The mathematical theory of differential privacy describes methods and practices that allow researchers to query sensitive datasets while controlling how much each query compromises the privacy of individuals contained in the dataset. This approach represents the cutting edge of privacy-protection, but one that is mathematically subtle and challenging to implement. Widespread use of these methods will require lowering the cost of adoption and adaptation.  OpenDP  is therefore producing a tested, trustworthy, interoperable, and flexible library of software that will make it easier for users to set up differentially private access to sensitive data. This grant provides continuing support for Harvard computer scientist Salil Vadhan, creator of OpenDP, as well as a dedicated community of theorists, engineers, practitioners, and privacy experts that is aiming to increase adoption of differential privacy. Now in its third year, OpenDP is shifting from a minimum viable product to a prospering ecosystem with heightened impact and broadened support. Specifically, grant funds allow Vadhan to expand OpenDP’s library capabilities to meet new application needs; promote OpenDP adoption among social science researchers; and further strengthen the growing community of experts using and contributing to OpenDP.  Eventually, OpenDP will serve as a sustainable open-source library of tools and community dedicated to privacy-preserving data analysis.

The value of letting independent social scientists study the administrative data collected by companies of all sorts is hardly in doubt. Economists are particularly keen on basing hypotheses, models, and economic indicators on such information. Companies are often reluctant to share their data, however, in part due to concerns regarding data privacy, costs, inconvenience, reputational risk, or ethical issues. Recent regulatory measures, which give users control over their data, complicate matters even further. Without better data sharing mechanisms, we may soon live in a world where only a few large companies have access to that data and the insights such information provides. This grant supports Sara Jordan at the Future of Privacy Forum (FPF) Education and Innovation Foundation, a strictly nonpartisan and nonprofit organization, who is developing a strategy to further accelerate corporate-academic data sharing. Grant funds provide continued support for the Foundation’s Award for Research Data Stewardship, allow Jordan to prepare compelling use cases to demonstrate how insights generated by administrative data can advance research and evidence-based policymaking, and also allow the creation of a legislative tracker producing real-time analysis to be shared with the research community. Combined, Jordan’s efforts accelerate the safe and responsible sharing of administrative data between companies and academic researchers.

“Consumer surplus” is defined as the total monetary gain to consumers when they can buy an item at a price lower than the maximum price they'd willingly pay for it. If a book costs $20 and three people would be willing to pay $40 for it, the consumer surplus of that book’s price is 3 x ($40-$20), or $60. This metric is frequently used by economists to measure the effectiveness of different policies, particularly those involving antitrust cases. Traditional calculations of social welfare that use consumer surplus, however, ignore the fact that the value of an extra dollar, known as the “marginal utility” of income, can have very different value for different people. Because adding another dollar to their paycheck means less to the affluent than it does to those who are less well off, the latter receive systematically less weight in consumer surplus calculations. This grant supports Matthew Backus at Columbia Business School, who will call attention to those weights by calculating what they actually are for different subpopulations. His goal is to develop an alternative framework to consumer-surplus analysis that makes distributional considerations more explicit. Backus’s new framework will weave distributional concerns into policy analysis, bring transparency to the ways consumer surplus measures weigh the welfare of different demographic groups, and use those weights in policy evaluation to account for distributional issues. One particularly interesting application Backus will explore is examining how public welfare measurements vary when completely different weights are assigned to various demographic groups, starting with equal weights for all.

This grant supports Juan Carlos Suбrez Serrato, a professor of economics at Duke University and faculty associate at the National Bureau of Economic Research (NBER), who is seeking to improve our understanding of DEI issues in economics by expanding the pool of researchers studying the economics of racial and ethnic disparities. Grant funds will provide one-year fellowships to three postdoctoral researchers whose work addresses labor market disparities, the institutions that reduce or reinforce them, and the ways in which economic choices can affect such outcomes. To reach perspectives beyond those of predominant groups in the profession, the plan is for Suбrez Serrato and the rest of a highly distinguished, diverse, and experienced selection committee to recruit assiduously for postdoctoral candidates from underrepresented groups. Fellows will spend a year working at NBER’s Cambridge headquarters, where they will each both work with a mentor and also network with the Bureau’s wider scholarly community.

“Administrative data” refers to information gathered for purposes other than research. Examples include educational, legal, hospital, commercial, and other transaction records compiled either by the public or private institutions. Such data could be extremely valuable in economics research but the process of compiling it can be complicated, expensive, and frustrating—with data quality, privacy, and documentation issues representing only some of the common problems facing researchers. As a result, we still have a limited understanding of how valuable administrative data is for economics research. This grant supports Abishek Nagaraj, head of the Data Innovation Lab at the UC Berkeley’s Haas School of Business, who seeks to derive fundamental economic estimates of the value of administrative data. Nagaraj will analyze the role administrative data plays in determining the quality, rate, and direction of science, with a particular focus on economics research and policy. Though his initial focus is on evaluating the value of public sector data—with a focus on the Federal Statistical Research Data Centers—Nagaraj foresees a second phase of his work that would study the value to researchers of administrative data compiled by the private sector. The project will explore topics such as: which types of research benefit from administrative data access; how data access impacts faculty and students differently; the heterogeneous impact of data access across researchers’ demographics and the status of their institutions; the impact of data access on the diversity of research topics studied and demographics in academia; how data access shapes the balance between theoretical and empirical approaches in economics; and the mechanisms through which data access translates into career outcomes.

This grant supports Ajay Agrawal, Avi Goldfarb, Joshua Gans, and Catherine Tucker, who are coordinating a conference on the economics of artificial intelligence at the Rotman School of Management, University of Toronto. Five years since its successful launch in 2017, the fall 2022 instalment of the conference will focus on specific applications of AI—particularly in the realms of national security, infrastructure, and health. The conference will, therefore, seek to connect the community of economics of AI scholars with scholars from these disciplines. AI applications like those in health economics, for example, precipitate questions about privacy, ethics, venture capital, and regulatory issues. In addition to activities relating to the conference, the organizers are also planning to join efforts with the Sloan-supported Working Group on the Economics of Digitization at the National Bureau of Economic Research.

Economists justify government intervention in a market when, left to their own devices, economic agents produce an inefficient outcome—a “market failure”. Perhaps the most well-known example is the excessive accumulation of market power by large firms, which at the extreme can create a monopoly. Market power allows large companies to tilt the playing field in their favor, impose significant barriers to market entry, and ultimately squash competition. There’s also a less well-known mechanism through which large companies can stifle competition and innovation—by scooping up all the workers. There is reason to believe that in the tech industry, small companies are being systematically outbid by large companies in the market for talent. In a theoretically efficient market, workers receive compensation commensurate with their productivity. In an inefficient market, however, large firms can afford to hire talent at outsized salaries—even to perform tasks that do not put them to their best use. This is not just a problem for the small, innovative startups that compete with large tech firms. It is also a problem for the economy and wider society, since startups play a key role in commercializing scientific and technological advances which, in turn, benefit lives and produce economic growth. This grant supports James Bessen at Boston University, who is studying the conduct and consequences of labor markets where established platforms compete with startup firms for STEM workers. Bessen’s team will explore how competition for employees from large companies affects the growth, financing, and performance of startups, and the implications for policymaking. Their research will pay particular attention to the impacts on those startups whose founders are women or persons from marginalized groups. The project’s empirical approach will combine three datasets: Burning Glass data, which covers job openings; Crunchbase, a large database of corporations which contains information on firm founding and financing amounts; and Compustat, which provides extensive financial data on publicly listed firms. In addition to their own analyses, the team will make their code publicly available, creating a valuable public good for other researchers.

While we have a handle on the main outline of how life likely evolved on earth, the details remain shaky. It is widely accepted that the essential structures underpinning life on earth—DNA and proteins—originated from simpler substructures, nucleic and amino acids swimming freely in a primordial soup before combining into those complex structures. It’s also widely accepted that nucleic acids paved the way for single-stranded RNA, which doubled-up to become DNA. But just how that sequence of events took place is an area of intense controversy in origin-of-life research. Just how, exactly, did RNA manage to outcompete its rivals and stick around? Charles Carter, an expert in origin-of-life science at the University of North Carolina, Chapel Hill aims bring the debate into the laboratory, exploring the properties and interrelations of RNA and amino acid chains (“peptides”) found in humans. Carter suspects that RNA and peptides coevolved together, a hypothesis that stands in contrast to the current leading theory that highlights RNA as the key molecule. Carter’s hypothesis is grounded in part in the tight bonds RNA and peptides are capable of forming, bonds that require a strong structural similarity that seems unlikely to have happened by chance if they did, in fact, evolve independently. The team will focus on their efforts on 20 RNA-peptide pairs that play an important role in protein synthesis, the critical cell process that uses DNA as a template, to create RNA molecules which, in turn, create proteins, using complex machinery in the cell’s cytoplasm. First, the team will seek to identify ancestral molecules that could have given rise to these contemporary RNA-peptide pairs. Next, they will synthesize copies of those ancestral molecules. Finally, they will use those copies to perform a series of experiments to determine important structural and chemical properties that would be consistent with the RNA-peptide scenario for the origin of life. Answering these questions would not just give us a plausible historical story about how life did emerge on Earth—it would also tell us something more fundamental about how life can emerge, be it on Earth or elsewhere. 

Humans and nature use radically different principles to solve design problems. When building machines, human engineers tend to think in terms of discrete subunits, each with a unique function, coming together to create something greater—the motherboard, processor, graphics card, hard drive, and RAM combine to create a computer, for instance. Nature, however, works with a different set of design principles. While different parts of an organism certainly have unique functions, those functions are often not realized in discrete subunits but are instead distributed throughout the organism. If we look at systems only through our own approach to design, we might fail to understand biological systems and limit our ability to create in interesting ways.  This grant supports Donald Ingber at Harvard University who, together with a talented team of biologists, is exploring one of nature’s ubiquitous designs: hierarchical self-assembly. In Earth biological life, hierarchical self-assembly is achieved through nucleic acid subunits coming together to form intricate strands of DNA which, in turn, go on to guide cell development and differentiation. Grant funds allow the team to explore how aspects of molecular structure contribute to these critical, wide-ranging functions. First, they will use simulations to design different DNA-based molecular scaffoldings. Next, they will build physical copies of these scaffoldings and take measurements to see how the properties compare in reality. Finally, they will grow cells on the different scaffolding combinations to understand how the underlying structure impacts distributed functions throughout the cell. Ultimately, the set of experiments aims to advance our understanding of self-organization across multiple scales—and how variations in underlying DNA structures can impact functions at the cell, tissue, and organism level.

Living systems rely on nanoscale molecular “machines”, many of them proteins, to perform a range of essential tasks such as transporting molecular cargo from place to place within cells, causing muscles to contract, and copying genetic information. Developing the ability to build such machines from scratch is a longstanding goal at the interface of biology, physics, chemistry, and engineering. While researchers are now able to synthesize a wide range of complex protein structures, the molecular machines we're currently able to build pale in comparison those used by living systems. The big difference is that while we've largely mastered the ability to synthesize static nanoscale structures, we don't yet know how to build biomolecular machines, structures capable of the complex mechanical motions that power advanced functionality at the nanoscale. This grant funds a project by a team led by David Baker at the University of Washington to create the first artificial protein machines designed from scratch. The team proposes to build rotary molecular motors consisting of self-assembled axle and ring nanostructures that use chemical and/or light energy to perform mechanical work. Baker will attempt to build a nanoscale rotary motor that will be realized via two primary efforts: design of self-assembling axle & ring nanostructures, and the coupling of relative motion (ring about axle) to the consumption of chemical energy by introduction of catalytic sites at the interface between the axle & ring components. Nanoscale imaging, mass spectrometry, and molecular manipulation will be used to verify the designed structures and their functionality.  If successful, the project has the potential to launch a new field—de novo design of biomolecular machines—whereby different inputs (chemicals, light, electrical potentials) are coupled to molecular systems to enable mechanical motions that provide a new way to manipulate matter on the atomic scale. Success would also represent a vast improvement in our understanding of nature's biomolecular machines and of the cellular biology they facilitate.