses is the cause of general distress to the system. A holiday is also useful to “ring-fence” one or …

Efficient Difference-In-Differences And Event Study Estimators, Xiaohong Chen, Pedro H. C. Sant’Anna, Haitian Xie

Cowles Foundation Discussion Papers

This paper investigates efficient Difference-in-Differences (DiD) and Event Study (ES) estimation using short panel data sets within the heterogeneous treatment effect framework, free from parametric functional form assumptions and allowing for variation in treatment timing. We provide an equivalent characterization of the DiD potential outcome model using sequential conditional moment restrictions on observables, which shows that the DiD identification assumptions typically imply nonparametric overidentification restrictions. We derive the semiparametric efficient influence function (EIF) in closed form for DiD and ES causal parameters under commonly imposed parallel trends assumptions. The EIF is automatically Neyman orthogonal and yields the smallest variance among …

Complementarities In High School And College Investments, John Eric Humphries, Juanna Schrøter Joensen, Gregory Veramendi

Cowles Foundation Discussion Papers

This paper examines how high school specialization shapes college investment decisions and their subsequent returns through dynamic complementarities. Using Swedish administrative data, we estimate a dynamic Roy model that accounts for selection on multidimensional skills, family background, prior investments, and unobserved heterogeneity. We identify the model using rich skill measures and quasi-experimental variation in program popularity. For marginal students, STEM specialization in high school increases wages by 9%, with more than half this return attributed to dynamic complementarities that enhance the productivity of subsequent college investments. Consequently, we find that counterfactual policies encouraging high school STEM specialization generate twice the …

Tracing Transformations: (Digitized) World War Ii Correspondence Through The Lens Of The Records Continuum Model, Milan M. Van Lange, Carlijn Keijzer

Journal of Contemporary Archival Studies

This paper traces the transformations of a historical archival collection of personal wartime correspondence. Most of the records in this collection were created in the context of the German Occupation of the Netherlands during World War II (1940–1945) and/or the War of Independence in Indonesia (1945-1950). The collection was gathered and curated by the NIOD Institute for War, Holocaust, and Genocide Studies in Amsterdam from 1945 onwards and was digitised between 2020 and 2023. We navigate the convergence of past and future transformations of the records and illuminate the intricate web of considerations and decisions of historical preservation, archiving, digitisation, …

Scalable Targeting Of Social Protection: When Do Algorithms Out-Perform Surveys And Community Knowledge?, Emily Aiken, Anik Ashraf, Joshua E. Blumenstock, Raymond P. Guiteras, Ahmed Mushfiq Mobarak

Discussion Papers

Innovations in big data and algorithms are enabling new approaches to target interventions at scale. We compare the accuracy of three different systems for identifying the poor to receive benefit transfers — proxy means-testing, nominations from community members, and an algorithmic approach using machine learning to predict poverty using mobile phone usage behavior — and study how their cost-effectiveness varies with the scale and scope of the program. We collect mobile phone records from all major telecom operators in Bangladesh and conduct community-based wealth rankings and detailed consumption surveys of 5,000 households, to select the 22,000 poorest households for $300 …

Scalable Targeting Of Social Protection: When Do Algorithms Out-Perform Surveys And Community Knowledge?, Emily Aiken, Anik Ashraf, Joshua E. Blumenstock, Raymond Guiteras, Ahmed Mushfiq Mobarak

Cowles Foundation Discussion Papers

Innovations in big data and algorithms are enabling new approaches to target interventions at scale. We compare the accuracy of three different systems for identifying the poor to receive benefit transfers — proxy means-testing, nominations from community members, and an algorithmic approach using machine learning to predict poverty using mobile phone usage behavior — and study how their cost-effectiveness varies with the scale and scope of the program. We collect mobile phone records from all major telecom operators in Bangladesh and conduct community-based wealth rankings and detailed consumption surveys of 5,000 households, to select the 22,000 poorest households for $300 …

Quantile-Optimal Policy Learning Under Unmeasured Confounding, Zhongren Chen, Siyu Chen, Zhengling Qi, Xiaohong Chen, Zhuoran Yang

Cowles Foundation Discussion Papers

We study quantile-optimal policy learning where the goal is to find a policy whose reward distribution has the largest α-quantile for some α P p0, 1q. We focus on the offline setting whose generating process involves unobserved confounders. Such a problem suffers from three main challenges: (i) nonlinearity of the quantile objective as a functional of the reward distribution, (ii) unobserved confounding issue, and (iii) insufficient coverage of the offline dataset. To address these challenges, we propose a suite of causal-assisted policy learning methods that provably enjoy strong theoretical guarantees under mild conditions. In particular, to address (i) and (ii), …

Ypfs Lessons Learned Oral History Project: An Interview With William White, William White, Mercedes Cardona

Documents

No abstract provided.

Towards Multi-Modal Multi-Document Understanding Capabilities In Foundation Models, Chuhan Li

Computer Science Theses

Contemporary foundation models are predominantly evaluated on isolated documentor image-understanding tasks, thereby overlooking the inherent multimodal multi-document reasoning that characterizes scientific inquiry. To bridge this gap, M3SCIQA is introduced, aMulti-Modal,Multi-document Scientific Question Answering benchmark crafted to test foundation models in practical scientific research settings. A comprehensive evaluation of 18 leading foundation models shows a substantial performance gap between models and human experts. Detailed error analysis reveals persistent deficiencies in both scientific visual reasoning tasks and long-range retrieval. Addressing the former, SPACECUE offers a concise yet effective visual prompting that overlays grid coordinates and Semantic-SAM masks …

Towards Practical And Real-Time Decoding Of Quantum Hypergraph Codes, Binghong (Leo) Li

Computer Science Theses

Quantum error correction (QEC) enables scalable quantum computation by detecting and correcting physical errors. However, decoding remains a key bottleneck—particularly for quantum low-density parity-check (qLDPC) codes, whose hypergraph structures demand complex reasoning. Most existing decoders are either too slow for real-time use or lack formal guarantees, and often struggle to generalize across diverse quantum hardware.

This thesis introduces the Minimum-Weight Parity Factor (MWPF) algorithm, a unified and certifiable decoding formulation that extends minimum-weight perfect matching to general hypergraph-based stabilizer codes. We focus on making MWPF practical and performant through a two-phase decoding architecture, consisting of a fast search phase and …

Exploring The Limits Of Multimodal Foundation Models For Visual Temporal Reasoning And Gesture Recognition Tasks, Ziyao Shangguan

Computer Science Theses

Multimodal foundation models (MFMs) have demonstrated impressive capabilities in static vision-language tasks such as image captioning, video summarization, and cross modal retrieval. However, their ability to reason over time—especially in gesture-rich video inputs—remains limited. This thesis investigates the temporal reasoning capabilities of MFMs in the context of gesture understanding, a critical component for enabling more expressive human-robot interaction. Through a preliminary study, we show that prompting-based strategies offer only marginal improvements in temporal reasoning, despite producing accurate frame-by-frame descriptions.

To more rigorously evaluate these limitations, we introduce TOMATO, a benchmark designed to assess visual …

Gene X Environment Interactions: Polygenic Scores And The Impact Of A Early Childhood Intervention In Colombia, Orazio Attanasio, Gabriella Conti, Pamela Jervis, Costas Meghir, Aysu Okbay

Discussion Papers

We evaluate impacts heterogeneity of an Early Childhood Intervention, with respect to the Educational Attainment Polygenic Score (EA4 PGS) constructed from DNA data based on GWAS weights from a European population. We find that the EA4 PGS is predictive of several measures of child development, mother’s IQ and, to some extent, educational attainment. We also show that the impacts of the intervention are significantly greater in children with low PGS, to the point that the intervention eliminates the initial genetic disadvantage. Lastly, we find that children with high PGS attract more parental stimulation; however, the latter increases more strongly in …

Revisiting The Lasting Impacts Of Incarceration, John Eric Humphries, Cécile Macaire, Aurélie Ouss, Megan Stevenson, Winnie Van Dijk

Cowles Foundation Discussion Papers

Using newly-linked administrative and commercial data from Virginia spanning 25 years, we study the consequences of incarceration. While previous research has examined labor market outcomes and recidivism, we focus on two of the primary channels through which low-income households build wealth: asset ownership (homes and cars) and human capital formation. To identify causal effects, we use a matched differencein-differences design. In line with much of the literature on the impact of incarceration in the U.S., we find no evidence of scarring effects on labor market outcomes or changes in recidivism beyond the incapacitation period. However, we find that incarceration leads …

Ypfs Lessons Learned Oral History Project: An Interview With Paul Tucker, Paul Tucker, Rosalind Z. Wiggins, Salil Gupta

Documents

No abstract provided.

Value Creation And Value Capture In Indian Garment Sector Bargaining, Russell Morton

Discussion Papers

This paper examines creation and distribution of surplus from global value chains (GVCs) in low- and middle-income country (LMIC) domestic supply chains. While GVC participation can enhance growth and productivity, low prices paid to small input suppliers raise concerns that gains from GVC participation accrue to the large exporters (the buyers). Supply-chain transactions often occur in bargained agreements with non-price terms that increase small supplier surplus, such as quantity stability and other insurance-like terms. Therefore, low input prices reflect both buyers’ share of surplus generated by non-price terms and buyer capture. I enrich a Nash bargaining model to study how …

From Delay To Payday: Easing Bureaucrat Access To Implementation Information Strengthens Social Protection Delivery, Eric Dodge, Yusuf Neggers, Rohini Pande, Charity Troyer Moore

Discussion Papers

Poor implementation frequently impedes social protection delivery in low-state-capacity settings. In such environments, reducing managers’ information acquisition costs may improve their grasp of frontline issues and program performance. However, in the presence of managerial rent-seeking, increasing the principal’s information may be key, and doing so just for managers can backfire. To assess this trade-off and implications for program performance, we collaborated with two Indian states to randomly vary bureaucrat access to PayDash, a digital platform for real-time tracking of worker payment processing for the national rural workfare program. In treated districts, PayDash expedited bureaucrat processing of workfare payments by 17%, …

Lessons Learned: J. Christopher Flowers, Mary Anne Chute Lynch

Journal of Financial Crises

J. Christopher Flowers has been managing director, CEO, and chairman of the private investment firm J.C. Flowers & Co. LLC for many years. During the Global Financial Crisis (GFC) of 2007–2009, Flowers was involved with investing in some of the largest banks and financial institutions in the world and advising and consulting with them on possible acquisitions, mergers, and sales as several of these firms began to collapse. In the fall of 2008, Flowers worked closely with the Bank of America (BofA) on proposals to acquire Lehman Brothers and Merrill Lynch, and he developed a plan for private investors to …

Lessons Learned: Giorgio Gobbi, Mercedes Cardona

Journal of Financial Crises

Giorgio Gobbi joined the Economic Research Department of the Bank of Italy in 1990 and was assigned to carry out analysis and research on the banking industry. From 1998 to 2004, he headed the department’s Financial Intermediaries Office and represented the bank at the International Monetary Fund, the Bank of International Settlements, and the European Central Bank. Starting in 2007, Gobbi headed the Financial Structure and Intermediaries Division within the bank’s Structural Economic Analysis Department. He was appointed deputy head of the Financial Stability Unit in 2013 and became head of the Financial Stability Directorate in 2014.

United States: Lehman Brothers Broker-Dealer Emergency Liquidity Program, 2008, Ayodeji George, Steven Kelly

Journal of Financial Crises

On Sunday, September 14, 2008, a deal to sell the United States investment bank Lehman Brothers Holdings Inc. (LBHI) to United Kingdom–based Barclays fell apart. US authorities informed LBHI that, given the lack of rescue funds, it would need to file for bankruptcy before Monday morning to avoid additional chaos for the firm and markets. However, authorities understood Barclays was still interested in buying Lehman’s broker-dealer subsidiary, Lehman Brothers Inc. (LBI). Federal Reserve and Treasury officials were concerned about the impact that the sudden failure of LBI could have on financial markets. LBI had $87 billion in secured overnight repurchase …

United States: First Republic Bank Emergency Liquidity Program, 2023, Salil Gupta, Jack French, Steven Kelly

Journal of Financial Crises

First Republic Bank, a California-based institution with $212.6 billion in assets, lost $25 billion in deposits on Friday, March 10, 2023, following the closing of Silicon Valley Bank that morning. On Sunday, March 12, First Republic announced that it had access to $70 billion in unused liquidity owing to its borrowing capacity at the Federal Reserve, the Federal Home Loan Bank System, and JPMorgan Chase (JPMC). But First Republic’s depositors withdrew a further $40 billion of deposits on Monday, March 13. First Republic’s borrowings from the Federal Reserve rose as high as $109 billion between Friday, March 10, and Wednesday, …