China’s market opening and employee ownership nudged firms to allocate labor more efficiently, and corporate AI adoption trimmed overstaffing while boosting patenting. Across listed firms, Stock Connect improved external monitoring of labor decisions, ESOPs lowered excess cash and aligned incentives, and AI adoption was followed by reduced labor over‑investment and stronger innovation outcomes.

Employment, capital market, and corporate policies

Li, Hanying · January 01, 2026 · Research Online (University of Wollongong)

openalex quasi_experimental medium evidence 8/10 relevance Summary only summary available; pdf_status=paywall DOI Source PDF

Structured author observations

Linked only from stored provider relations; the raw author line above is never matched by name.

OpenAlex

Latest observation: August 25, 2026

Li, Hanying provider ID

Using Chinese listed-firm panels, the thesis finds that Stock Connect liberalization and ESOP adoption improve firms' labor investment efficiency and corporate behavior, while firm-level AI adoption reduces labor over-investment and raises patent quantity and quality.

Citation observations

Cumulative provider counts captured on specific dates; providers are never combined.

0 cumulative citations

OpenAlex · Observed August 25, 2026

View corpus context

Over the past decades, the global labor market has experienced significant restructuring, reflected by persistent mismatches between labor supply and demand. While prior accounting and finance literature has predominantly focused on the capital component of Cobb-Douglas production function, little attention has been paid to labor input. Yet human capital is a more dynamic and innovative production factor in modern corporate governance, playing a decisive role in firm success and, ultimately, in economic growth. As a “proactive resource” that can freely decide whether to join, stay, and contribute efforts, human capital should be strategically managed and effectively incentivized, making firms’ labor investment decisions a critical area of inquiry. Understanding how such decisions vary across contexts and the factors that shape them remain an open question. This thesis thus aims to extend the literature on the determinants of labor investment and the consequences of employee treatments by examining the effects of nationwide reforms and firms’ strategic adjustments in the Chinese setting, consisting of three empirical studies.Chapter 2 investigates the impact of stock market liberalization on firms’ labor investment decisions. Many countries have progressively liberalized their stock markets by removing restrictions on foreign equity investment and encouraging cross-border capital flows, which facilitates foreign investors’ participation and monitoring, mitigates financial constraints, and improves the information environment for domestic firms. Exploiting the Mainland-Hong Kong Stock Connect program as an exogenous shock and employing a staggered difference-in-differences (DID) approach with a sample of 1,938 Chinese listed firms from 2010 to 2020, this chapter shows that stock market liberalization enhances labor investment efficiency by correcting both under-investment and over-investment in labors. This main effect remains robust after addressing endogeneity concerns, using alternative samples and different proxies for labor investment efficiency, and the consideration of non-labor investments. Improved stock liquidity and a more transparent information environment thus enhanced external monitoring, are the two key channels driving these improvements. Cross-sectional analyses indicate that the effect is stronger for non-state-owned enterprises (non-SOEs). Taking a broad view of labor investments, I find that stock market liberalization also fosters more employee-friendly practices through better wages and welfare. Overall, chapter 2 underscores the significant role of nationwide financial reform in shaping firms’ labor investment decisions.Chapter 3 uncovers an underexplored relation between firm’s strategic employee-related decisions and cash management policy. Framing employee stock ownership plans (ESOPs) as a mechanism for aligning the interests between employees and firms, this chapter investigates the impact of ESOPs on corporate cash holding. Exploiting the gradual adoption of ESOPs in China since 2014 as a quasi-natural experiment, this chapter shows that ESOPs adoption significantly reduces corporate cash holdings by generating incentive and signaling effects that mitigate precautionary motives and by enhancing internal governance to curb agency-driven motives for excessive cash retention. The negative effect is stronger among firms with fewer free-rider issues, greater rank-and-file employee participation, and longer employee stockholding durations. Furthermore, firms adopting ESOPs adjust their cash holdings more swiftly to optimal levels post-adoption, boosting cash’s market value. Chapter 3 highlights ESOPs’ distinctive role in aligning interests, contributing to the broader literature on employee incentive and corporate financial policies.Chapter 4 examines the role of artificial intelligence (AI) adoption in shaping labor investment efficiency in China. Featuring AI technology as a means of labor displacement and utilizing data from Chinese A-share listed firms from 2011 to 2021, I find that firms adopting AI technologies experience significant improvements in labor investment efficiency in the subsequent year, mainly by addressing labor over-investment issues, and achieve better innovation performance in terms of both quantity and quality of patents. Information asymmetry and corporate governance are two plausible channels through which AI adoption exerts influence, while financial constraints, labor skills, and industry competition moderate the effect. The results are robust when employing two-stage least square (2SLS) with instrumental variables, using alternative samples, adopting alternative variable definitions, and considering high-dimensional fixed effects. Chapter 4 highlights the labor-saving role of AI technology, contributes to the broader literature on the determinants of labor investment efficiency, and provides additional evidence on the firm-level impact of AI adoption.Overall, this thesis fits into the dynamics of the global labor market, drawing on the rapid development of China’s labor, capital, and technological markets, it offers additional evidence on factors that enhance labor investment efficiency and, consequently, improve firm performance. It also provides insights into the role of employee incentive plans in China in shaping corporate policies.

Summary

Main Finding

Nationwide financial liberalization, firm-level employee incentives, and firm adoption of AI each materially reshape firms’ labor investment efficiency and related corporate outcomes in China. Specifically: (1) stock market liberalization (Mainland–Hong Kong Stock Connect) improves labor investment efficiency and employee-friendly practices through better liquidity and transparency; (2) employee stock ownership plans (ESOPs) reduce excessive corporate cash holdings by aligning incentives and strengthening internal governance; and (3) firm-level AI adoption improves labor investment efficiency (mainly by correcting over-investment) and raises innovation quantity and quality. These effects are robust across multiple specifications and operate through distinct governance, information, incentive, and market channels.

Key Points

Chapter 2 — Stock market liberalization
- Natural experiment: Mainland–Hong Kong Stock Connect treated as exogenous shock.
- Sample: 1,938 Chinese listed firms, 2010–2020.
- Method: staggered difference-in-differences (DID).
- Finding: Liberalization corrects both labor under- and over-investment (raises labor investment efficiency); also increases wages and welfare (more employee-friendly).
- Mechanisms: improved stock liquidity and a more transparent information environment → stronger external monitoring.
- Heterogeneity: larger effects for non-SOEs.
- Robustness: alternative samples, alternate efficiency proxies, controls for non-labor investments, endogeneity checks.
Chapter 3 — ESOPs and corporate cash management
- Context: gradual ESOP adoption in China since 2014 as quasi-natural experiment.
- Main result: ESOP adoption significantly reduces corporate cash holdings.
- Mechanisms: incentive and signaling effects (reduce precautionary motives), improved internal governance (curbs agency-driven cash hoarding).
- Heterogeneity: stronger when free-rider problems are limited, when rank-and-file participation is higher, and with longer employee stockholding duration.
- Additional: firms adjust cash toward optimal levels faster post-adoption; market values cash more efficiently.
Chapter 4 — AI adoption and labor investment efficiency
- Sample: Chinese A-share listed firms, 2011–2021.
- Method: panel regressions with 2SLS instrumental-variable checks, high-dimensional fixed effects, various robustness checks.
- Finding: AI adoption leads to improved labor investment efficiency in the subsequent year, primarily by reducing labor over-investment; also increases patent quantity and quality.
- Mechanisms: reduced information asymmetry and improved corporate governance help reallocate labor; effects moderated by financial constraints, workforce skill composition, and industry competition.
- Robustness: alternative samples, variable definitions, IV estimation.
Cross-cutting
- Labor is a dynamic, strategic production factor that responds to macro reforms, governance innovations, and technology adoption.
- Different policy levers (market liberalization, employee incentives, AI diffusion) affect firms’ labor, financial, and innovation outcomes through complementary channels.

Data & Methods

Data sources: Chinese A-share/listed firm financials and disclosures; firm-level HR/payroll indicators and welfare proxies; corporate patent records; ESOP adoption announcements; timing of Mainland–Hong Kong Stock Connect.
Time windows:
- Chapter 2: 2010–2020 (1,938 listed firms).
- Chapter 3: post-2014 ESOP adoption era (panel of listed firms; exact sample varies by test).
- Chapter 4: 2011–2021 panel of A-share firms.
Identification strategies:
- Staggered DID exploiting phased roll-out of Stock Connect (Chapter 2).
- Quasi-experimental variation from staggered ESOP adoption (Chapter 3).
- Panel regressions with 2SLS/IV and firm & time fixed effects; robustness to high-dimensional fixed effects (Chapter 4).
Outcome measures:
- Labor investment efficiency: deviations of actual labor input from model-implied/optimal levels or other proxy measures (alternative definitions tested).
- Corporate cash holdings: cash/asset ratios and measures of deviation from optimal cash.
- Innovation: patent counts and quality metrics (citations, granting rates).
- Employee outcomes: wages, welfare expenditures.
Robustness and checks: alternative samples, alternative variable definitions, IV/2SLS, tests for heterogeneous effects, controls for non-labor investment shifts.

Implications for AI Economics

Firm-level AI adoption can act as a labor-displacing but efficiency-enhancing technology:
- AI reduces labor over-investment and reallocates labor toward more productive uses, generating measurable gains in innovation output.
- Policy and welfare assessments must account for both displacement and gains in firm-level productivity/innovation.
Interaction with governance and information environments matters:
- The effectiveness of AI in improving labor allocation is conditional on corporate governance quality and information transparency; weak governance or high information frictions can blunt AI’s reallocative gains.
- Financial-market reforms and employee incentive structures (e.g., ESOPs) can complement AI by strengthening monitoring and aligning incentives, facilitating labor reallocation and efficient cash use.
Distributional and transitional concerns:
- While firm performance and innovation rise, AI-driven correction of labor over-investment can have worker-level consequences (reallocation, upskilling needs). Policies should pair AI diffusion with training, social insurance, and incentives to preserve workforce attachment and re-employment.
Measurement and research directions for AI economics:
- Use firm-level adoption timing and granular HR data to separate substitution vs. complementarity effects across occupations and skill groups.
- Examine long-run dynamics: how do early AI adopters’ labor mixes, wage structures, and cash policies evolve relative to laggards?
- Investigate spillovers: do AI-induced labor reallocations at adopters affect local labor markets, wage-setting, or industry competition?
- Study interactions: how do financial liberalization and employee incentives modify the macro-level impact of AI on employment, inequality, and aggregate productivity?
Policy takeaways:
- Complement AI deployment with improved governance, employee participation mechanisms (e.g., ESOPs), and active labor-market policies to capture productivity gains while mitigating worker dislocation.
- Financial-market reforms that increase transparency and liquidity amplify firms’ capacity to adjust labor inputs efficiently in response to technological change.

Assessment

Paper Typequasi_experimental Evidence Strengthmedium — The thesis uses credible quasi-experimental tools (staggered DID, event studies, 2SLS) and extensive robustness checks across multiple chapters, supporting plausibly causal claims at the firm level; however, residual concerns remain about parallel trends in staggered DID designs, potential instrument validity/strength for 2SLS, measurement error in AI adoption and labor-investment proxies, and external validity beyond Chinese listed firms. Methods Rigorhigh — Multiple identification strategies are applied appropriately (staggered DID, event studies, IV/2SLS), high-dimensional fixed effects are included, various robustness checks and alternative specifications are reported, and heterogeneity and channel analyses are provided — indicating strong econometric rigor for observational data. SampleFirm-level panel of Chinese A-share listed companies roughly covering 2010–2021 (Chapter 2: sample of 1,938 listed firms, 2010–2020; Chapter 3: firms with staggered ESOP adoption since 2014; Chapter 4: 2011–2021 listed-firm panel with firm-year measures of AI adoption, labor inputs, cash holdings, and patent outcomes). Themesproductivity labor_markets adoption innovation IdentificationChapter 2: Staggered difference-in-differences exploiting the exogenous roll-out of the Mainland–Hong Kong Stock Connect as a shock to foreign investor access and monitoring (treated vs. control listed firms; event-study and robustness checks). Chapter 3: Quasi-natural experiment using the staggered/gradual adoption of ESOPs since 2014 with DID-style comparisons, heterogeneity tests, and event-study dynamics. Chapter 4: Panel regressions of firm-level AI adoption on subsequent labor investment efficiency with lead/lag tests, two-stage least squares (2SLS) using instrumental variables, and high-dimensional fixed effects to control for time-invariant and some time-varying confounders. GeneralizabilityRestricted to Chinese A-share listed firms — excludes unlisted, small, and informal firms, Findings may not generalize to countries with different ownership structures, labor market institutions, or corporate governance regimes, AI adoption measurement likely relies on firm disclosures/keywords and may misclassify adoption intensity or type, ESOP design and implementation in China differ from Western contexts, limiting transferability, Potential time-period specificity (2010s China) — results may change as AI and financial markets evolve

Claims (12)

Claim	Direction	Outcome	Confidence & Evidence	Details
Stock market liberalization via the Mainland–Hong Kong Stock Connect enhances firms' labor investment efficiency by correcting both under-investment and over-investment in labor. Organizational Efficiency	positive	labor investment efficiency (correction of under- and over-investment)	Reading fidelity high Study strength high	n=1938 0.8
Improved stock liquidity is a key channel through which stock market liberalization improves firms' labor investment efficiency, by enhancing external monitoring. Market Structure	positive	stock liquidity / external monitoring (mechanism)	Reading fidelity high Study strength medium	n=1938 0.48
A more transparent information environment is another channel by which stock market liberalization enhances labor investment efficiency. Governance And Regulation	positive	information transparency / information environment (mechanism)	Reading fidelity high Study strength medium	n=1938 0.48
The positive effect of stock market liberalization on labor investment efficiency is stronger for non-state-owned enterprises (non-SOEs) than for SOEs. Organizational Efficiency	positive	differential change in labor investment efficiency by ownership (non-SOE vs SOE)	Reading fidelity high Study strength medium	n=1938 0.48
Stock market liberalization also fosters more employee-friendly practices, manifested in higher wages and better welfare provision. Wages	positive	wages and employee welfare	Reading fidelity medium Study strength medium	n=1938 0.29
Adoption of employee stock ownership plans (ESOPs) significantly reduces corporate cash holdings. Organizational Efficiency	negative	corporate cash holdings	Reading fidelity high Study strength medium	not reported 0.48
The negative effect of ESOP adoption on cash holdings is stronger among firms with fewer free-rider problems, greater rank-and-file employee participation, and longer employee stockholding durations. Organizational Efficiency	negative	cash holdings (heterogeneous treatment effects)	Reading fidelity medium Study strength medium	not reported 0.29
Firms adopting ESOPs adjust their cash holdings more swiftly toward optimal levels after adoption, and ESOP adoption increases the market value of cash. Firm Revenue	positive	speed of cash-holding adjustment and market value of cash	Reading fidelity medium Study strength medium	not reported 0.29
Firm-level adoption of artificial intelligence (AI) technologies leads to significant improvements in labor investment efficiency in the subsequent year, primarily by reducing labor over-investment. Organizational Efficiency	positive	labor investment efficiency (reduction of over-investment) one year after AI adoption	Reading fidelity high Study strength medium	not reported 0.48
AI-adopting firms achieve better innovation performance, in both the quantity and quality of patents, following AI adoption. Innovation Output	positive	patent quantity and patent quality (innovation performance)	Reading fidelity high Study strength medium	not reported 0.48
Information asymmetry reduction and corporate governance improvements are plausible channels through which AI adoption affects labor investment efficiency and innovation outcomes. Governance And Regulation	positive	information asymmetry / corporate governance (mechanisms)	Reading fidelity medium Study strength medium	not reported 0.29
The effect of AI adoption on labor investment efficiency is moderated by financial constraints, labor skills, and industry competition. Organizational Efficiency	mixed	moderation of AI effect by financial constraints, labor skills, and industry competition	Reading fidelity medium Study strength medium	not reported 0.29