Introduction

Cultural consumption has gained increased importance in both academic and policy discussions (Murtin and Zanobetti, 2024). In China, since the turn of the millennium, the strategic significance of the cultural industry in national socioeconomic development has markedly increased (Keane, 2009); the term cultural and creative industries (CCI) was officially sanctioned at the 15th Communist Party Central Committee, which has since adopted CCIs as a national policy and pillar of the economy. The cultural industry plays a pivotal role in the global economy by generating both cultural and economic value. The demand for cultural consumption constitutes a crucial impetus for the growth of the cultural industry (Richards, 1996), exerting a direct effect on its production and reproduction processes.

In view of the importance of subjective well-being, prior research has examined its determinants, such as environmental pollution (Levinson, 2012), income level (Easterlin, 2010; Kahneman and Deaton, 2010), and household registration status (Jiang et al., 2012). In terms of cultural consumption, participation in cultural activities can positively influence an individual’s emotional, cognitive and social development and can instill a sense of purpose, identity and belonging that are vital for one’s mental and emotional happiness (Grossi, 2014). Research has demonstrated that individuals who take part in cultural activities are more inclined to report elevated levels of life satisfaction, self-esteem and positive emotions than their nonparticipating counterparts are (Brown et al., 2015).

It is widely agreed that the internet can promote the development of the cultural industry (Betzler et al., 2021, Khlystova et al., 2022). Not only musicians and artists but also museums and other cultural organizations have adopted digital tools to interact with customers and audiences, offering services online (e.g., virtual exhibitions, recorded tours, concerts, and lectures) and highlighting the importance of using digital technology as a survival tool (Agostino et al., 2020, Betzler et al., 2021). The COVID-19 pandemic has further highlighted the importance of using digital tools, embedding them into organizational business models, and integrating them to generate synergies (Khlystova et al., 2022). For China, the digital era has brought about new opportunities for the innovative development of the cultural industry. Emerging technologies, which are often represented by artificial intelligence, play a significant role in generating high-quality cultural content, promoting new mechanisms for cultural consumption, innovating cultural dissemination models, and strengthening cultural governance paradigms. Shanghai, which focuses on digital technology, aims to become a leading center for cultural and creative industries (CCIs) worldwide (Arkaraprasertkul, 2019).

The relationships between cultural activities and socioeconomic outcomes have been firmly established in the academic literature (see Table 1). Scholars have studied the impacts of conspicuous consumption expenditures in Australia (Wu, 2020), cultural consumption in Italy (Denti et al., 2022), and arts attendance in the UK (Hand, 2018) on life satisfaction and well-being; however, relatively little research has been conducted on China’s cultural consumption. Currently, in China, cultural consumption already has a relatively solid economic foundation. China recognizes intangible cultural heritage as a rich resource for its regional CCIs (cultural and creative industries) and as a means by which to increase its global soft power (Ma, 2022). The report of the 20th National Congress of the Communist Party of China proposed implementing a strategy driven by major cultural industry projects and systematically planned the strategic goal of building a strong cultural nation by 2035. The cultural industry can meet people’s growing needs for a better life and is an important carrier for building a ‘culturally strong nation.’ Given the disparities present in cultural background and economic development, there may be substantial variations in the relationship between cultural consumption and subjective well-being across different nations. Therefore, it is necessary to conduct research in the context of China.

Table 1 Literature on cultural activities and well-being.
Full size table

The present study makes several contributions. Firstly, it expands the growing literature on cultural activities and socioeconomic outcomes. Research in this domain has investigated the linear relationship between cultural consumption and residents’ subjective well-being (Becchetti et al., 2012; Węziak-Białowolska et al., 2019); only a few studies have explored the nonlinear relationship. Secondly, we establish the threshold for cultural consumption relative to total household expenditure to explore the issue of excessive consumption. Thirdly, our research investigates how cultural consumption impacts subjective well-being using longitudinal data from the China Family Panel Studies, filling a research gap in China where this topic has not been extensively examined (Wu, 2020; Denti et al., 2022; Hand, 2018). Only a few studies have used longitudinal datasets to explore the relationship between the two concepts (Becchetti et al., 2012, Wu, 2020). The present investigation utilizes data from the China Family Panel Studies (CFPS). The CFPS is a nationally representative longitudinal survey that has been conducted biennially since 2010 with the objective of tracking changes in China’s society, economy, population, education, and health through the analysis of samples at the individual, family, and community levels. Consequently, this study investigates the nonlinear effect of cultural consumption on residents’ subjective well-being in China while considering the moderating impact of internet use on this relationship.

Literature review and research hypotheses

Cultural consumption involves spending on cultural activities, such as the consumption of music, visual arts (including activities such as visiting museums, attending art exhibitions, and going to heritage sites), cinema, and theater. Participation in cultural activities and the consumption of cultural products can be referred to simply as “cultural participation” or “cultural consumption” (Murtin and Zanobetti, 2024). Wheatley and Bickerton (2017) explored the relationships between participation in arts, culture and sports and subjective well-being. They believe that cultural goods produce positive leisure experiences and that the frequency of participation is crucial for the positive impact produced by artistic activities and sports activities.

Research increasingly focuses on understanding the contribution of leisure activities to life satisfaction (Sirgy et al., 2017). If leisure-time activities and hobbies are reasonably considered to contribute to happiness, then similar reasoning can be applied to cultural activities. Denti et al. (2022) explored the impact of cultural consumption on hate incidents via data from the Italian region. The results show that cultural consumption is effective in combating hate. Vegheș (2020) suggested that greater cultural participation leads to better quality of life. Bertacchini et al. (2021) reported that there is a positive correlation between cultural participation and overall life satisfaction, which includes health, leisure, friendship relationships, and economic and job conditions. However, the author does not distinguish between subjective well-being and objective well-being. Cultural products can provide utility to consumers (Throsby, 1994), and consumers satisfy their needs through consumption. Cultural consumption can help individuals build knowledge and skills that can lead to personal and professional growth (Kasser, 2004). The positive emotional experience brought about by consumption can be expressed by consumer subjective well-being and reflects a persistent and stable positive emotional experience (Diener, 1984).

Different Chinese generations are experiencing a “U” shaped life satisfaction trend (Zhang et al., 2022). In the context of cultural consumption, Fisher and Specht (1999) reported that elderly individuals who participate in artistic activities experience stronger feelings of purpose and personal growth in life. Building on these findings, this study proposes that cultural consumption among the elderly population in China will lead to stronger subjective well-being than that present among young people. Additionally, Chui and Wong (2016) highlighted gender differences in life satisfaction and happiness, and Zhang et al. (2022) analysed Chinese household data and concluded that women tend to report higher levels of life satisfaction than men do. On the basis of these insights, we suggest that cultural consumption among Chinese women results in stronger subjective well-being than among men of the same age group. On the basis of the previous literature, this study proposes the following:

Hypothesis 1: Cultural consumption contributes to enhancing residents’ subjective well-being.

Hedonic consumption is the primary means through which consumers seek happiness and enjoyment (Alba and Williams, 2013). The utilitarian alternative tends to be chosen over the hedonic alternative when the two are presented jointly (Okada, 2005). The phenomenon of hedonic adaptation suggests that the satisfaction derived from the repeated consumption of cultural goods diminishes rapidly with increasing frequency (Sheldon and Lyubomirsky, 2012). In the context of cultural consumption, overindulgence in cultural entertainment may have negative consequences for consumer happiness, such as financial and psychological burdens (Mrad and Cui, 2020, Olsen et al., 2022). Therefore, this study proposes the following hypothesis:

Hypothesis 2: Excessive cultural consumption may lead to a decrease in subjective well-being. There is an inverted U-shaped relationship between residents’ cultural consumption and their subjective well-being when households experience excessive consumption.

Internet use can be defined in terms of the frequency with which activities are performed (Scheerder et al., 2017). Individuals who use the internet are more likely to engage in cultural participation than nonusers are (Schehl et al., 2019). Prior studies have indicated a positive correlation between active internet use among older adults in Finland and their engagement in various leisure activities (Näsi et al., 2012). The internet has significantly influenced cultural consumption by transforming how consumers participate in cultural activities and consume cultural products. The COVID-19 pandemic has accelerated the adoption of digital tools by not only musicians and artists but also museums and other cultural organizations to interact with customers and offer online services such as virtual exhibitions, recorded tours, concerts, and lectures (Agostino et al., 2020, Lawton et al., 2022). Digital technologies such as Kindle, Facebook, and WhatsApp have become integral to organizational business models, highlighting their importance in the cultural industry (Khlystova et al., 2022). Research suggests that online experiential activities can improve consumer happiness (Skard et al., 2021). Thus, internet use moderates the relationship between cultural consumption and subjective well-being, with higher internet use frequency being associated with greater subjective well-being derived from cultural consumption. Therefore, this study proposes the following hypothesis:

Hypothesis 3: Cultural consumption contributes to enhancing residents’ subjective well-being, and internet use strengthens this relationship.

Data, model, and descriptive statistics

Data source

The data utilized in this study are drawn from the 2010, 2014, and 2018 databases of CFPS. Numerous researchers have employed the CFPS databases to conduct research. For example, Gong and Song (2021) utilized the CFPS2014 database to conduct an empirical analysis of the impact of financial literacy on family cultural consumption. Ju et al. (2022) examined the impact of the planetary boundary layer height (PBLH) on residents’ health. Leng (2022) considered the relationship between the digital revolution and the income of rural households in China. Li and Lu (2024) examined the impact of internet access on enhancing women’s bargaining power within households. The CFPS database comprises family questionnaires and individual questionnaires. After family IDs are matched and missing values of related variables are eliminated, an effective sample size of 47,949 from 31 provinces is obtained. Table 2 summary statistics for sociodemographic background data such as age, gender, and education level.

Table 2 Summary statistics of the sociodemographic background, means (percentage).
Full size table

On the basis of the studies by Cattaneo et al. (2024) and Akcigit et al. (2022), we use the ‘binscatter’ command to visualize the bivariate relationship. By incorporating province fixed effects, we conduct regressions on the relationship between cultural consumption and subjective well-being for the years 2010, 2014, and 2018; the trend changes across these years are displayed in Fig. 1. The results show that in 2010, the relationship between cultural consumption and residents’ well-being was relatively linear; as cultural consumption increased, residents’ well-being also increased at a fixed rate, without complex nonlinear characteristics between the variables. However, in 2014 and 2018, the relationship between cultural consumption and residents’ well-being exhibited an inverted U shape, indicating a nonlinear effect. At a certain level of cultural consumption, residents’ well-being reaches a maximum and then begins to decline.

Fig. 1: Scatter plot and fitted lines between cultural consumption (lnCultcons) and subjective well-being (SWB) for the years 2010, 2014, and 2018.
figure 1

It shows the scatter plot and fitted lines representing the relationship, with data points and lines for 2010 (blue), 2014 (red), and 2018 (green). In 2010, cultural consumption is plotted along the X-axis, while subjective well-being increases from bottom to top along the Y-axis. In 2014 and 2018, the relationship exhibits an inverted U-shape.

Full size image

Model

To investigate the impact of cultural consumption on residents’ subjective well-being across different regions in 2010, 2014, and 2018, we employ a fixed-effects regression model. Our model is specified as follows.

$${{SWB}}_{{it}}={\beta }_{0}+{\beta }_{1}\mathrm{ln}{{Cultcons}}_{{it}}+{\gamma }_{0}{C}_{{it}}+{\alpha }_{i}+{\mu }_{t}+{\delta }_{{it}}+{\varepsilon }_{{it}}$$
(1)

In Eq. (1), \({{SWB}}_{{it}}\) is the dependent variable representing the respondent’s subjective well-being for province i in year t, and \(\mathrm{ln}{{Cultcons}}_{{it}}\) is the key independent variable representing the natural logarithm of cultural consumption. \({\beta }_{0}\) represents the intercept item,\({\beta }_{1}\) is the coefficient of \(\mathrm{ln}{{Cultcons}}_{{it}}\), and \({C}_{{it}}\) is a matrix of control variables, which are related mainly to age, gender, health status, marital status, highest educational attainment, household per capita income, household registration, housing ownership, and nonmortaging financial liabilities. \({\alpha }_{i}\) denotes the fixed effects for province, \({\mu }_{t}\) denotes the fixed effects for year, and \({\delta }_{{st}}\) denotes the interaction fixed effects. \({\varepsilon }_{{it}}\) is the error term. The descriptive statistical results of the variables are shown in Table 3. Due to space limitations, definitions for each indicator are included in Table A1 in the Appendix.

Table 3 Descriptive Statistics for entire population and different demographic segments.
Full size table

Empirical results

Baseline regression results

This study uses Stata 17 software to conduct regression analysis. Table 4 reports the estimation results without including any fixed effects (Column 1) and the province-level fixed effects model (Column 2) without control variables, as well as the province‒year interaction fixed effects estimation results with control variables (Column 3). Introducing province and year interaction fixed effects further enhances the validity of causal inference. These interaction terms capture dynamic changes in the relationship between cultural consumption and subjective well-being across different provinces and years. For instance, cultural consumption may exhibit varying effects in different years due to economic development, policy changes, or social events. The interaction fixed effects help to control for these temporal variations, thereby strengthening the reliability of the study findings.

Table 4 Baseline results.
Full size table

The results show that the coefficient of lnCultcons is 0.0104 and positive and significant at the 1% level after controlling for both province‒year interaction fixed effects and additional control variables. This indicates that a one-unit increase in lnCultcons is associated with a 0.0104 increase in subjective well-being. For the control variables, the personal characteristic variables and family characteristic variables of the respondents affect residents’ subjective well-being to varying degrees. The results indicate that age, health status, marital status, highest educational attainment, household per capita income, and household registration positively and significantly affect subjective well-being, whereas being male and having nonmortaging financial liabilities negatively affect subjective well-being.

Moderating effect

Previous theoretical analysis has shown that with the advent of the digital economy era, digital technology has greatly promoted the development of the cultural industry and enriched the application scenarios of cultural products (Agostino et al., 2020, Lawton et al., 2022). Therefore, this study employs two variables as proxy measures for internet usage, namely, “internet access” according to Li and Lu (2024) and “frequency of internet use” according to Leng (2022). Thus, it explores the moderating role of internet use in the relationship between cultural consumption and residents’ subjective well-being. Eq (2) is listed below:

$$\begin{array}{l}{{SWB}}_{{it}}={\alpha }_{0}+{\alpha }_{1}\mathrm{ln}\,{{Cultcons}}_{{it}}+{\alpha }_{2}{{Internet}}_{{it}}+{\alpha}_{3}{\mathrm{ln}\,{Cultcons}}_{{it}}\times {{Internet}}_{{it}}\\\quad\qquad\quad\;+{\gamma }_{1}{C}_{{it}}+{\alpha }_{i}+{\mu }_{t}+{\delta }_{{it}}+{\varepsilon }_{{it}}\end{array}$$
(2)

Columns 1 and 2 in Table 5 present the linear regression results incorporating measures of internet usage frequency and internet access, respectively. Both models control for province fixed effects, time fixed effects, and their interaction and include a set of control variables. The complete results are detailed in Table A2 in the appendix. In Column 1, the coefficient for the interaction term with internet usage frequency is 0.0014 (p < 0.001). Internet usage frequency is measured as the monthly frequency of social and entertainment activities online. In Column 2, the coefficient for the interaction term with internet access is 0.0312 (p < 0.001). Internet access is represented as a binary variable (1 for using mobile devices or computers to access the internet, 0 otherwise). These findings indicate that the effect of cultural consumption varies significantly depending on the frequency of internet use and whether individuals have internet access. The internet facilitates cultural consumption by enhancing convenience, introducing online consumption avenues such as virtual reality and streaming media, and broadening access to diverse cultural experiences.

Table 5 Moderation effects of internet use.
Full size table

Figure 2 illustrates the marginal effects of lnCultcons on subjective well-being across different levels of internet use. The left panel shows that as the frequency of internet use increases, the predicted subjective well-being shows a slight positive trend, with a noticeable increase in variance at higher frequencies. Conversely, the right panel shows a more pronounced positive relationship between cultural consumption and subjective well-being among individuals who have internet access than among those who do not, as indicated by the steeper slope and narrower confidence intervals.

Fig. 2: Marginal effects plots of cultural consumption by internet use (Frequency of internet use and Internet access).
figure 2

The left panel shows the effects by frequency of internet use, while the right panel demonstrates the effects by internet access. Internet use trends from left to right along the X-axis, while the marginal effects of lnCultcons on subjective well-being rise from bottom to top along the Y-axis.

Full size image

Inverted U‐shaped relationships

To further analyse the nonlinear relationship between cultural consumption and subjective well-being, we use the method proposed by Haans et al. (2016) to test nonlinear relationships (Eq. (3)). The results are shown in Fig. 3. The complete results are detailed in Table A3.

$$\begin{array}{l}{{SWB}}_{{it}}={\omega }_{0}+{\omega }_{1}\mathrm{ln}\,{{Cultcons}}_{{it}}+{\omega }_{2}{\mathrm{ln}{Cultcons}}_{{it}}^{2}+\\\qquad\qquad\quad{\gamma }_{2}{C}_{{it}}+{\alpha }_{i}+{\mu }_{t}+{\delta }_{{it}}+{\varepsilon }_{{it}}\end{array}$$
(3)
Fig. 3: Inverse U-shaped relationship between cultural consumption and subjective well-being.
figure 3

The influence of residents’ cultural consumption on subjective well-being is positive up to a certain point, after which it becomes negative. This indicates the presence of a saturation point.

Full size image

Specifically, the extreme point of \(\mathrm{ln}{Cultcons}\) is located at 5.67, with a 99% Fieller confidence interval of [4.83, 7.78]. The range of values of the independent variable is [0, 10.3]. This finding indicates that subjective well-being tends to increase as cultural consumption increases from low to moderate levels. However, beyond a certain point, further increases in cultural consumption are associated with a decline in subjective well-being. The overall test strongly supports the presence of an inverse U shape (Fig. 3). The influence of residents’ cultural consumption on subjective well-being is positive up to a certain point and then becomes negative; that is, there is a saturation point. After the logarithm of the extreme point is restored, the value is approximately CNY 289.55. After this point, there may be hedonic adaptation, where further increases in expenditure do not lead to corresponding increases in subjective well-being. This can be understood as excessive consumption of cultural activities having a negative impact on consumer welfare and causing financial and psychological burdens (Mrad and Cui, 2020, Olsen et al., 2022).

Based on Hansen’s (2000) research, we employed a threshold regression model to investigate the nonlinear relationship between cultural consumption and subjective well-being in the Chinese context. We defined the percentage of household expenditure spent on cultural consumption as the threshold variable. According to the results, the sample is divided into two regions, and the threshold value is 16%. When the rate is low, cultural consumption has a stable or positively increasing impact on subjective well-being (coefficient: 0.024). Conversely, the high rate (more than 16%) reveals a negative relationship (coefficient: −0.179). Results indicate that in Chinese households, cultural consumption expenditure should be controlled within 16% of total household expenditure to maximize subjective well-being. The complete results are detailed in Table A4.

Endogeneity test

In our model, we opt for instrumental variable estimation. Becchetti et al. (2012) investigate the relationship between social leisure and life satisfaction, which employ two instrumental variables, namely, the probability that individuals as old as the respondent may retire and the probability that individuals one year younger than the respondent may retire. Following this approach, we employ a similar metric as our instrumental variable. Retirement often provides more leisure time, which is correlated with leisure activities; however, it does not directly affect life satisfaction or well-being, thereby meeting the conditions for instrumental variables. Owing to significant missing data for retirement age and retirees in the CFPS, we use external data, the number of retirees in basic pension insurance per province and year, as our instrumental variables. After individuals retire, they begin to receive pension benefits. Retirement generally provides individuals with more leisure time, which is closely associated with participation in leisure activities. This has a positive influence on cultural consumption. However, retirement does not directly impact life satisfaction or overall well-being, which aligns with the conditions for a valid instrumental variable.

We employ two-stage least squares (2SLS) to test for endogeneity. In Column (1) of Table 6, we present the results from the first-stage regression. The instruments are significantly positively correlated with cultural consumption, indicating that higher retirement numbers are associated with increased cultural consumption, as hypothesized. Column (2) reports the second-stage regression results estimated via instrumental variables. We first examine the validity of the instruments. The Cragg–Donald Wald F statistic exceeds 10, suggesting that the instruments are relevant. Thus, the model is identified, confirming the effectiveness of the instruments (p = 0.0004). Consistent with our estimation, we find that cultural consumption significantly enhances residents’ subjective well-being.

Table 6 Regression results of instrumental variable estimation.
Full size table

Robustness results

To ensure the robustness of the results, this study substitutes “satisfaction with one’s own life” and “confidence in one’s own future” for subjective well-being. The results shown in Table 7. Columns 1 and 2 of Table 7 indicate that after the dependent variable is replaced, cultural consumption has a significant positive effect on both life satisfaction and future confidence. We have considered inflation factors and adjusted the values of cultural consumption for regression analysis. The inflation rate is equal to the M2 growth rate minus the GDP growth rate. The coefficient for lnIRCultcons in the regression model is estimated at 0.0107, with statistical significance at the 1% level. This adjustment aims to ensure comparability of cultural consumption across different years, accounting for inflationary effects, as shown in Column 3 of Table 7. The complete results are detailed in Table A5. Combining the two test methods, the test results indicate the robustness of the previous conclusions.

Table 7 Robustness check.
Full size table

Heterogeneity analysis

The purpose of this analysis is to investigate whether the relationship between cultural consumption and subjective well-being varies significantly across different demographic segments.

Age groups

Zhang et al. (2022) categorized age into three groups on the basis of Chinese demographic trends: an elderly group (born in the 1940s and 1950s), a middle-aged group (born in the 1960s and 1970s), and a young group (born in the 1980s and 1990s). Their study identified a “U”-shaped trend in life satisfaction across different Chinese generations. Following this approach, our study categorizes individuals into the same three groups. As shown in Table 8, the results indicate significant differences in the impact of cultural consumption on subjective well-being across the age groups. The complete results are detailed in Table A6. For the youth group, cultural consumption does not significantly nonlinearly affect subjective well-being. The middle-aged group exhibits an inverted U-shaped relationship, suggesting that moderate cultural consumption enhances subjective well-being, while excessive consumption may have negative effects. Conversely, the elderly group shows a notably positive impact, although there is a risk of an inverted U-shaped trend. The Chow test yields a statistic of 33.66 with a P > F (24, 47913) = 0.00, indicating significant differences within the sample.

Table 8 Heterogeneity analysis results by gender and age.
Full size table

Gender

Chui and Wong (2016) highlighted gender differences in happiness and life satisfaction. Zhang et al. (2022) analysed CFPS data and concluded that women tend to report higher levels of life satisfaction than men do. Additionally, significant differences in leisure activities have been observed between elderly men and women (Zhang and Feng, 2024). In this study, individuals are divided into two groups: males and females. The results demonstrate that cultural consumption affects subjective well-being more for women than for men. This finding suggests opportunities to further engage the female consumer segment in cultural consumption. The Chow test yields a statistic of 5.14, with a P > F (10, 47927) = 0.00, indicating significant differences within the sample.

Discussion and conclusions

Building on this conclusion, this paper examines the inverted U-shaped relationship between cultural consumption and subjective well-being across a significant household sample in China from 2010, 2014, and 2018. Our results provide several policy implications. Subjective well-being encompasses positive emotions and the subjective evaluation of life satisfaction (Ahuvia and Friedman, 1998). Cultural consumption can provide residents with enduring and stable positive emotional experiences, highlighting the importance of arts and culture in education (Murtin and Zanobetti, 2024). We recommend initiating public art education campaigns, such as regional art festivals. These initiatives help more individuals develop lasting cultural consumption habits. Additionally, cultural consumption can be integrated with tourism to enrich residents’ cultural lives (Fan and Luo, 2022), promote national culture (Gao et al., 2021), and drive the development of the cultural industry.

Beyond a certain threshold, increased cultural consumption is associated with a decline in subjective well-being, indicating a saturation point. This threshold is estimated at approximately CNY 289.55. This phenomenon aligns with the concept of “hedonic adaptation,” where further increases in expenditure do not correspondingly increase subjective well-being (Sheldon and Lyubomirsky, 2012). The excessive consumption of cultural products and services can negatively impact consumer welfare, leading to financial and psychological burdens (Mrad and Cui, 2020, Olsen et al., 2022). Our findings suggest that the percentage should remain within the 16% to maximize subjective well-being. This finding indicates that at higher proportions of expenditure, increases in cultural consumption may not correspondingly increase subjective well-being. To support this, we propose targeted subsidies and incentives—such as cultural consumption vouchers and ticket subsidies—in regions where cultural consumption falls below optimal levels. In areas where consumption nears or exceeds this threshold, efforts should focus on improving the quality and diversity of cultural offerings to ensure sustainable engagement without encouraging excessive consumption.

Our results support the finding that the effect of cultural consumption varies significantly depending on the frequency of internet use and whether individuals have internet access. Internet use can enhance the relationship between cultural consumption and subjective well-being. With the assistance of the internet, traditional cultural industries are continually optimized, and their scale is expanded (Khlystova et al., 2022). The online consumption scenarios enabled by digital technology make cultural consumption more convenient and offer consumers a broader range of cultural experiences (Agostino et al., 2020, Betzler et al., 2021). Given the transformative role of digital technology in cultural consumption, fostering digital adoption is crucial for China’s cultural industry development. This includes providing funding and policy support to entities undergoing digital transformation (Raimo et al., 2021). Initiatives that combine culture and technology could include promoting short videos, online streaming, and online performances, which leverage the growing demand for digital cultural content and make cultural participation more accessible to a wider audience.

Furthermore, heterogeneity analysis results indicate significant differences across the age groups. The middle-aged group and the elderly group exhibit an inverted U-shaped relationship, and the elderly group shows a notably positive impact. Also, the results demonstrate that cultural consumption affects subjective well-being more for women than for men. The paper proposes promoting cultural consumption by addressing retirement and technological advancements, which provide individuals with more leisure time for various activities (Zhang and Feng, 2024). Thus, encouraging and supporting communities and cultural institutions to offer diverse cultural activities tailored for the elderly is recommended. These activities could encompass art exhibitions, concerts, theatrical performances, book clubs, handicraft classes, etc., with the aim of catering to the diverse interests and needs of older adults. Encourage low-cost or free cultural activities in communities to increase residents’ participation and reduce the negative impact of high spending. Economic support or subsidies for cultural activities, such as free or discounted tickets and transportation subsidies, should be provided to low-income elderly individuals to ensure affordability. Moreover, recognizing that cultural consumption contributes more to the well-being of women than men does suggests promoting awareness among women about the importance of cultural consumption and its positive impact on subjective well-being. This can be achieved through media campaigns, community activities, and other channels of communication.

The paper conducted heterogeneity analyses by age group, dividing the sample into three categories. However, these groups resulted in relatively small sample sizes, which may affect the reliability of the results. This limitation stems from the time span and intervals in the survey database, which constrained the ability to increase the sample size further. In addition, while the instrumental variable approach effectively addresses reverse causality, there is still room for improvement in the selection of instruments. Future research could explore more refined or alternative instruments to enhance the robustness of the results.