Estimating Conditional Average Treatment Effects (CATE) at the individual level is central to precision marketing, yet systematic benchmarking of uplift modeling methods at industrial scale remains limited. We present UpliftBench, an empirical evaluation of four CATE estimators: S-Learner, T-Learner, X-Learner (all with LightGBM base learners), and Causal Forest (EconML), applied to the Criteo Uplift v2.1 dataset comprising 13.98 million customer records. The near-random treatment assignment (propensity AUC = 0.509) provides strong internal validity for causal estimation. Evaluated via Qini coefficient and cumulative gain curves, the S-Learner achieves the highest Qini score of 0.376, with the top 20% of customers ranked by predicted CATE capturing 77.7% of all incremental conversions, a 3.9x improvement over random targeting. SHAP analysis identifies f8 as the dominant heterogeneous treatment effect (HTE) driver among the 12 anonymized covariates. Causal Forest uncertainty quantification reveals that 1.9% of customers are confident persuadables (lower 95% CI > 0) and 0.1% are confident sleeping dogs (upper 95% CI < 0). Our results provide practitioners with evidence-based guidance on method selection for large-scale uplift modeling pipelines.

Financial statement auditing is conducted under a risk-based evidence approach to obtain reasonable assurance. In practice, auditors often perform additional sampling or related procedures when an initial sample does not provide a sufficient basis for a conclusion. Across jurisdictions, current standards and practice manuals acknowledge such extensions, while the statistical design of sequential audit procedures has not been fully explored. This study formulates audit sampling with additional, sequentially collected items as a sequential testing problem for a finite population under sampling without replacement. We define null and alternative hypotheses in terms of a tolerable deviation rate, specify stopping and decision rules, and formulate exact sequential boundary conditions in terms of finite-population error probabilities. For practical implementation, we calibrate those boundaries by Monte Carlo simulation at least-favorable deviation rates. The exact design yields ex ante control of decision error probabilities, and the simulation-based implementation approximates that design while allowing the computation of expected stopping times. The framework is most naturally suited to attribute auditing and deviation-rate auditing, especially tests of controls, and it can be extended to one-sided, two-stage, and truncated designs.

A central challenge in mechanism design is to identify mechanisms whose performance is robust under uncertainty about the environment. The maxmin optimality criterion is commonly used for this purpose, but it often yields a large and economically uninformative set of mechanisms. This paper proposes a lexicographic approach to refining the maxmin criterion and characterizes the efficiency of optimal mechanisms. In canonical screening and auction environments, the strongest refinement $\unicode{x2013}$ proper robustness $\unicode{x2013}$ selects ex post efficient mechanisms. By contrast, in a public good provision environment, it identifies the precise form of optimal inefficiencies, which become severe in large economies.

The value of proof-of-work cryptocurrencies critically depends on miners having incentives to follow the protocol. However, the Bitcoin mining protocol proposed by Nakamoto (2008) and implemented in practice is well known not to constitute an equilibrium: Eyal and Sirer (2018) construct a profitable deviation called ``selfish mining'' which relies on strategically delaying disclosure of newly mined blocks rather than publishing them immediately. We propose inertial mining, a novel mining protocol. When miners follow inertial mining, they produce the outcome intended by Nakamoto, i.e., a single longest chain. But unlike the Bitcoin mining protocol, inertial mining constitutes an equilibrium (assuming no miner controls more than half of the mining power). Indeed, neither selfish mining nor any other deviation is profitable. Furthermore, inertial mining only changes miners' behavior in the event of off-path forks, and can be implemented in Bitcoin without any changes to its consensus mechanism or blockchain architecture.

Direct Air Carbon Capture and Storage (DACCS) can mitigate hard-to-abate emissions, e.g. from transport or industry. However, there is a wide variety of cost estimates for DACCS, driven, to a significant extent, by differences in electricity cost. At the same time, there is a notable gap in research that integrates direct air capturing systems into long-term energy system models. We separate direct air capturing, carbon transport, and carbon storage and integrate them into a European capacity expansion model for a fully decarbonised electricity system in 2050. We explore how two dimensions affect the total system costs of DACCS. The first dimension is the availability of CO2 storage locations: In one analysis, storage locations are restricted to offshore storage locations in the North Sea only, i.e. depleted natural gas fields. The alternative analysis comprises suitable storage locations distributed across Europe, including onshore. We find that limiting CO2 storage to North Sea sites increases overall capture costs by approximately 10 %. The second dimension is whether DACCS is analysed as stand-alone or integrated into the electricity system. We differentiate between three alternatives: fully isolated, fully integrated, and retrospectively added to an existing system. We find that neglecting system integration - i.e. treating direct air capture system as a stand-alone technology - increases capture costs by up to 30 %.

This paper studies a dominance relation among scoring rules with respect to avoiding the selection of the Condorcet loser. In a voting model with three or more alternatives, we say that a scoring rule $f$ Condorcet-loser-dominates (CL-dominates) another scoring rule $g$ if the set of profiles where $f$ selects a Condorcet loser is a proper subset of the set where $g$ does. We show that the Borda rule not only CL-dominates all other scoring rules, but also is the only scoring rule that CL-dominates some scoring rule.

We estimate the effect of cigarette price and tax increases on smoking rates using Eurobarometer survey data from 27 European Union countries between 2012 and 2020. Following a difference-in-differences approach, we compare individuals exposed to large price and tax increases with those in stable price and tax environments. Estimation is based on a difference-in-differences estimator with double machine learning, which relaxes the functional form assumptions typically imposed by parametric approaches such as two-way fixed effects. Our results indicate that tax increases reduce smoking rates among individuals who smoke at least once per month and among daily smokers. The reduction is primarily driven by individuals aged 15-24. We examine the sensitivity of our findings to functional form assumptions and treatment definitions. While estimates are robust to alternative functional form assumptions, they are sensitive to whether the treatment is defined as binary or continuous.

This paper introduces a novel characteristics-based specification for linear demand to investigate endogenous product design. Characteristics are allowed to affect both consumers' product valuations and to what extent these compete. I demonstrate how such a specification can help linear demand deliver fresh insights in how firms may optimally design existing goods, as well as predict demand for new products. The framework is novel in its broad applicability to settings with any finite number of goods, firms, and product characteristics, with both vertical and horizontal differentiation across different market structures, and under asymmetry.

We estimate the impact of environmental law enforcement on violence in the Brazilian Amazon. The introduction of the Real-Time Deforestation Detection System (DETER), which enabled the government to monitor deforestation in real time and issue fines for illegal clearing, significantly reduced homicides in the region. To identify causal effects, we exploit exogenous variation in satellite monitoring generated by cloud cover as an instrument for enforcement intensity. Our estimates imply that the expansion of state presence through DETER prevented approximately 1,477 homicides per year, a 15\% reduction in homicides. These results show that a replicable environmental enforcement policy produces social benefits.

The credibility revolution advances the use of research designs that permit identification and estimation of causal effects. However, understanding which mechanisms produce measured causal effects remains a challenge. The dominant current approach to the quantitative evaluation of mechanisms relies on the detection of heterogeneous treatment effects (HTEs) with respect to pre-treatment covariates. This paper develops a framework to understand when the existence of such heterogeneous treatment effects can support inferences about the activation of a mechanism. We show first that this design cannot provide evidence of mechanism activation without additional, generally implicit, exclusion assumptions. Further, even when these assumptions are satisfied, the presence of HTEs supports the inference that mechanism is active but the absence of HTEs is generally uninformative about mechanism activation. We provide novel guidance for interpretation and research design in light of these findings.

We establish the existence, uniqueness, and $W^{1,2,p}$-regularity of solutions to fully-nonlinear, parabolic obstacle problems when the obstacle is the pointwise supremum of functions in $W^{1,2,p}$ and the nonlinear operator is required only to be measurable in the state and time variables. In particular, the results hold for all convex obstacles. Applied to stopping problems, they provide general conditions under which a decision maker never stops at a convex kink of the stopping payoff. The proof relies on new $W^{1,2,p}$-estimates for obstacle problems when the obstacle is the maximum of finitely many functions in $W^{1,2,p}$.

Household welfare dynamics are often difficult to investigate due to lack of long-term panel data. Existing methods, such as pseudo-panel and synthetic panel, offer widely used solutions based on repeated cross-section designs, but they do not exploit within-household variation in rotating panel designs, which provide very useful information for estimating long-run dynamics. This paper applies grouped fixed effects (GFE) to estimate poverty mobility and persistence in a rotating panel setting, using National Household Survey on Living Conditions and Poverty (ENAHO) in Peru. Using observed transitions, we show that GFE-implied poverty transitions closely track the data. In a one-step-ahead validation that holds out each household's final observed year, predicted transition shares remain close to realized transition shares, indicating that the method captures short-run entry and exit dynamics out of sample. When benchmarked against synthetic panel point estimates, the GFE approach delivers transition measures that are closer to observed transitions on average, while also providing an interpretable grouping structure that supports richer descriptions of poverty persistence and mobility.

We study identification and inference in nonlinear dynamic systems defined on unknown interaction networks. The system evolves through an unobserved dependence matrix governing cross-sectional shock propagation via a nonlinear operator. We show that the network structure is not generically identified, and that identification requires sufficient spectral heterogeneity. In particular, identification arises when the network induces non-exchangeable covariance patterns through heterogeneous amplification of eigenmodes. When the spectrum is concentrated, dependence becomes observationally equivalent to common shocks or scalar heterogeneity, leading to non-identification. We provide necessary and sufficient conditions for identification, characterize observational equivalence classes, and propose a semiparametric estimator with asymptotic theory. We also develop tests for network dependence whose power depends on spectral properties of the interaction matrix. The results apply to a broad class of economic models, including production networks, contagion models, and dynamic interaction systems.

Production function estimates underpin the measurement of firm-level markups, allocative efficiency, and the productivity effects of policy interventions. Since Olley and Pakes (1996), every major proxy variable estimator has identified the production function through a first-order Markov assumption on unobserved productivity; I show that misspecification of this assumption generates persistent upward bias in the materials elasticity that propagates into overestimated markups and inflated treatment effects. I replace the Markov restriction with conditional independence across three intermediate input demands, a static condition grounded in input market segmentation, and establish nonparametric identification from a single cross-section. I develop a GMM estimator and establish consistency and asymptotic normality. Monte Carlo simulations confirm that the proposed estimator is unbiased across Markov and non-Markov environments, while the standard estimator exhibits persistent bias of up to 63 percent of the true materials elasticity. In 502 Japanese manufacturing industries, the proposed method yields systematically lower markups than the standard method across the entire distribution (median 0.93 vs. 1.03), reducing the share of industries with markups above unity from 54 to 37 percent. In a difference-in-differences analysis of the 2011 Tohoku earthquake, the standard method overstates the productivity loss by 0.40 percentage points, roughly $3.6 billion (400 billion yen) per year.

This study investigates how entrepreneurial orientation (EO) affects governance of SaaS platforms in SMEs, including strategy alignment and long-term governance performance. This study uses SaaS as a hybrid governance model to examine how transaction cost variables affect strategic alignment and how EO moderates these associations. The research uses multi-study design. Study 1 examined 180 UK and US entrepreneurs' survey data using PLS-SEM with reflecting constructs. Study 2 used a quasi-experimental approach using a secondary dataset from 238 European start-ups to operationalize variables using industry-based indicators. The study found an inverted U-shaped association between human asset specificity, SaaS usage frequency, and SMEs' strategic objectives. Risk-taking deepens the link between human asset distinctiveness and SaaS strategic alignment, while proactiveness strengthens the link to long-term success. Both studies show that SaaS strategic alignment has an inverted U-shaped connection with long-term performance, suggesting that excessive SaaS dependence may harm governance-enabled strategic outcomes. This paper introduces SaaS as a hybrid governance paradigm and examines its strategic influence on SMEs, utilizing transaction cost theory and EO perspectives. It shows the non-linear effects of SaaS adoption on strategic alignment and performance, emphasizing entrepreneurial decision-making in digital technology adoption.

Economic complexity - a group of dimensionality-reduction methods that apply network science to trade data - represented a paradigm shift in development economics towards materializing the once-intangible concept of capabilities as inferrable and quantifiable. Measures such as the Economic Complexity Index (ECI) and the Product Space have proven their worth as robust estimators of an economy's subsequent growth; less obvious, however, is how they have come to be so. Despite ECI drawing its micro-foundations from a combinatorial model of capabilities, where a set of homogeneous capabilities combine to form products and the economies which can produce them, such a model is consistent with neither the fact that distinct product classes draw on distinct capabilities, nor the interrelations between different products in the Product Space which so much of economic complexity is based upon. In this paper, we extend the combinatorial model of economic complexity through two innovations: an underlying network which governs the relatedness between capabilities, and a production function which trades the original binary specialization function for a fine-grained, product-level output function. Using country-product trade data across 216 countries, 5000 products and two decades, we show that this model is able to accurately replicate both the characteristic topology of the Product Space and the complexity distribution of countries' export baskets. In particular, the model bridges the gap between the ECI and capabilities by transforming measures of economic complexity into direct measures of the capabilities held by an economy - a transformation shown to both improve the informativeness of the Economic Complexity Index in predicting economic growth and enable an interpretation of economic complexity as a proxy for productive structure in the form of capability substitutability.

Traditional threshold-based stock networks suffer from subjective parameter selection and inherent limitations: they constrain relationships to binary representations, failing to capture both correlation strength and negative dependencies. To address this, we introduce statistically validated correlation networks that retain only statistically significant correlations via a rigorous t-test of Pearson coefficients. We then propose a novel structure termed the largest strong correlation balanced module (LSCBM), defined as the maximum-size group of stocks with structural balance (i.e., positive edge-sign products for all triplets) and strong pairwise correlations. This balance condition ensures stable relationships, thus facilitating potential hedging opportunities through negative edges. Theoretically, within a random signed graph model, we establish LSCBM's asymptotic existence, size scaling, and multiplicity under various parameter regimes. To detect LSCBM efficiently, we develop MaxBalanceCore, a heuristic algorithm that leverages network sparsity. Simulations validate its efficiency, demonstrating scalability to networks of up to 10,000 nodes within tens of seconds. Empirical analysis demonstrates that LSCBM identifies core market subsystems that dynamically reorganize in response to economic shifts and crises. In the Chinese stock market (2013-2024), LSCBM's size surges during high-stress periods (e.g., the 2015 crash) and contracts during stable or fragmented regimes, while its composition rotates annually across dominant sectors (e.g., Industrials and Financials).

The main goal of this paper is to settle a conceptual framework for cooperative game theory in which the notion of composition/aggregation of games is the defining structure. This is done via the mathematical theory of algebraic operads: we start by endowing the collection of all cooperative games with any number of players with an operad structure, and we show that it generalises all the previous notions of sums, products and compositions of games considered by Owen, Shapley, von Neumann and Morgenstern, and many others. Furthermore, we explicitly compute this operad in terms of generators and relations, showing that the Möbius transform map induces a canonical isomorphism between the operad of cooperative games and the operad that encodes commutative triassociative algebras. In other words, we prove that any cooperative game is a linear combination of iterated compositions of the 2-player bargaining game and the 2-player dictator games. We show that many interesting classes of games (simple, balanced, capacities a.k.a fuzzy measures and convex functions, totally monotone, etc) are stable under compositions, and thus form suboperads. In the convex case, this gives by the submodularity theorem a new operad structure on the family of all generalized permutahedra. Finally, we focus on how solution concepts in cooperative game theory behave under composition: we study the core of a composite and describe it in terms of the core of its components, and we give explicit formulas for the Shapley value and the Banzhaf index of a compound game.

Human-annotated preference data play an important role in aligning large language models (LLMs). In this paper, we study two connected questions: how to monitor the quality of human preference annotators and how to incentivize them to provide high-quality annotations. In current practice, expert-based monitoring is a natural workhorse for quality control, but it performs poorly in preference annotation because annotators are heterogeneous and downstream model performance is an indirect and noisy proxy for annotation quality. We therefore propose a self-consistency monitoring scheme tailored to preference annotation, and analyze the statistical sample complexity of both methods. This practitioner-facing analysis identifies how many inspected samples are needed to reliably assess an annotator and shows when self-consistency monitoring can outperform expert-based monitoring. We then use the resulting monitoring signal as the performance measure in a principal-agent model, which lets us study a second sample-complexity question: how many monitored samples are needed before simple contracts perform close to the ideal benchmark in which annotation quality is perfectly observable. Under this continuous action space, we show that this shortfall scales as $Θ(1/\sqrt{\mathcal{I} n \log n})$ for binary contracts and $Θ(1/(\mathcal{I}n))$ for linear contracts, where $\mathcal{I}$ is the Fisher information and $n$ is the number of samples; we further show that the linear contracts are rate-optimal among general contracts. This contrasts with the known result that binary contracts are optimal and of $\exp(-Θ(n))$ when the action space is discrete \citep{frick2023monitoring}.

Can large language models (LLMs) learn a decision maker's preferences from observed choices and generate preference-consistent recommendations in new situations? We propose a portable Simulate-Recommend-Evaluate framework that tests preference learning from revealed-choice data by comparing LLM recommendations with optimal choices implied by known preference primitives. We apply the framework to choice under uncertainty using the disappointment aversion model. Recommendation accuracy improves as models observe more choices, but learning is heterogeneous across preference types and LLMs: GPT learns risk aversion better than disappointment aversion, Gemini performs best in high disappointment-aversion regions, and Claude shows the broadest effective learning across parameter regions.