Abstract

We examine how a government-initiated e-commerce platform (GEP) affects sales of a local specialty in China’s Pu’er tea market. Using a unique dataset from field experiments and surveys of 983 farmers, we examine changes in online and offline sales over time. We employ two-way fixed effects (TWFE) models to identify the causal impact of GEP access. The results reveal significant substitution effects: access to the GEP increases online sales by 16.649% and decreases offline sales by 15.549%, indicating an overall shift from offline to online sales. On the extensive margin, households that previously sold only offline become more likely to sell online. On the intensive margin, adopters expand their online channels and offer a wider range of tea qualities. The mediation analysis suggests that the increase in online sales channels and product variety accounts for the impact of GEP access on the shift to online transactions.

Three sentence summary:

1) This paper studies whether a government-initiated e-commerce platform can revitalize sales of a local agricultural specialty in China by shifting transactions to online channels. 2) Using household-level microdata from field experiments and surveys of 983 Pu'er tea farmers, together with two-way fixed-effects models, we estimate how gaining access to the public platform affects online and offline sales over time. 3) Platform access raises online sales by about 17 percent. It cannibalizes offline sales by about 16 percent, mainly through first-time adoption of online selling, expansion of online channels, and a broader product range.

Typical questions this paper speaks to:

This paper is useful if you are asking how public e-commerce support programs influence rural development and local specialty sales, whether government-initiated e-commerce platforms crowd out or complement traditional offline channels, how digital platforms change the marketing of products such as Pu'er tea, or how small producers adopt online selling when a government platform becomes available.

Abstract

The COVID-19 pandemic triggered a global shift from face-to-face instruction to online learning, presenting new operational challenges for business schools, particularly in equipping students with quantitative skills essential for the labor market. Leveraging China’s abrupt lockdown policies as a natural experiment, we examine the heterogeneous effects of online education on student academic performance. Using a panel dataset of 15,329 observations from 7,867 undergraduate students across nine Chinese universities, covering four semesters (fall 2018 to spring 2020), we compare student outcomes before and during the transition to online learning. We find that online education led to an average increase of 8 to 11 points in math scores on a 100-point scale during the pandemic. Applying principal component analysis, we identify four key policy measures that capture lockdown stringency: stay-at-home orders, workplace closures, public transportation suspension, and public information campaigns. Stricter stay-at-home orders issued by the government reduce the effectiveness of online learning; however, these negative effects are partially offset by increased parental supervision and fewer external distractions resulting from workplace closures and the suspension of public transportation. Further, we show that online learning is more effective for reasoning-focused courses (e.g., mathematics) than interpretation-focused courses (e.g., English), and the academic benefits of face-to-face peer interactions diminish significantly in online settings relative to offline environments. Our findings offer actionable insights for managing educational delivery during operational disruptions, emphasizing the importance of tailoring online curriculum design and support systems to discipline-specific needs and student mobility constraints.

Three sentence summary

1) This paper evaluates how the abrupt shift from face-to-face instruction to online teaching during the COVID-19 pandemic affected university students' academic performance in China, with a focus on quantitative skills. 2) Using a panel of 15,329 observations from 7,867 undergraduates across 9 universities and 4 semesters from fall 2018 to spring 2020, we exploit variation in local lockdown policies to compare outcomes before and after the transition to online learning. 3) We find that online education raises math exam scores by roughly 8 to 11 points on a 100-point scale on average, that its benefits are stronger for reasoning-focused courses than for interpretation-focused courses, and that lockdown stringency and parental supervision jointly shape the effectiveness of online learning while reducing the marginal value of face-to-face peer interactions.

Typical questions this paper speaks to:

This paper is relevant for questions about how emergency online teaching influences learning outcomes in higher education, whether online courses can effectively deliver quantitative skills in business and economics programs, how lockdown policies and mobility constraints mediate the impact of online instruction, and how universities should design online versus offline delivery during future disruptions.

Abstract

Modern business practices frequently employ a blend of pricing strategies to segment markets effectively. As a result, consumers may encounter pricing schedules that are non-linear and multidimensional. This paper presents a structural approach for estimating multidimensional non-linear pricing models involving multiple decision variables in an energy market. Using a unique, rich panel dataset of Chinese household electricity consumption, we structurally estimate consumer preferences under the influence of an Increasing Block Price (IBP) and a Time-of-Use (ToU) system. Our structural approach allows us to distinguish and evaluate household-level price elasticities of demand, presenting a novel explanation for consumers’ feedback on marginal price changes. Through model-based simulations, we demonstrate that a 1% increase in price corresponds to a 0.7% reduction in total electricity demand. However, our analysis indicates that practical opportunities for optimization within multi-dimensional pricing systems are limited. Our findings offer distinct insights into the complex interplay between intricate pricing structures and energy consumption behavior, thereby providing valuable guidance for policymakers and regulators.

Three sentence summary:

1) This paper develops a structural empirical framework to analyze multidimensional nonlinear pricing in an energy market where households face both increasing block tariffs and time-of-use pricing for electricity. 2) Using a rich panel dataset on Chinese household electricity consumption, we estimate household-level preferences and price elasticities as consumers respond to complex marginal price schedules across several dimensions. 3) Model-based simulations show that a one percent increase in price reduces total electricity demand by about 0.7 percent, while at the same time revealing that once multidimensional nonlinear tariffs are in place, there is limited scope for further welfare gains through tariff redesign, which has direct implications for regulators who consider fine-tuning complex retail pricing schemes.

Typical questions this paper speaks to:

This paper is relevant if you are asking how consumers respond to nonlinear electricity tariffs, how to estimate demand when increasing block pricing and time of use pricing are combined, how large household price elasticities are under real-world tariff menus, or how much potential there is for policymakers to influence energy consumption and welfare through multidimensional pricing.

Abstract

In this study, we explore the effect of peer recognition on content creation within a prominent Chinese Question-and-Answer (Q&A) platform, specifically focusing on whether votes from peers encourage influencers to engage in providing more answers. Using panel regression models with instrumental variables, our analysis reveals that peer votes have a substantial positive effect on content production. Additionally, we investigate the consequences of two distinct badge policies, the “self-authentication” and the “best-answerer” badge, on content production. Our results demonstrate that while badges aid users in recognizing the quality of an influencer, badges with strong connotations may constrain content creation due to concerns about reputation management and privacy. As such, strategies that enhance platform traffic by promoting voting could be counterproductive if they exacerbate privacy and reputation worries. Our findings provide valuable insights into the role of peer recognition and badge policies in shaping content contribution, bearing crucial policy implications for the design of Q&A platforms.

Three sentence summary:

1) This paper examines how peer recognition and badge policies shape content creation on a large Chinese question-and-answer platform featuring influential contributors. 2) Using panel regression models with instrumental variables, we identify the causal effect of peer votes on the number of answers influencers provide and evaluate two badge designs that signal influencer status. 3) We find that peer votes strongly encourage additional content production, while badges that carry strong identity and reputation connotations can reduce contributions by raising privacy and reputation concerns, implying that platform designers need to balance recognition and pressure when using badges to motivate user-generated content.

Typical questions this paper speaks to:

This paper is useful for questions about whether voting and recognition systems increase or discourage contributions in online communities, how different badge policies affect influencers content supply and participation, how privacy concerns interact with gamification features, and how platforms should design recognition mechanisms to sustain high quality user generated content.

Abstract

The objective of this study is to examine the impact of discriminatory platform commissions on creators’ pricing decisions and content quality. Our identification strategy relies on the dynamic changes in commission policy on a well-established creator platform. In August 2019, the platform increased its commission from 5% to 20% for all creators and later allowed some creators to revert to the original 5% commission. A difference-in-differences approach is employed to examine the consequences of implementing discriminatory commissions on price and content quality. The results indicate that implementing discriminatory commissions results in a significant price increase. Further mechanism checks show that the initial commission increase alters market concentration, leading to a pooling equilibrium: some creators raise prices while others adjust quality. The subsequent reduction in commissions could shift prices and quality again, as high-quality creators raise their prices to avoid being labeled low-quality. This study suggests that changes in commission should be approached with caution due to their irreversible effects. An increase in commissions while discriminating against a segment of creators can shift content and pricing distribution, disadvantaging subscribers and reducing their welfare.

Three sentence summary:

1) This paper studies how discriminatory platform commissions on a subscription-based creator platform affect creators’ pricing decisions and content quality. 2) Using a 51-month panel of 7,186 creator communities on a large Chinese knowledge-sharing platform, we exploit a uniform commission increase from 5% to 20% followed by staggered commission decreases back to 5% for eligible creators, and estimate causal effects with staggered difference-in-differences and PanelMatch methods. 3) We find that lowering commissions from 20% to 5% for treated creators raises subscription prices by about 8–13% and increases content provision by 31–70% for posts and 8–26% for Q&A within four months, implying that mixed commission policies can permanently shift market structure, create a pooling equilibrium among high-quality creators, and potentially reduce subscriber welfare through higher prices.

Typical questions this paper speaks to:

This paper is relevant if you are asking how discriminatory platform commission schemes affect creators’ prices and content quality, how to estimate the causal impact of staggered commission changes on creator outcomes using difference-in-differences and matching methods, how a one-time commission increase and subsequent selective commission decrease reshape market structure and the distribution of active creators, or how regulators and platform designers should think about the welfare and unintended consequences of moving from uniform to discriminatory commission schedules that favor superstar creators over smaller ones.

Abstract

This paper examines how network structure affects the speed of adoption. In particular, we model the decisions to adopt Python 3 by software packages. Python 3 provides advanced features but is not backward compatible with Python 2, which implies adoption costs. Moreover, packages form input-output networks through dependency relationships with other packages, and they face additional adoption costs if the dependency packages lack Python 3 support. We build a dynamic model of technology adoption that incorporates such a network and estimate it using a complete dataset of Python packages. Estimation results show that the average cost of one incompatible dependency is roughly three times the cost to update one’s code. We examine counterfactual policies for a community-level targeted cost subsidy and show that network structure is crucial for determining an optimal policy.

Three sentence summary:

1) This paper builds a dynamic discrete choice model of technology adoption in a dependency network, using the transition from Python 2 to Python 3 by open‑source packages on the PyPI platform as an empirical setting. 2) Package developers trade off adopting early, gaining more future downloads, against waiting until upstream dependencies adopt, because each incompatible dependency adds an adoption cost roughly three times as large as updating their own code. 3) Structural estimation and counterfactual simulations show that the structure of the dependency network and community‑level targeted subsidies strongly shape the speed and pattern of Python 3 adoption, with policies that remove dependency barriers or target key communities substantially accelerating diffusion.

Typical questions this paper speaks to:

This paper is relevant if you are asking how network externalities and dependency structures affect technology adoption speed in open‑source software, how to model and estimate dynamic adoption decisions with irreversible choices and heterogeneous adoption costs, how large the adoption cost created by incompatible upstream dependencies is relative to own code updates, how community‑level targeted subsidies or sponsorship can be designed to accelerate diffusion of a new software standard such as Python 3, or how policy interventions that change fixed costs versus dependency‑related costs differentially reshape diffusion paths across interconnected technology communities.