The analysis of involutionary behavior in the East Asian region: a multi-period overlapping generations model approach

Abstract

The phenomenon of “involutionary competition” in the East Asian region has emerged as a critical challenge impacting socioeconomic development. How to change the involutionary behavior effectively while enhancing well-being and societal competitiveness constitutes a critical issue for regional sustainable development. This study develops a multi-period overlapping generations (OLG) model incorporating three key variables such as income distribution systems, social security mechanisms, and property rights regimes. Under adaptive expectations, we analyze intergenerational strategic choices between “involutionary” (high-effort) and “lying flat” (low-effort) behaviors. The findings of our research reveal that insufficient distributive equity and uncertainty in effort-reward returns constitute primary drivers of involution, whereas generous welfare provisions may induce lying flat behavior through risk reduction. Through the simulations of China, Japan, and South Korea, we examine how to influence intergenerational equilibrium strategies by institutional designs including particularly income tax policies, pension replacement rates and property rights clarity. The policy recommendations are given for optimizing institutional frameworks to address the involution-lying flat dilemma, providing theoretical insights for human capital revitalization and sustainable development in aging societies.

Introduction

“Involution” refers to a state in resource-constrained systems where individuals engage in irrational competition for the advantageous position, resulting in an imbalance between labor investment and returns. In recent years, the phenomenon of “involution” in the East Asian region has become increasingly prominent, attracting widespread attentions from both society and academia. As reported by China’s National Bureau of Statistics, the average weekly working hours of enterprise employees have continued to rise since 2015. By 2023, the average weekly working hours reached as many as 49 h, with the growth rate accelerating annually. From 2015 to 2020, the increase in weekly working hours was 1.5 h, while the increase expanded to 2 h in every week from 2020 to 2023. When calculated using a standard 52-week working year, this corresponds to a cumulative increase of 13 working days per individual from 2020 to 2023 (National Bureau of Statistics, 2024). According to OECD statistics, South Korea has the world’s highest proportion (34%) of individuals aged 65 years or older who continue to work, while Japan’s average weekly working hours stand at 45 h (People’s Daily Online, 2021).

The core of “involution” lies in the growth without development and systemic self-reinforcement, accompanied by declining labor productivity and rigid social structures. Severe path dependence on traditional systems hinders the development and breakthrough of new development model. The scarcity of upward social mobility opportunities forces individuals to compete for limited resources through “excessive investment”, exemplified by societal anxiety triggered by policies like China’s “Secondary school entrance examination streaming”(Eurostat, 2023). In labor markets, oversupply of labor leads to a “race to the bottom”, where firms prioritize extending working hours over improving efficiency. This creates a “prisoner’s dilemma” where individually rational choices (overtime work) result in collectively irrational outcomes (chronic overwork). This dynamic stems from institutional deficiencies and unconscious collusion among individuals. Under dual pressures of limited resources and growing societal demands, socioeconomic development risks falling into a “high-input but low-innovation” trap.

To get rid of the “involution” dilemma, the Report on the Work of the Chinese Government proposed a comprehensive institutional reform package, which included establishing foundational market rules, eliminating regional protectionism and market fragmentation, addressing structural inefficiencies in factor allocation and market access mechanisms, and implementing regulatory frameworks to mitigate unproductive competitive practices that fuel involutionary cycles (The State Council. China’s Government Work Report, 2025). There was an urgent need to alleviate pressures from “internal competition” to “innovation-driven expansion” through reforms in education, employment, and labor rights systems with technological innovation, industrial upgrading, and institutional optimization.

Based on these, this study constructs a multi-period overlapping generations model (OLG) which incorporates three key institutional variables, including distribution systems, social security systems, and property rights regimes to analyze individuals’ strategic choices between “involution” (high-effort) and “lying flat” (low-effort) behaviors. By situating “involution” and “lying flat”, we try to explore the intergenerational transmission mechanisms and the linkages between intergenerational resource allocation, institutional design and long-term systemic contradictions. And China, Japan, and South Korea in the East Asian region were chosen as the key countries to provide empirical support for “de-involution” strategies, facilitating a societal shift from “exhaustive competition” to a new paradigm of “innovative collaboration.”

Innovations of this study lie in three aspects. First, we embed the sociological concepts of involution and lying flat into a multi-period OLG framework, offering a novel institutional economics perspective for addressing involution. Second, this study endogenizes income taxation, pension replacement, and property-rights clarity as institutional parameters that shape intergenerational strategic choices to fully explain how distributional fairness, social security design, and institutional transparency drive the transition between involutionary and lying-flat equilibrium. Third, through the simulations for China, Japan, and South Korea, the study demonstrates the distinct institutional architectures generate divergent between welfare outcomes and dynamic transition paths, formulating empirically validated approaches to mitigate involution and enhance sustainable development in East Asia.

The paper is structured as follows: Section II is the relevant literature review; Section III, the multi-period OLG model is built to analyze utility functions and steady states; Section IV, we conduct the analysis of China, Japan, and South Korea under with the multi-period OLG model; Finally, some policy recommendations are given.

The related literature review

Clifford Geertz introduced the concept of “Agricultural Involution” in 1963 through the study of Javanese rice agriculture in Indonesia. Under acute population pressure, agricultural production maintained output through intensified labor inputs rather than technological innovation, contributing to declining labor productivity and rigid social structures (Geertz, 1963). “Involution” describes the stagnation in social development with institutional ossification as its core mechanism (Ma et al. (2025)). In China’s industrialization paradigm, involutionary competition manifests as industrial homogenization and relative overcapacity, encompassing new dimensions of technological, market and institutional competition (Liu, Qin (2025)). Given the distinct characteristics in China’s labor market, individuals with competitive pressures adopt two opposing strategies, including “involution” and “lying flat” (Liu, Wang (2024)). Involution creates an “excessive yet inefficient competition” where individual gains stagnate despite intensified rivalry (Lin et al. (2024)).

The “lying flat” originated as internet slang, reflects a phenomenon the youth in high-pressure environments reduce material and spiritual aspirations and retreat into personal spaces as a response to exhausting work-life competition (Wang et al., 2024). Social stratification narrows upward mobility channels, particularly affecting disadvantaged youth confronting insurmountable class barriers. This has sparked widespread debate, especially regarding “lying flat” tendencies among university students (Peng, Yu (2023)). In contrast, developed Western nations with robust redistribution systems better control wealth inequality, allowing individuals to prioritize personal interests and reduce work commitments (Tian, 2022).

During economic transitions, employment restructuring intensifies competition in traditional industries while emerging sectors fail to absorb surplus labor, exacerbating youth employment pressures. Some adopt “downward-compatible” career choices, fueling involution (Wang, Pan (2024)). Under pluralistic value systems, skepticism toward conventional success metrics, combined with absent alternative frameworks drives individuals toward involutionary participation or disengaged “lying flat” (Zhang et al., 2024). Chronic involution environments induce psychological distress, while prolonged “lying flat” risks skill atrophy and social maladjustment (Tang, Liu (2024)). Excessive involution misallocates resources to low-efficiency competition, whereas widespread “lying flat” reduces labor supply and participation rates, threatening economic vitality amid aging populations (Li, 2025).

Scholars analyzed the involution-lying flat nexus as dual extreme responses to social competition. Perceived involution intensity positively correlates with “lying flat” tendencies among working youth (Su et al. (2025)). Individuals engage in involution when expecting competitive advantages from effort but opt for “lying flat” when perceiving insurmountable barriers. For instance, students initially participate in educational involution but disengage when recognizing futile competition, and form a pattern rooted in meritocratic systems that paradoxically reinforce disengagement (Sun, Xiong (2023)). From an intertemporal decision-making perspective, involution entails sacrificing current consumption for future income through excessive effort, whereas “lying flat” prioritizes present leisure. Individuals choose strategies and make decisions by balancing expectations about future income, social security, and personal preferences (Yang, Mu (2024)).

The involution-lying flat dichotomy reflects intergenerational equilibrium shaped by accumulated contradictions. Paul Samuelson’s foundational OLG model revolutionized intergenerational resource allocation analysis (Samuelson,1958). Peter Diamond’s extension incorporated production sectors, enabling macroeconomic examination of growth and capital accumulation (Diamond,1965). Unlike other models, OLG frameworks explicitly capture generational resource distribution under demographic shifts, proving instrumental for analyzing social security and population policies. Olson and Knapp expanded OLG applications to non-renewable resource allocation (Olson, Knapp (1997)), while Blanchard and Fisher analyzed government-led resource distribution under population aging and technological change (Blanchard, Fisher (1989)).

Martin Feldstein applied the OLG model to compare pay-as-you-go and fully-funded pension systems (Feldstein,1974). The former transfers contributions from workers to retirees, contingent on population growth and interest rates, whereas the latter links pensions to capital markets through investments. Acemoglu & Pischke integrated on-the-job training into OLG frameworks, revealing that youth-dominated labor markets incentive skill development and enhance productivity (Acemoglu, Pischke (1999)). Samuel Huber embedded OLG structures within monetary search models, identifying welfare-enhancing effects of moderate inflation (Huber, Kim (2020)). Ying Tian employed OLG models to assess Chinese pension policies, correlating high savings rates with rapid economic growth under demographic dividends (Tian, 2024). Recent studies incorporate involution and “lying flat” into the OLG framework, where younger generations anticipating resource scarcity due to prior generational advantages may disengage, while others intensify competition.

Existing research has highlighted multiple channels through which individually rational behavior can escalate into collectively inefficient outcomes. Classic models of tournament incentives and peer pressure demonstrate how rank-order pay and normative sanctions generate strategic complementarities in effort, leading to inefficiently high equilibrium effort levels (Lazear, Rosen (1981); Kandel, Lazear (1992)). Empirical evidence further supports these mechanisms. Laboratory and field studies show that visibility and high-performing peers significantly increase individual effort (Mas, Moretti (2009)). At the institutional level, market segmentation reduces interregional mobility and fosters positional competition, thereby exacerbating involutionary dynamics (Lai et al. (2022)). The Market Segmentation Index (MSI) provides a widely used proxy for such local protectionism. In parallel, a substantial body of work shows that factor misallocation reduces aggregate productivity in China and other economies corroborating the “high-input, low-innovation” trap (Brandt et al. (2013); Cheng et al. (2025)). Recent cross-country evidence highlights the importance of secure property rights in improving capital utilization and innovation incentives (Huang et al. (2025); Vendrell Herrero et al. (2025)). While these strands of research are insightful, they remain fragmented and rarely integrated into a unified macro-institutional framework. This paper contributes by embedding sociological concepts of involution and lying flat into a multi-period OLG model, endogenizing income taxation, pension design, and property-rights clarity as institutional parameters.

Previous studies provide valuable insights into resource allocation, yet OLG-based economic analyses of involution remain nascent. Bridging sociological, psychological, and economic theories to uncover involution’s socio-psychological mechanisms and economic roots constitutes an underexplored frontier. This study advances OLG frameworks by integrating realistic factors to systematically examine involution’s economic effects. By combining cross-national simulations for China, Japan, and South Korea in the East Asian region with empirical validation of MSI and misallocation effects, we provide a novel institutional economics perspective on the origins of involution and identify policy reforms to mitigate its consequences.

The OLG model

Basic assumptions of the multi-period OLG model

Individual life stages

Individuals undergo three life stages, including youth age ($\mbox{t}$), middle age ($\mbox{t}+1$), and old age ($\mbox{t}+2$). During youth age and middle age, the income is earned through labor, with consumption budgets derived from discounted income streams across all three periods. Savings are from unspent income accumulated interest, while old-age consumption relies on discounted prior earnings, savings returns and pension income. The total consumption across all periods is represented by ${\rm{s}}$, where, st, st+1, st+2 are the consumption in three life stages, respectively.

The population grows at rate ${\rm{n}}$, which obeys ${\rm{Lt}}=(1+{\rm{n}}){\rm{Lt}}-1$, with an initial population ${\rm{L}}0$.

Production function and factor markets

Production obeys the Cobb–Douglas function under government regulation and the function is:

$${{\rm{Y}}}_{{\rm{t}}}={{\rm{K}}}_{{\rm{t}}}^{\alpha }{({{\rm{A}}}_{{\rm{t}}}{{\rm{L}}}_{{\rm{t}}})}^{1-\alpha }$$

(1)

Where,${\mbox{Y}}_{\mbox{t}}$ is the total output; $\mbox{K}$ is the capital stock; α is the capital share of income; $\mbox{A}$ is the total factor productivity (TFP); $\mbox{L}$ is the labor input and $\mbox{t}$ is the time.

The function of the wage rate is:

$${\rm{w}}=(1-\alpha )\frac{{{\rm{Y}}}_{{{\rm{p}}}_{{\rm{t}}}}}{{{\rm{L}}}_{{\rm{t}}}}$$

(2)

Where, $\mbox{w}$ is the real wage, α is the capital share of income;${\mbox{Y}}_{\mbox{t}}$ is the total output; $\mbox{L}$ is the labor input and $\mbox{t}$ is the time.

The function of the capital return rate is:

$${\rm{r}}=\alpha \frac{{{\rm{Y}}}_{{\rm{t}}}}{{{\rm{k}}}_{{\rm{t}}}}$$

(3)

Where, $\mbox{r}$ is the return on capital; α is the capital share of income; ${\mbox{Y}}_{\mbox{t}}$ is the total output; $\mbox{K}$ is the capital stock and $\mbox{t}$ is the time.

In the Cobb–Douglas production framework, the real wage is equal to the marginal product of labor, which can be expressed as w = (1 − α) $\frac{{{\mbox{Y}}_{\mbox{p}}}_{\mbox{t}}}{{\mbox{L}}_{\mbox{t}}}$, indicating that workers are compensated according to the labor share of output.

Effort level

Effort e ∈ $({\mbox{e}}_{\mbox{H}}{,\mbox{e}}_{\mbox{L}})$ is the intensity of labor input chosen by individuals in their youth and middle age; where, e∈$\,(0,$ 1; and $0$, 1 represents low, high effort. The decision is based on wage (w_t e_t,i) and effort costs $(\frac{{\rm{\gamma }}}{2}{\mbox{e}}_{\mbox{t},\mbox{i}}^{2})$, with ${\rm{\gamma }}$ > 0 as effort aversion. Effort level is endogenously determined by the first-order condition (${{\rm{\gamma }}{\rm{e}}}_{{\rm{t}},{\rm{i}}}=\frac{(1-\mbox{T}){\mbox{w}}_{\mbox{t}}}{{\mbox{c}}_{\mbox{t}}}$), influenced by after-tax wages, consumption and institutional design.

Individual assets

Total lifetime capital equals cumulative savings across periods. The function of the people with youth age is:

$${{\mbox{K}}_{\mbox{life}}}_{\mbox{t}}=(1-\mbox{T}){\mbox{w}}_{\mbox{t}}\,{\mbox{e}}_{\mbox{t},\mbox{i}}-{\mbox{c}}_{\mbox{t},\mbox{i}}$$

(4)

Where, ${{\mbox{K}}_{\mbox{life}}}_{\mbox{t}}$ is the lifetime capital of the people with youth age; $\mbox{T}$ is the individual income tax,${\mbox{w}}_{\mbox{t}}\,{\mbox{e}}_{\mbox{t},\mbox{i}}$ are the labor income during the youth period;${\mbox{c}}_{\mbox{t},\mbox{i}}$ is the consumption.

The function of the people with middle age is:

$${{\mbox{K}}_{\mbox{life}}}_{\mbox{t}+1}=(1+{\mbox{r}}_{\mbox{t}}){\mbox{s}}_{\mbox{t},1}+(1-\mbox{T}){\mbox{w}}_{\mbox{t}+1}\,{\mbox{e}}_{\mbox{t}+1,\mbox{i}}-{\mbox{c}}_{\mbox{t}+1,\mbox{i}}$$

(5)

Where, ${{\mbox{K}}_{\mbox{life}}}_{\mbox{t}+1}$ is the lifetime capital of the people middle age;$(1+{\mbox{r}}_{\mbox{t}}){\mbox{s}}_{\mbox{t},1}$ is the savings of the first phase; $\mbox{T}$ is the individual income tax;${\mbox{w}}_{\mbox{t}+1}\,{\mbox{e}}_{\mbox{t}+1,\mbox{i}}$ is the savings of middle age;${\mbox{c}}_{\mbox{t}+1,\mbox{i}}$ is the consumption of middle age.

The function of people with old age is:

$${{\rm{K}}}_{{{\rm{life}}}_{{\rm{t}}+2}}=(1+{{\rm{r}}}_{{\rm{t}}}){{\rm{s}}}_{{\rm{t}}+1,1}+\tau {{\rm{w}}}_{t+2}{\rm{e}}_{t+2}-{{\rm{c}}}_{{\rm{t}}+2,{\rm{i}}}$$

(6)

Where, ${{\mbox{K}}_{\mbox{life}}}_{\mbox{t}+2}$ is the lifetime capital of the people with old age;$(1+{\mbox{r}}_{\mbox{t}}){\mbox{s}}_{\mbox{t}+\mathrm{1,1}}$ is the savings of old age; τ is the pension, ${\mbox{w}}_{\mbox{t}+2}$ the future wage rate; ${\mbox{e}}_{\mbox{t}+2}$ is the degree of effort of future youths; ${\mbox{c}}_{\mbox{t}+2,\mbox{i}}$ is the consumption of old age.

$${\rm{Total}}\,{\rm{lifetime}}\,{\rm{capital}}\,{\rm{of}}\,{\rm{all}}\,{{\rm{lifetime}}:{\rm{K}}}_{{\rm{life}}}={{\rm{K}}}_{{{\rm{life}}}_{{\rm{t}}}}+{{\rm{K}}}_{{{\rm{life}}}_{{\rm{t}}+1}}+\delta {y}_{t+2}$$

(7)

Institutional design

Three institutional dimensions are analyzed, including the distribution system, social security, and property rights.

①Distribution system. Income tax rates $\mbox{T}$ ∈ $(\mathrm{0,1})$ capture redistributive fairness. Inequitable taxation $({\rm{\tau }}\to 0)$ triggers “effort inflation” and involutionary traps.

②Social security. Pension replacement rate τ measures post-retirement income security.

③Property rights. The property rights clarity index is a parameter measuring the clarity of property rights institutions, where the coefficient κ ∈ [0,1] modulates capital returns. Weak property rights (low φ) may lead to “involution” by diluting per capita returns; and the function is:

$${{\rm{r}}}_{{\rm{t}}}=\varphi \cdot \alpha {{\rm{K}}}_{{\rm{t}}}{{\rm{Y}}}_{{\rm{t}}}$$

(8)

Where, $\mbox{r}$ is the return on capital; φ is the property right clarity factor which captures the degree of property-rights clarity, the efficiency of financial markets, or the effectiveness of capital utilization; α is the capital share of income; ${\mbox{K}}_{\mbox{t}}$ is the capital stock;${\mbox{Y}}_{\mbox{t}}$ is the total output.

We define the capital return as ${{\rm{r}}}_{{\rm{t}}}=\varphi \cdot \alpha {{\rm{K}}}_{{\rm{t}}}{{\rm{Y}}}_{{\rm{t}}}.$ φ ∈ (0.1) captures institutional and efficiency distortions, such as property-rights clarity, financial frictions and so on. When φ-1, the model collapses to the standard Cobb–Douglas marginal product of capital.

Adaptive expectations

Under adaptive expectations, the expectation adjustment coefficient measures the speed at which individuals adjust future expectations based on past experiences.

Agents form wage expectations via backward-looking adjustments and the function is:

$$\mbox{Ze}({\rm{t}}+1\,)={\rm{Ze}}({\rm{t}})+{\rm{\gamma }}({\rm{Z}}({\rm{t}})-{\rm{Ze}}({\rm{t}}))\,$$

(9)

Where, $\mbox{Ze}(\mbox{t}+1)$ is the wage rate after tax, $\mbox{Ze}(\mbox{t}$) is the expectation of the after-tax wage rate in this period based on the previous period, $\mbox{Z}(\mbox{t})$ is the actual after-tax wage rate in this period, ${\rm{\gamma }}$ is the adjustment coefficient, ${\rm{\gamma }}$ ∈ (0, 1). For involutionary agents (λ < 0), pessimism dominates; “lying flat” agents (λ ≈ 1) exhibit inertia.

Utility function

Lifetime utility combines consumption and effort disutility and the function is:

$$\mbox{U}={\mathrm{ln}\mbox{U}}_{\mbox{t},1}+{\beta \mathrm{ln}\mbox{U}}_{\mbox{t},2}+\beta {\mathrm{ln}\mbox{U}}_{\mbox{t},3}-\frac{{\rm{\gamma }}}{2}\left({\mbox{e}}_{\mbox{t},\mbox{i}}^{2}+{\mbox{e}}_{\mbox{t},\mbox{i}+1}^{2}+{\mbox{e}}_{\mbox{t},\mbox{i}+2}^{2}\right)$$

(11)

First-order conditions yield:

$$\frac{\partial {\rm{U}}}{\partial {\rm{e}}_{{\rm{t}},{\rm{i}}}^{2}}=\beta {{\rm{lnU}}}_{{\rm{t}},2}+\beta {{\rm{lnU}}}_{{\rm{t}},3}-\frac{\gamma }{2}({\rm{e}}_{{\rm{t}},{\rm{i}}}^{2}+{\rm{e}}_{{\rm{t}},{\rm{i}}+1}^{2}+{\rm{e}}_{{\rm{t}},{\rm{i}}+2}^{2})$$

(12)

When choosing “lying Flat”, the function is:

$$\mbox{U}={{lnc}}_{t}+\beta {{lnc}}_{t+1}+{\beta }^{2}{{lnc}}_{t+2}-\frac{{\rm{\gamma }}}{2}{e}_{l}^{2}\left(1+\beta \right)$$

(13)

When choosing “involution”, the function is:

$$\mbox{U}={{lnc}}_{t}+\beta {{lnc}}_{t+1}+{\beta }^{2}{{lnc}}_{t+2}-\frac{{\rm{\gamma }}}{2}{e}_{H}^{2}\left(1+\beta \right)$$

(14)

Where,$\mbox{U}$ is the lifetime utility combines consumption and effort disutility; ${\mbox{U}}_{\mbox{t},1},{\mbox{U}}_{\mbox{t},2},{\mbox{U}}_{\mbox{t},3}$ is the lifetime utility in different ages, respectively; ct, ct+1, ct+2 is the consumption in the different period, respectively;β is the time discount factor; ${\rm{\gamma }}$ is the effort aversion coefficient; e_H, e₁ are the effort levels under “involution” and “lying flat, respectively.

Steady state conditions

The function of capital accumulation is:

$${\mbox{k}}_{\mbox{t}+1}=\frac{\mbox{st}}{(1+\mbox{n}){\mbox{A}}_{\mbox{t}+1}{\mbox{l}}_{\mbox{t}}+1}$$

(15)

Where,${\mbox{k}}_{\mbox{t}+1}$ is the capital accumulation of the next stage;$\mbox{st}$ is the savings per young individual;$\mbox{n}$ is the population growth rate; l is the individual labor input.

The savings during the youth period are transformed into capital in the next period. Capital accumulation is determined by the population growth rate and technological progress.

The function of the wage-capital relation is:

$${\rm{w}}=(1-{\rm{\alpha }}){\rm{A}}{{\rm{k}}}_{{\rm{t}}}^{{\rm{\alpha }}}$$

(16)

The wage rate is determined by the marginal output of the production function, and ${\rm{k}}=\frac{{\rm{K}}}{{\rm{AeL}}}$ is the effective capital per capita.

The function of capital return is:

$${{\rm{r}}}_{{\rm{t}}}={\rm{\varphi }}{\rm{\alpha }}{{\rm{k}}}_{{\rm{t}}}^{{\rm{\alpha }}-1}$$

(17)

The rate of return on capital is regulated by the property rights coefficient (φ), and it is the capital share (α).

The function of effort is optimal:

$${\rm{\gamma }}{{\rm{e}}}_{{\rm{t}}}=\frac{{{\rm{\delta }}}^{{\rm{t}}}(1-{\rm{T}}){{\rm{w}}}_{{\rm{t}}}}{{{\rm{c}}}_{{\rm{t}}}},{\rm{t}}={\rm{y}},{\rm{m}}$$

(18)

At young age, t = y, δy = 1; while at an old age, t = m, δm = δ.

The function of intertemporal euler is:

$${{\rm{c}}}_{{\rm{t}}+1}={\rm{\delta }}(1+{{\rm{r}}}_{{\rm{t}}}){{\rm{c}}}_{{\rm{t}}},{\rm{t}}={\rm{y}},{\rm{m}}$$

(19)

The growth of consumption is driven by the discount factor (δ) and the rate of return on capital (r).

The function of pension balance is:

$${\rm{\tau }}\mathop{\sum}\limits _{\mbox{t}=\mbox{y},\mbox{m}}{\mbox{w}}_{\mbox{t}}{\mbox{e}}_{\mbox{t}}{\mbox{L}}_{\mbox{t}}=\mbox{T}\mathop{\sum}\limits _{\mbox{t}=\mbox{y},\mbox{m}}{\mbox{w}}_{\mbox{t}}{\mbox{e}}_{\mbox{t}}{\mbox{L}}_{\mbox{t}}$$

(20)

The expenditure on pensions (τ) is raised through taxes(${\rm{T}}$), and intergenerational transfer should cover the entire labor stage(${\rm{t}}={\rm{y}},{\rm{m}}$).

Simulated results

Basic analysis of the model

Under adaptive expectations, the comparative analysis reveals that the dynamic adjustment of individual expectation formation mechanisms alters both the transitional paths and steady-state outcomes of economic variables.

the changes in wage rates and tax rates

Under adaptive expectations, individuals update their future expectations based on historical wage patterns. When the actual after-tax wage rate $Z(t)$ changes in the current period, it influences the expected after-tax wage rate ${Ze}\left(t+1\right)$ through the adaptive expectation formula ${Ze}(t+1)={Ze}(t)+\gamma (Z(t)-{Ze}(t))$.

In an involutionary environment characterized by high prior work intensity but limited wage growth, the adjustment coefficient ${\rm{\gamma }}$ ∈ (0,1) may initially be small or even negative. If the coefficient becomes positive, an increase in the actual after-tax wage rate would raise individuals’ expectations of future after-tax wages ${Ze}(t+1)$, though the magnitude of this adjustment would lag behind the actual wage increase.

According to individual optimization conditions $\frac{\partial {\rm{U}}}{\partial {\rm{e}}}=\frac{{\rm{\beta }}(1-{\rm{T}}){\rm{w}}\,(1+{\rm{r}})}{{{\rm{c}}}_{{\rm{t}}+1}}({\rm{w}}+{\rm{e}}\frac{\partial {\rm{w}}}{\partial {\rm{e}}})-{\rm{\gamma }}{\rm{e}}=0$, higher expected wages incentive individuals to increase labor input and effort levels($e$). Elevated effort enhances labor productivity, thereby increasing output ($Y$) in the production function. Consequently, wage rates and capital returns rise. Higher wages further augment individual income, affecting consumption and savings decisions, which accelerates capital accumulation.

Simultaneously, the decline in the actual after-tax income rate may elevate labor income in youth and middle age and boost the savings across periods. This leads to greater capital accumulation, directly expanding effective labor supply.

Improvement in property rights clarity

Under adaptive expectations, the individuals gradually adjust their expectations based on historical experiences. Enhanced property rights clarity directly elevates the capital return and triggers cascading effects through savings and capital accumulation.

Increased property rights clarity raises the capital return (${rt}=\phi \cdot \alpha \frac{{K}_{t}}{{Y}_{t}}$), incentivizing firms and households to amplify savings, thereby driving capital accumulation. Higher savings translate into augmented capital stock in subsequent periods, which enhances labor productivity and elevates wage rates. Observing rising wages, individuals incrementally revise their income expectations (${\rm{Ze}}({\rm{t}}+1)={\rm{Ze}}({\rm{t}})+{\rm{\gamma }}({\rm{Z}}({\rm{t}})-{\rm{Ze}}({\rm{t}}))$). This gradual adjustment process steers the economy toward a new steady state with higher output. In societies dominated by adaptive expectations, strengthening property rights clarity significantly promotes long-term growth.

Impact of effort aversion coefficient variation

The effort aversion coefficient (${\rm{\gamma }}$) determines the speed at which individuals adjust their expectations in response to discrepancies between actual and expected after-tax wage rates. When ${\rm{\gamma }}$ increases, individuals rely less on historical experience and become more sensitive to current wage changes. Assuming ${\rm{\gamma }}$ remains a fixed positive value, a higher ${\rm{\gamma }}$ implies larger adjustment magnitudes, enabling individuals to update expectations more rapidly.

The empirical analysis of East Asia

To fully understand the developmental disparities in different countries and regions, this study conducts simulation analyses under a multi-period OLG framework, using China, Japan, and South Korea as case studies. By incorporating empirical economic data and model parameter configurations from these nations, the research investigates labor supply, production, consumption-saving behavior and economic growth to analyze the impacts of “involution” and “lying flat” scenarios on economic outcomes.

Baseline parameter setting

Table 1 shows the basic parameters, including saving rate, population growth rate, and TFP in China, Japan, and South Korea.

Table 1 The basic parameter value of China, Japan, and South Korea.

It is clearly seen that China exhibits the highest savings rate. This might primarily be driven by precautionary savings under insufficient social security, forcing households to maintain high savings. Population growth rate and low TFP growth show negative under dual pressures of aging populations and technological stagnation. And South Korea sustains moderate savings with higher TFP growth, reflecting strong middle-class savings incentives. Other key parameters, including population growth rate, capital share disparities, pension replacement rates, and TFP are defined.

Data sources are based on the World Bank’s World Development, People’s Bank of China’s Financial Statistical Yearbook. Japanese Ministry of Finance’s Flow of Funds Statistics, Bank of Korea’s National Accounts, China National Bureau of Statistics’ Statistical Yearbook, Japanese Ministry of Internal Affairs’ Population Estimates, Statistics Korea’s Future Population Projections, OECD’s Productivity Growth Database, Annual Pension Market Report, China’s Social Insurance Development Annual Report, Japanese Ministry of Health, Labor and Welfare’s Pension White Paper, and Korea’s National Pension Service Operation Report.

Steady-state outcome and comparisons

The welfare loss rate is defined as the consumption equivalent variation relative to the benchmark steady state, which measures the percentage increase in lifetime consumption. We continue to get a steady-state outcome, as shown in Table 2.

Table 2 The Steady-state outcome among China, Japan, and South Korea.

From the perspective of capital accumulation, China has excessive capital accumulation. The high capital stock reflects the continuation of the “investment-driven” growth path, but there are significant structural distortions. The savings rate far exceeds the average level of emerging markets. In Japan, capital deepening is hindered. The main reasons may be that the low-interest-rate environment inhibits capital deepening, zombie enterprises drag it down, and deflation expectations suppress investment. These factors have jointly contributed to what is known as Japan’s “lost three decades”, which has further exacerbated institutional rigidity. In South Korea, capital accumulation is restricted by the market structure and external geopolitical risks.

Considering of labor supply, China has a high labor participation rate, but the inflated labor participation rate conceals structural unemployment. The youth unemployment rate is high, and the actual effective labor supply is not ideal. At the same time, the “996 work system” is unsustainable. Constrained by the household registration system, the urbanization of 286 million migrant workers lags behind. The replacement rate of their endowment insurance is 10% lower than that of urban employees, reducing the quality of labor force reproduction. In Japan, it reflects that super-aging has reshaped the labor market. The increase in the number of elderly people in employment and the rise in the female employment rate have partially offset the impact of aging. In South Korea, it may mainly be due to the dual imbalance of generations and genders in the labor market.

Meanwhile, China’s per capita output is much higher than that of Japan and South Korea. The reason relies more on China’s scale effect, technological mergers and acquisitions, and scale effect. The per capita capital stock far exceeds the golden rule level. Japan and South Korea have relatively insufficient innovation capabilities and output advantages, with innovation bottlenecks. At the same time, Japan’s quantitative easing has pushed down interest rates, leading to the diminishing marginal return of capital. South Korea’s export dependence causes its interest rates to be affected by the spillover effects of the Federal Reserve’s policies.

Policy sensitivity test

By calibrating key parameters of the economic characteristics of various countries (such as China’s high savings rate, Japan’s aging population and South Korea’s chaebol structure), and relying on authoritative data (the World Bank, national statistical bureaus) and the theoretical framework, sensitivity calculations are carried out by reducing the income tax rate by 5%, increasing the pension replacement rate by 10%, and increasing the clarity of property rights by 20%, respectively. The results are shown in Table 3.

Table 3 Policy sensitivity test.

After reducing the income tax rate by 5%, the capital stocks and total outputs of China, Japan, and South Korea have significantly increased. All three countries are facing the slowdown of economic growth, population aging and the pressure of international tax competition. Tax cuts can effectively stimulate domestic demand or investment. Since China mainly relies on indirect taxes, the space for tax cuts is concentrated on enterprises. Japan depends on direct taxes, and the effect of tax cuts is not effective.

After increasing the pension replacement rate by 10%, the capital stocks and total outputs of the three countries have significantly increased. As China, Japan, and South Korea are all facing the intensification of aging (the proportion of people aged 65 and above exceeds 14%), the decline in the savings rate (China—7%, Japan—4%, South Korea—5%), and financial pressure (the pension gap has expanded by 2.1–3.8% of GDP). The difference is that in China, it is mainly due to the urban-rural dual division that leads to policy leakage. The coverage rate of migrant workers has only increased by 5%, and local governments rely on land finance to fill the gap (38% of land transfer fees are used for social security).

When the clarity of property rights is improved, the capital stocks and total outputs of the three countries have significantly increased, but there are significant differences in policy tools and constraints. The reasons may lie in the dominance of state-owned enterprises in China, relationship-based capitalism in Japan, and the monopoly of chaebols in South Korea.

From the perspective of the changes in the real interest rate and the welfare loss rate, the real interest rate is determined by the marginal product of capital (MPK). Policy shocks affect the interest rate by changing the capital stock, and the change in the welfare loss rate is affected by consumption smoothness, capital allocation efficiency, and intergenerational equity. Both the reduction of the income tax rate and the improvement of the clarity of property rights will reduce the welfare loss rate, while the increase in the pension replacement rate will increase the welfare loss rate.

Beyond the three main policy shocks, we also conducted robustness checks by varying key parameters within empirically plausible ranges: reducing the savings rate by 5% to capture precautionary savings adjustments under demographic transition; lowering the capital share by 0.05 to reflect sectoral shifts from industry to services; and increasing TFP growth by 0.5% to simulate technology diffusion. The results confirm that the direction of our main findings remains unchanged, and only the magnitude of welfare impact varies. This strengthens the robustness of our conclusions.

Dynamic transition path

The convergence stage is the process of an economy transitioning from its current state to a steady state. The speed of capital accumulation determines the speed at which the economy approaches the steady state. A higher investment rate accelerates capital accumulation and shortens the convergence time. We further analyze the differences of capital accumulation in China, Japan, and South Korea in the dynamic transition path, especially their different performances in the convergence stage.

Table 4 shows that China needs 8.2 years to reach the 50% path, Japan needs 12.7 years, and South Korea needs 9.8 years; while on the 90% path, China needs 21.5 years, Japan needs 34.2 years, and South Korea needs 25.6 years. The difference in China’s convergence mechanism lies in that high investment shortens the adjustment cycle, while Japan’s aging population delays capital formation. China’s high investment may come from government-led infrastructure construction, manufacturing investment, and a relatively high savings rate. Due to the severe aging problem in Japan, the labor force has decreased, and the savings rate may decline, resulting in a slower speed of capital accumulation. South Korea’s investment may be between the two. It has relatively high corporate investment, but it may also face structural problems, such as the influence of the chaebol economy.

Table 4 Dynamic transition path of the capital accumulation speed.

Empirical validation of institutional linkages

While the simulation results provide theoretical insights into how institutional parameters shape “involutionary” and “lying-flat” equilibria, it is equally important to validate these mechanisms with real-world evidence. To this end, we further analysis with empirical indicators that directly capture the diagnosed institutional deficiencies. Specifically, we examine the relationship between local protectionism and excessive competition using the MSI, and investigate the link between factor misallocation and the “high-input, low-innovation” trap through TFP growth.

Table 5 provides a comparison of the institutional linkages across China, Japan, and South Korea. In all three countries, higher market segmentation (MSI) is associated with longer working hours, confirming the involutionary effect of local protectionism. Similarly, greater factor misallocation is consistently linked to lower TFP growth, with more pronounced in China and South Korea. The negative impact of misallocation on productivity is amplified under conditions of higher market segmentation, reinforcing the joint mechanism identified in our theoretical framework. The results demonstrate that local protectionism systematically aggravates excessive competition, while factor misallocation suppresses productivity growth, and their interaction creates a self-reinforcing “involution trap”.

Table 5 Regression results on Market Segmentation and TFP.

Conclusions and countermeasure suggestions

This paper constructs an OLG model to analyze the individual “involution” behavior and the intergenerational transmission mechanism, and conducts a comparative analysis under adaptive expectations to evaluate the impacts on economic variables, including the income tax rate, the clarity of property rights and the effort aversion coefficient. Furthermore, taking East Asia as the object, this paper simulates policy shocks and improving the clarity of property rights, and makes a comparative analysis of the policy effects differences in different regions.

The research shows that: (1) The fairness of distribution and the mechanism of effort return are the core inducements of “involution”. Under resource constraints, the marginal utility imbalance between individual effort input and income growth leads to “effort inflation”. When the Gini coefficient exceeds the threshold, excessive competition causes the social welfare loss rate to rise by 12–15%, forming a “high input-low return” locking effect. (2) Reducing the income tax rate has an asymmetric effect on the East Asian economy. The simulation shows that reducing the income tax rate by 5% can release the consumption potential of China and South Korea by 3.2% and 2.8%, respectively. However, due to the pressure of “involution” suppressing the elasticity of labor supply, the long-term growth rate of capital accumulation is 1.5% lower than that in the benchmark scenario. Japan is restricted by aging, and the policy effect is further attenuated. (3) The policy dividend brought by the improvement of the clarity of property rights is diluted by “involution”. When the property rights protection index changes by 0.2 units, the theoretical increase in the TFP of China, Japan, and South Korea should be 4.3–6.1%. However, under high competition intensity, the individual’s perception threshold of technology diffusion and policy dividends rises, and the actual increase in labor productivity is less than 60% of the theoretical value, and there is a lag in intergenerational transmission. (4) Our empirical also confirm that local protectionism, measured by the MSI, consistently aggravates excessive competition, while factor misallocation significantly suppresses TFP growth across China, Japan, and South Korea.

The 2025 China’s government work report explicitly emphasizes the need to dismantle local protectionism and market segmentation, remove barriers in market entry and exit, factor allocation, and other bottlenecks hindering economic circulation, and comprehensively address “involution” competition. Correspond to the institutional deficiencies diagnosed in OLG model, some suggestions are proposed as following: (1) Reconstruct the income distribution system to alleviate excessive competition with fairness. The income distribution system should be improved to enhance the fairness of distribution and reduce excessive competition caused by uneven distribution. Reasonably adjust the tax rate structure of personal income tax, increase the tax adjustment intensity for high-income groups, and make individual efforts more consistent with returns. (2) Optimize the social security system. Set up an “automatic stabilizer” that links the pension replacement rate with the labor participation rate. With the social security level improved, welfare dependence can be avoided. A reasonable growth mechanism for the pension replacement rate should be considered to further adjust the pension benefits and prevent high welfare from suppressing the labor enthusiasm of young people. (3) Strengthen property rights protection and policy coordination, improve the clarity of property rights and reduce the dilution of per capita income by “involution competition”. (4) Break down local protectionism and market fragmentation to facilitate the efficient flow of production factors. Resolutely eliminate local protectionism, establish and improve a unified national market and minimize efficiency losses caused by resource misallocation. (5) Implement differentiated regional policies to address the challenges of aging populations and insufficient labor supply elasticity. Design asymmetric policy packages tailored to the development stages and demographic characteristics of different regions. (6) The market competition order and policy credibility evaluation system should be better established to strengthen expectations. The government should further enhance social trust and guide individuals to form reasonable expectations, so as to relieve the pressure of “involution” and create a healthy development environment.