Dynamic Causality Between the Housing Boom and Technological Innovation in China: A Sub-Sample Rolling-Window Analysis
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe document analyzes the impact of housing booms (HB) on technological innovation (TI) in China, using advanced causal analysis methods and dynamic models. However, improvements in structure, clarity, and contextualization of the results could significantly enhance the academic and practical impact of the work.
The title, “Is Housing Boom a Blessing or Curse for Technological Innovation?”, could be improved by including references to the geographical (China) and methodological context.
The abstract clearly describes the topic but should be more specific about the methodological tools used (e.g., rolling-window bootstrap causality tests) and the main findings.
The introduction section provides good context but requires a better transition between the discussion of housing booms and technological innovation. The unique contribution of the paper compared to existing literature should be articulated more clearly.
The literature reviewed is comprehensive, but it could be synthesized to reduce length and improve readability. However, recent references linking technological innovation to global housing policies are missing. (citations)
The methodological section is well-structured but could expand on the explanation of the rolling-window bootstrap method, providing examples of previous studies demonstrating its effectiveness. It would be useful to clarify the criteria for selecting control parameters, such as financing constraints, and to discuss the robustness of the results concerning variations in model parameters.
The description of the results is detailed but requires greater emphasis on their practical relevance. Graphs and tables could be complemented with more concise explanations.
The discussion is well-developed but lacks a dedicated section on the study's limitations. For instance, reliance on Chinese data might limit its generalizability.
Some repetitions could be eliminated to improve the cohesion of the text.
Recommended literature:
Resampling techniques for real estate appraisals: Testing the bootstrap approach / Del Giudice, Vincenzo; Salvo, Francesca; De Paola, Pierfrancesco. - In: SUSTAINABILITY. - ISSN 2071-1050. - 10:9(2018), p. 3085. [10.3390/su10093085]
Rogoff, K., & Yang, Y. (2024). CHINA’S REAL ESTATE CHALLENGE.
Atkinson, R. D., & Atkinson, R. D. (2024). China Is Rapidly Becoming A Leading Innovator in Advanced Industries. Information Technology and Innovation Foundation: Washington, DC, USA.
Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34-58.
Li, J., Lyu, P., & Jin, C. (2023). The Impact of Housing Prices on Regional Innovation Capacity: Evidence from China. Sustainability, 15(15), 11868.
Author Response
HB+TI Response: Reply to the Reviewer 1 The document analyzes the impact of housing booms (HB) on technological innovation (TI) in China, using advanced causal analysis methods and dynamic models. However, improvements in structure, clarity, and contextualization of the results could significantly enhance the academic and practical impact of the work. 1. The title, “Is Housing Boom a Blessing or Curse for Technological Innovation?”, could be improved by including references to the geographical (China) and methodological context. Response: Thank you so much for your comments. We have revised the article title, and the new title includes the geographical (China) and methodological context of the study, which is shown as follows : The title: Dynamic Causality Between Housing Boom and Technological Innovation in China: A Sub-Sample Rolling-Window Analysis 2. The abstract clearly describes the topic but should be more specific about the Response: Thank you so much for your comments. We have highlighted methodological tools used and the main findings in abstract Section, which is shown as follows (Pages 1): Abstract: This paper employs bootstrap rolling-window tests to investigate the dynamic causal relationship between housing boom (HB) and technological innovation (TI) in China. Through sub-sample analysis, we reveal a dual impact of HB on TI: during periods dominated by the liquidity effect, HB exerts a positive influence on TI, whereas during periods dominated by the crowding out effect, HB negatively affects TI. Furthermore, the study identifies a significant positive effect of TI on HB, suggesting that TI serves as a predictor of real estate development trends. This research not only provides empirical evidence on the bidirectional interaction between HB and TI, but also offers valuable insights for policymakers in balancing the development of the real estate market and TI. 3. The introduction section provides good context but requires a better transition between the discussion of housing booms and technological innovation. The unique contribution of the paper compared to existing literature should be articulated more clearly. Response: Thank you so much for your comments. To create a smoother connection between the discussion of HB and TI, we have reorganized the content of the first paragraph of the introduction and added a transitional sentence. To more clearly articulate the unique contributions of this paper, we have supplemented the content of the third paragraph of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been acknowledged globally. However, the impact of HB on TI remains a topic of debate, lacking a unified consensus. This is due to its complex and variable nature, which depends on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000, China’s real estate sector has experienced significant growth, largely attributed to the rapid economic development and the housing commercialization policy. At present, China have the biggest real estate market in the whole world (Shi et al., 2018). At the same time, China is also the world’s fastest-growing in terms of housing prices (Wang et al., 2024). High housing prices bring high profits and high return on investment (Yang et al., 2024). According to statistics from the National Bureau of Statistics of China (NBSC), the return on investment in China’s real estate industry is twice as high as the return on investment in manufacturing. The high return not only attracts a significant amount of capital from non-real estate enterprises (Li et al., 2021), but also attracts all kinds of financial capital (Wang et al., 2024). The total investment in real estate increases 30 times, from the initial 0.36 trillion yuan in 1998 to 11.09 trillion yuan in 2023. The participation of a large number of non-real estate enterprises further intensifies HB (Meng et al., 2018). However, the main business and technical expertise of non-real estate companies are not in the real estate sector. Therefore, China's HB, as well as the expansion of real estate investment, may have an impact on TI. So, China provides a good case study for examining the relationship between HB and TI. This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, it has largely overlooked interplay between these two phenomena. Our study examines the bidirectional influences between HB and TI, providing a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024a) to capture the dynamic interplay across various time periods. Third, most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 4. The literature reviewed is comprehensive, but it could be synthesized to reduce length and improve readability. However, recent references linking technological innovation to global housing policies are missing. (citations) Response: Thank you so much for your comments. We have synthesized the literature on "The Impact of HB on TI" and supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2-3): 2. Literature review 2.1 The impact of HB on TI The literature extensively explores the impact of HB on TI, revealing both positive and negative impacts. On one hand, HB can enhance corporate financing and investment, thereby supporting TI. Studies like Aghion et al. (2012) highlight that real estate serves as valuable collateral, boosting firms' debt capacity. Similarly, Caballero et al. (2006) note that during bubble growth, high investment levels are sustained through leveraging bubble assets. Chaney et al. (2012) and Corradin et al. (2015) emphasize how rising house prices and housing wealth alleviate credit constraints, enabling innovation investment. Rong et al. (2020) and Gan (2007) further illustrate how real estate value fluctuations directly influence R&D investment, with booms stimulating and busts constraining it. On the other hand, HB can crowd out R&D investment, negatively affecting TI. Saint-Paul (1992) and Battiati (2019) argue that speculative bubbles divert resources from productive investments, including R&D. Chakraborty et al. (2018) find that banks in robust housing markets prioritize mortgage lending over commercial lending, reducing R&D funding. In China, most studies suggest HB crowds out TI funds. Chen et al. (2015) and Chu et al. (2024) observe that rising housing prices encourage real estate investment at the expense of non-real estate sectors, particularly harming less innovative firms. Rong et al. (2016), Yang et al. (2022), and Wang (2017) document how real estate speculation and diversification reduce R&D intensity and innovation output. Yin et al. (2022) and Jia et al. (2021) highlight how housing price increases shift bank credit toward real estate, crowding out manufacturing loans. Wang et al. (2021) and Liang et al. (2024) confirm that the capital relocation effect outweighs the collateral effect, hindering corporate competitiveness and innovation. Chu et al. (2023) note that urban HB lead to investment distortions, thereby weakening overall research and innovation capacity. Rong et al. (2016) find that HB negatively impact the innovation capabilities of manufacturing firms. Li et al. (2023) reveal that HB not only suppress firms' innovation input but also weaken innovation output and the capacity to transform innovation achievements. However, some studies present contrasting views. Mao (2021) and He et al. (2022) find that HB boosts both the quantity and quality of innovation, while Lin et al. (2021) link rising housing prices to increased city innovation and talent attraction. Cao et al. (2015) show that real estate value shocks enhance financing capacity and stimulate innovation. Chen et al. (2024) highlight that HB enhance urban innovation through specific mechanisms, such as attracting and concentrating talent and generating spatial spillover effects that benefit neighboring cities. Other studies suggest a more nuanced relationship. Han et al. (2017) and Miao et al. (2014) argue that while HB increases investment through collateral, it may displace other investments. Yu et al. (2021) and Liu et al. (2024) reveal that housing price increases initially boost urban innovation but eventually hinder it, with spillover effects impacting surrounding cities. Chu et al. (2024) explore the nonlinear relationship between housing prices and corporate innovation, identifying a threshold of 2.82% in housing prices growth rates, beyond which the negative impact on innovation significantly intensifies. 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 5. The methodological section is well-structured but could expand on the explanation of the rolling-window bootstrap method, providing examples of previous studies demonstrating its effectiveness. It would be useful to clarify the criteria for selecting control parameters, such as financing constraints, and to discuss the robustness of the results concerning variations in model parameters. Response: Thank you so much for your comments. We have expanded on the explanation of the rolling-window bootstrap method, Which is shown as follows (Pages 5): 4.3 Bootstrap sub-sample rolling-window causality test To tackle the problem of parameter structural changes, we utilize the sub-sample rolling-window causality test developed by Balcilar et al. (2010). This method involves dividing the entire sample into smaller sub-samples with a fixed window width to test for causality, and then rolling these sub-samples from the beginning to the end of the full sample. The specific steps are as follows: in a time series of length T, set the sub-sample length to f, and define the end of each sub-sample as τ = f, f+1, ..., T, thus constructing T-f sub-samples. Based on the RB-adjusted LR causality test, each sub-sample yields an empirical result for the causality test. By aggregating all observed p-values and LR statistics in chronological order, the results of the rolling window causality test for the sub-samples can be obtained. Equation (4) describes the impact of HB on TI. (4) Here, represents the number of bootstrap repetitions, and denotes the bootstrap estimator derived from the VAR model in Equation (4). Similarly, Equation (5) is used to analyze the impact of TI on HB, where represents the bootstrap estimator obtained from the VAR model in Equation (5). (5) This study employs a 90% confidence interval and uniformly removes the top and bottom 5% of the bootstrap-estimated values to eliminate excessively large or small values, ensuring the accuracy of the test (Su et al., 2024c). We have clarifide the rationale for selecting financial control as a control variable, as detailed below (Pages 6): 5. Data source and descriptive analysis ........Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. ........ To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. 6. The description of the results is detailed but requires greater emphasis on their practical relevance. Graphs and tables could be complemented with more concise explanations. Response: Thank you so much for your comments. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 8-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 7. The discussion is well-developed but lacks a dedicated section on the study's limitations. For instance, reliance on Chinese data might limit its generalizability. Response: Thank you so much for your comments. We have supplemented the comparative analysis content with India. so,we have discussed the study's limitations as follows (Pages 16-17): The causal relationship between HB and TI not only shows a time-varying structure but also varies across regions, populations, and other dimensions. This study focuses on temporal variability using data from China, and conducted a comparative analysis with India. Account for cultural, economic, and institutional differences, which may limit the generalizability of the findings to other contexts. Future research should prioritize investigating these variations to uncover nuanced dynamics and enhance broader applicability. 8. Some repetitions could be eliminated to improve the cohesion of the text. Response: Thank you so much for your comments. We have eliminated Some repetitions as follows: ......... HB mitigates the risk associated with real estate investment while enhancing returns, thereby stimulating investment in the real estate industry. Consequently, this leads to a decrease in investment in non-real estate industries, resulting in a crowding out effect on investments of TI. ......... because of high returns and low risks of real estate. Then, from the perspective of financing, the crowding out effect is formed again ........ In essence, the relationship between HB and TI is intricate and challenging to delineate, resulting in diverse innovation activities and outcomes. ........ Our observations indicate that HB positively influences TI during certain periods, aligning with the liquidity effect. Conversely, during other periods, TI experiences effects from HB, consistent with the crowding out effect. Additionally, fluctuations in TI lead to corresponding changes in HB direction during some impactful periods, indicating that TI can stimulate HB. Concentrate into: Our observations indicate that HB positively and negative influences TI, TI only stimulate HB. 9. Recommended literature: Resampling techniques for real estate appraisals: Testing the bootstrap approach / Del Giudice, Vincenzo; Salvo, Francesca; De Paola, Pierfrancesco. - In: SUSTAINABILITY. - ISSN 2071-1050. - 10:9(2018), p. 3085. [10.3390/su10093085] Rogoff, K., & Yang, Y. (2024). CHINA’S REAL ESTATE CHALLENGE. FINANCE & DEVELOPMENT MAGAZINE，28-32. Atkinson, R. D. (2024). China Is Rapidly Becoming A Leading Innovator in Advanced Industries. Information Technology and Innovation Foundation: Washington, DC, USA. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34-58. Li, J., Lyu, P., & Jin, C. (2023). The Impact of Housing Prices on Regional Innovation Capacity: Evidence from China. Sustainability, 15(15), 11868. Response: Thank you so much for your recommended literature. These recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited them as references. Reply to the Reviewer 2 I have several suggestions for improving this work. 1) I suggest improving the introduction section by highlighting the research gaps, significance of the research, aims and novelties, and research questions. Response: Thank you so much for your comments. In order to more clearly articulate and highlight the research gaps, the significance of the research, the aims and novelties, as well as the research questions, I have rewritten the first and third paragraphs of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The primary aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). The dynamics of HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, the escalation in housing prices associated with HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This, in turn, enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). When the liquidity effect predominates, the impact of HB on TI is positive, acting as a boon for TI. Conversely, when the crowding-out effect is more pronounced, the impact turns negative, rendering HB a bane for TI. The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been extensively documented and acknowledged globally. However, the effect of HB on TI remains a subject of debate, lacking a unified consensus due to its complex and variable nature, which hinges on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000 ...... This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, or vice versa, it has largely overlooked the intricate and multifaceted interplay between these two phenomena. Our study breaks new ground by examining the bidirectional influences between HB and TI, offering a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. As a result, the nuanced and often overlooked dynamics between HB and TI remain understudied. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024) to capture the dynamic interplay between HB and TI across various time periods. This methodological innovation enables a more refined understanding of their relationship, shedding light on previously uncharted aspects of this critical economic interplay. Third, while most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. Fourth, previous studies have largely concentrated on data intervals before 2015, leaving a significant gap in understanding the effects of the new HB on TI. Our research addresses this limitation by extending the data coverage from 2000 to 2024. This extended timeframe encompasses pivotal events such as China's economic new normal, the COVID-19 pandemic, the Sino-US trade war, and the collapse of Silicon Valley Bank. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 2) I recommend adding more studies related to housing like "The real estate industry in Turkey: a time series analysis. The Service Industries Journal, 2021, 41(5-6), 427-439." to the literature review section. Response: Thank you so much for your recommended literature. This recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited it as reference. 3) The policy implications need to improve by considering the various stakeholders involved. Response: Thank you so much for your comments. I have revised the policy implications by taking into account the perspectives of all relevant stakeholders. Which is shown as follows (Pages 16): Balancing HB and TI is a complex issue involving multiple stakeholders. Governments, enterprises, and banks each play key roles. Here are recommendations: First, governments should stabilize housing prices by refining tax policies to prevent excessive capital from flowing into the real estate market, which could exacerbate the crowding-out effect of HB on TI. Additionally, governments should enhance policy support and increase funding for TI to improve investment returns and make the sector more attractive. Second, enterprises should prioritize R&D and TI to bolster their core competitiveness. They should avoid over-reliance on real estate investments and instead pursue diversified growth strategies to thrive in the evolving economic landscape. Third, banks should optimize their credit structures by controlling the proportion of real estate loans and increasing financing for technology-driven firms. Offering a variety of financial services support sustainable innovation. 4) I suggest adding the robustness check the test the validity of the findings by both liquidity and crowding out effects. Response: Thank you so much for your comments. To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14-15): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply enhances liquidity, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding-out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding-out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. Reply to the Reviewer 3 Article Title: The Dynamic Relationship Between Housing Boom (HB) and Technological Innovation (TI): An Examination Using the Bootstrap Rolling-Window Method Article Type: Research Article Abstract Evaluation: The abstract clearly presents the main objective of the study, which is to examine the dynamic relationship between the housing boom (HB) and technological innovation (TI). The use of the bootstrap rolling-window method for analysis is appropriately mentioned, and it provides a solid overview of the key findings. Additionally, the article outlines China's economic context and policy recommendations based on the results. However, some aspects, such as the methodological approach and specific empirical findings, could be more briefly summarized to provide a clearer preview of the article's core contribution. General Evaluation: The article effectively investigates the relationship between HB and TI, focusing on China’s economic development and technological advancements. By using an advanced analytical technique and examining data over time, the article provides a meaningful contribution to understanding how these two sectors interact. However, the study could be strengthened by expanding its analysis to include comparisons with other developing economies, offering a broader context for the findings. Methodology Evaluation: The article employs the bootstrap rolling-window method, a robust approach that captures the time-varying relationship between variables. This method is appropriate for the study’s objectives and provides reliable results. However, more detailed information about the data sources, variable selection, and model specifications would improve transparency and reproducibility. Clarifying the underlying assumptions and explaining how the rolling window method was applied could strengthen the methodological clarity. Results and Discussion: The article presents a thorough analysis of the results, explaining the positive and negative effects of the housing boom on technological innovation. The negative impact is attributed to the crowding-out effect of innovation funds due to increased investments in real estate. On the other hand, the positive effect is linked to the rising housing prices, which increase corporate real estate mortgage values and thus enhance financing liquidity. Additionally, the positive relationship between TI and HB is convincingly discussed at the national, regional, and firm levels. The policy implications derived from the study are relevant and practical. The recommendations to balance real estate investment and innovation funding are valuable for policymakers, and the suggestions for market-oriented policies to reduce speculative returns in real estate are well-founded. Constructive Criticism: 1. Literature Review: The literature review section could benefit from a more comprehensive discussion, incorporating additional references and studies related to HB and TI in other contexts. This would provide a broader theoretical foundation for the study. 2. Data Sources and Analysis Details: More information is needed regarding the data sources and the specifics of the analysis. The inclusion of these details would enhance the credibility of the study and allow for a better understanding of the analytical process. 3. Comparative Analysis: While the study focuses on China, comparing the results with other developing economies could provide a more generalized perspective on the relationship between HB and TI. Conclusion: Overall, the article makes a significant contribution to understanding the dynamic relationship between housing markets and technological innovation. The methodology is solid, and the results are well-presented. However, there is room for improvement in the literature review, the transparency of the data and methodology, and the inclusion of a comparative analysis. The article is a valuable contribution to the field, but it would benefit from minor revisions to enhance clarity and expand the analysis. Suggested Revisions: 1. Expand the literature review to include more references. Response: Thank you so much for your comments. We have supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2): 2 Literature review 2.1 The Impact of HB on TI 。。。。 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 2. Provide more details on data sources and the analytical methodology. Response: Thank you so much for your comments. We have further elaborated the data sources as follows (Pages 6): This paper examines the causal relationship between HB and TI using monthly data from 2000M1 to 2024M01. In 2000, China joins the World Trade Organization. This not only enables China's economy to integrate into the global market faster and better, but also enables Chinese enterprises to grow rapidly in global competition, and accelerates TI (Geng et al., 2021). In addition, since 2000, rapid economic development and wealth accumulation have given rise to strong housing demand, resulting in long-term HB (Jiang et al., 2020). There are two main methods to measure TI, input method and output method. Innovation output indicators, especially the number of patent applications, can better reflect the level of TI (Griliches,1990; Cornaggia et al.,2015). So, We choose the monthly number of successful patent applications to reflect the degree of national TI (Su et al.,2022). The data are obtained from the World Intellectual Property Organization. In addition, we use the Monthly Real Estate Climate Index (MRECI), issued by China's National Burea of Statistics, to reflect the degree of HB. MRECI is an index that comprehensively reflects the operation and fluctuation of China's real estate. The MRECI selects 2000 as the base year, setting its growth level at 100. Typically, a level of 100 points is considered the most suitable, with 95 to 105 points indicating a moderate level, below 95 indicating a low level, and above 105 indicating a high level. Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. The data of FC comes from official website, the People’s Bank of China. As for the indicators to measure FC, this paper selects the loan growth of financial institutions. As shown in Table 1. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 7-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 3. Include comparative analysis with other developing economies. Response: Thank you so much for your comments. We consider this is a good idea for in-depth research. We have selected another member of the BRICS nations, India, to conduct a comparative analysis. Due to space limitations, only the main Figures and brief analysis results are presented as follows (Pages 15-16). India shares many similarities with China in terms of economic development, real estate market, population size, TI and participation in globalization, making it an excellent comparative subject for analyzing China alongside other developing economies. This paper examines the bidirectional interaction between HB and TI in India, using monthly data from 2009M01 to 2024M01. We choose number of patent applications to reflect the degree of Indian TI. The data are obtained from the World Intellectual Property Organization (WIPO), We choose Real House Price Index for India (RHPII) to reflect the Indian HB. The data are obtained from Organisation for Economic Co-operation and Development(OECD). Figure 11-14 presents the empirical findings on the bidirectional relationship between HB and TI in India. By analyzing these four figures, we observe that from 2009M01 to 2024M01, HB exerts a significant positive influence on TI without any adverse effects. on the other hand, TI also actively promotes HB. These results inconsistent with findings from China. Some explanations are given: Liquidity effect channel: First, HB enhance the wealth perception of property owners, particularly among India's middle class and tech professionals. This wealth effect boosts their consumption capacity and investment confidence. Tech professionals may reinvest the increased value of their properties into entrepreneurship or tech projects, indirectly supporting TI. Second, in Indian cities with concentrated tech industries, housing prices rise rapidly, and market liquidity is high. Companies can sell properties or secure loans against them to raise funds, thereby increasing investments in TI. Crowding out effect channel: First, The Indian government and society prioritize the tech industry, viewing it as a core driver of economic growth. As a result, even amid rising housing prices, financial and policy resources continue to flow primarily into the tech sector rather than real estate. This prioritization minimizes the occurrence of the crowding-out effect. Second, India is a global hotspot for tech investments, attracting substantial foreign capital. These investments are predominantly directed toward the tech sector rather than real estate. The steady inflow of foreign capital provides ample funding for TI, offsetting any potential crowding-out effects caused by HB. Third, many tech companies, especially in software and internet sectors, operate with a light-asset model, reducing their reliance on real estate. They focus more on talent, technology, and market expansion rather than accumulating fixed assets. This approach limits the crowding-out effect of rising housing prices on tech companies. Figure 11. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI in India. Figure 12. The coefficients for the effect of HB on TI in India. Figure 13. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB in India. Figure 14. The coefficients for the effect of TI on HB in India.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Comments and Suggestions for AuthorsI have several suggestions for improving this work.
1) I suggest improving the introduction section by highlighting the research gaps, significance of the research, aims and novelties, and research questions.
2) I recommend adding more studies related to housing like "The real estate industry in Turkey: a time series analysis. The Service Industries Journal, 2021, 41(5-6), 427-439." to the literature review section.
3) The policy implications need to improve by considering the various stakeholders involved.
4) I suggest adding the robustness check the test the validity of the findings by both liquidity and crowding out effects.
Comments on the Quality of English LanguageIt's good.
Author Response
HB+TI Response: Reply to the Reviewer 1 The document analyzes the impact of housing booms (HB) on technological innovation (TI) in China, using advanced causal analysis methods and dynamic models. However, improvements in structure, clarity, and contextualization of the results could significantly enhance the academic and practical impact of the work. 1. The title, “Is Housing Boom a Blessing or Curse for Technological Innovation?”, could be improved by including references to the geographical (China) and methodological context. Response: Thank you so much for your comments. We have revised the article title, and the new title includes the geographical (China) and methodological context of the study, which is shown as follows : The title: Dynamic Causality Between Housing Boom and Technological Innovation in China: A Sub-Sample Rolling-Window Analysis 2. The abstract clearly describes the topic but should be more specific about the Response: Thank you so much for your comments. We have highlighted methodological tools used and the main findings in abstract Section, which is shown as follows (Pages 1): Abstract: This paper employs bootstrap rolling-window tests to investigate the dynamic causal relationship between housing boom (HB) and technological innovation (TI) in China. Through sub-sample analysis, we reveal a dual impact of HB on TI: during periods dominated by the liquidity effect, HB exerts a positive influence on TI, whereas during periods dominated by the crowding out effect, HB negatively affects TI. Furthermore, the study identifies a significant positive effect of TI on HB, suggesting that TI serves as a predictor of real estate development trends. This research not only provides empirical evidence on the bidirectional interaction between HB and TI, but also offers valuable insights for policymakers in balancing the development of the real estate market and TI. 3. The introduction section provides good context but requires a better transition between the discussion of housing booms and technological innovation. The unique contribution of the paper compared to existing literature should be articulated more clearly. Response: Thank you so much for your comments. To create a smoother connection between the discussion of HB and TI, we have reorganized the content of the first paragraph of the introduction and added a transitional sentence. To more clearly articulate the unique contributions of this paper, we have supplemented the content of the third paragraph of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been acknowledged globally. However, the impact of HB on TI remains a topic of debate, lacking a unified consensus. This is due to its complex and variable nature, which depends on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000, China’s real estate sector has experienced significant growth, largely attributed to the rapid economic development and the housing commercialization policy. At present, China have the biggest real estate market in the whole world (Shi et al., 2018). At the same time, China is also the world’s fastest-growing in terms of housing prices (Wang et al., 2024). High housing prices bring high profits and high return on investment (Yang et al., 2024). According to statistics from the National Bureau of Statistics of China (NBSC), the return on investment in China’s real estate industry is twice as high as the return on investment in manufacturing. The high return not only attracts a significant amount of capital from non-real estate enterprises (Li et al., 2021), but also attracts all kinds of financial capital (Wang et al., 2024). The total investment in real estate increases 30 times, from the initial 0.36 trillion yuan in 1998 to 11.09 trillion yuan in 2023. The participation of a large number of non-real estate enterprises further intensifies HB (Meng et al., 2018). However, the main business and technical expertise of non-real estate companies are not in the real estate sector. Therefore, China's HB, as well as the expansion of real estate investment, may have an impact on TI. So, China provides a good case study for examining the relationship between HB and TI. This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, it has largely overlooked interplay between these two phenomena. Our study examines the bidirectional influences between HB and TI, providing a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024a) to capture the dynamic interplay across various time periods. Third, most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 4. The literature reviewed is comprehensive, but it could be synthesized to reduce length and improve readability. However, recent references linking technological innovation to global housing policies are missing. (citations) Response: Thank you so much for your comments. We have synthesized the literature on "The Impact of HB on TI" and supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2-3): 2. Literature review 2.1 The impact of HB on TI The literature extensively explores the impact of HB on TI, revealing both positive and negative impacts. On one hand, HB can enhance corporate financing and investment, thereby supporting TI. Studies like Aghion et al. (2012) highlight that real estate serves as valuable collateral, boosting firms' debt capacity. Similarly, Caballero et al. (2006) note that during bubble growth, high investment levels are sustained through leveraging bubble assets. Chaney et al. (2012) and Corradin et al. (2015) emphasize how rising house prices and housing wealth alleviate credit constraints, enabling innovation investment. Rong et al. (2020) and Gan (2007) further illustrate how real estate value fluctuations directly influence R&D investment, with booms stimulating and busts constraining it. On the other hand, HB can crowd out R&D investment, negatively affecting TI. Saint-Paul (1992) and Battiati (2019) argue that speculative bubbles divert resources from productive investments, including R&D. Chakraborty et al. (2018) find that banks in robust housing markets prioritize mortgage lending over commercial lending, reducing R&D funding. In China, most studies suggest HB crowds out TI funds. Chen et al. (2015) and Chu et al. (2024) observe that rising housing prices encourage real estate investment at the expense of non-real estate sectors, particularly harming less innovative firms. Rong et al. (2016), Yang et al. (2022), and Wang (2017) document how real estate speculation and diversification reduce R&D intensity and innovation output. Yin et al. (2022) and Jia et al. (2021) highlight how housing price increases shift bank credit toward real estate, crowding out manufacturing loans. Wang et al. (2021) and Liang et al. (2024) confirm that the capital relocation effect outweighs the collateral effect, hindering corporate competitiveness and innovation. Chu et al. (2023) note that urban HB lead to investment distortions, thereby weakening overall research and innovation capacity. Rong et al. (2016) find that HB negatively impact the innovation capabilities of manufacturing firms. Li et al. (2023) reveal that HB not only suppress firms' innovation input but also weaken innovation output and the capacity to transform innovation achievements. However, some studies present contrasting views. Mao (2021) and He et al. (2022) find that HB boosts both the quantity and quality of innovation, while Lin et al. (2021) link rising housing prices to increased city innovation and talent attraction. Cao et al. (2015) show that real estate value shocks enhance financing capacity and stimulate innovation. Chen et al. (2024) highlight that HB enhance urban innovation through specific mechanisms, such as attracting and concentrating talent and generating spatial spillover effects that benefit neighboring cities. Other studies suggest a more nuanced relationship. Han et al. (2017) and Miao et al. (2014) argue that while HB increases investment through collateral, it may displace other investments. Yu et al. (2021) and Liu et al. (2024) reveal that housing price increases initially boost urban innovation but eventually hinder it, with spillover effects impacting surrounding cities. Chu et al. (2024) explore the nonlinear relationship between housing prices and corporate innovation, identifying a threshold of 2.82% in housing prices growth rates, beyond which the negative impact on innovation significantly intensifies. 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 5. The methodological section is well-structured but could expand on the explanation of the rolling-window bootstrap method, providing examples of previous studies demonstrating its effectiveness. It would be useful to clarify the criteria for selecting control parameters, such as financing constraints, and to discuss the robustness of the results concerning variations in model parameters. Response: Thank you so much for your comments. We have expanded on the explanation of the rolling-window bootstrap method, Which is shown as follows (Pages 5): 4.3 Bootstrap sub-sample rolling-window causality test To tackle the problem of parameter structural changes, we utilize the sub-sample rolling-window causality test developed by Balcilar et al. (2010). This method involves dividing the entire sample into smaller sub-samples with a fixed window width to test for causality, and then rolling these sub-samples from the beginning to the end of the full sample. The specific steps are as follows: in a time series of length T, set the sub-sample length to f, and define the end of each sub-sample as τ = f, f+1, ..., T, thus constructing T-f sub-samples. Based on the RB-adjusted LR causality test, each sub-sample yields an empirical result for the causality test. By aggregating all observed p-values and LR statistics in chronological order, the results of the rolling window causality test for the sub-samples can be obtained. Equation (4) describes the impact of HB on TI. (4) Here, represents the number of bootstrap repetitions, and denotes the bootstrap estimator derived from the VAR model in Equation (4). Similarly, Equation (5) is used to analyze the impact of TI on HB, where represents the bootstrap estimator obtained from the VAR model in Equation (5). (5) This study employs a 90% confidence interval and uniformly removes the top and bottom 5% of the bootstrap-estimated values to eliminate excessively large or small values, ensuring the accuracy of the test (Su et al., 2024c). We have clarifide the rationale for selecting financial control as a control variable, as detailed below (Pages 6): 5. Data source and descriptive analysis ........Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. ........ To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. 6. The description of the results is detailed but requires greater emphasis on their practical relevance. Graphs and tables could be complemented with more concise explanations. Response: Thank you so much for your comments. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 8-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 7. The discussion is well-developed but lacks a dedicated section on the study's limitations. For instance, reliance on Chinese data might limit its generalizability. Response: Thank you so much for your comments. We have supplemented the comparative analysis content with India. so,we have discussed the study's limitations as follows (Pages 16-17): The causal relationship between HB and TI not only shows a time-varying structure but also varies across regions, populations, and other dimensions. This study focuses on temporal variability using data from China, and conducted a comparative analysis with India. Account for cultural, economic, and institutional differences, which may limit the generalizability of the findings to other contexts. Future research should prioritize investigating these variations to uncover nuanced dynamics and enhance broader applicability. 8. Some repetitions could be eliminated to improve the cohesion of the text. Response: Thank you so much for your comments. We have eliminated Some repetitions as follows: ......... HB mitigates the risk associated with real estate investment while enhancing returns, thereby stimulating investment in the real estate industry. Consequently, this leads to a decrease in investment in non-real estate industries, resulting in a crowding out effect on investments of TI. ......... because of high returns and low risks of real estate. Then, from the perspective of financing, the crowding out effect is formed again ........ In essence, the relationship between HB and TI is intricate and challenging to delineate, resulting in diverse innovation activities and outcomes. ........ Our observations indicate that HB positively influences TI during certain periods, aligning with the liquidity effect. Conversely, during other periods, TI experiences effects from HB, consistent with the crowding out effect. Additionally, fluctuations in TI lead to corresponding changes in HB direction during some impactful periods, indicating that TI can stimulate HB. Concentrate into: Our observations indicate that HB positively and negative influences TI, TI only stimulate HB. 9. Recommended literature: Resampling techniques for real estate appraisals: Testing the bootstrap approach / Del Giudice, Vincenzo; Salvo, Francesca; De Paola, Pierfrancesco. - In: SUSTAINABILITY. - ISSN 2071-1050. - 10:9(2018), p. 3085. [10.3390/su10093085] Rogoff, K., & Yang, Y. (2024). CHINA’S REAL ESTATE CHALLENGE. FINANCE & DEVELOPMENT MAGAZINE，28-32. Atkinson, R. D. (2024). China Is Rapidly Becoming A Leading Innovator in Advanced Industries. Information Technology and Innovation Foundation: Washington, DC, USA. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34-58. Li, J., Lyu, P., & Jin, C. (2023). The Impact of Housing Prices on Regional Innovation Capacity: Evidence from China. Sustainability, 15(15), 11868. Response: Thank you so much for your recommended literature. These recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited them as references. Reply to the Reviewer 2 I have several suggestions for improving this work. 1) I suggest improving the introduction section by highlighting the research gaps, significance of the research, aims and novelties, and research questions. Response: Thank you so much for your comments. In order to more clearly articulate and highlight the research gaps, the significance of the research, the aims and novelties, as well as the research questions, I have rewritten the first and third paragraphs of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The primary aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). The dynamics of HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, the escalation in housing prices associated with HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This, in turn, enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). When the liquidity effect predominates, the impact of HB on TI is positive, acting as a boon for TI. Conversely, when the crowding-out effect is more pronounced, the impact turns negative, rendering HB a bane for TI. The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been extensively documented and acknowledged globally. However, the effect of HB on TI remains a subject of debate, lacking a unified consensus due to its complex and variable nature, which hinges on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000 ...... This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, or vice versa, it has largely overlooked the intricate and multifaceted interplay between these two phenomena. Our study breaks new ground by examining the bidirectional influences between HB and TI, offering a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. As a result, the nuanced and often overlooked dynamics between HB and TI remain understudied. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024) to capture the dynamic interplay between HB and TI across various time periods. This methodological innovation enables a more refined understanding of their relationship, shedding light on previously uncharted aspects of this critical economic interplay. Third, while most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. Fourth, previous studies have largely concentrated on data intervals before 2015, leaving a significant gap in understanding the effects of the new HB on TI. Our research addresses this limitation by extending the data coverage from 2000 to 2024. This extended timeframe encompasses pivotal events such as China's economic new normal, the COVID-19 pandemic, the Sino-US trade war, and the collapse of Silicon Valley Bank. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 2) I recommend adding more studies related to housing like "The real estate industry in Turkey: a time series analysis. The Service Industries Journal, 2021, 41(5-6), 427-439." to the literature review section. Response: Thank you so much for your recommended literature. This recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited it as reference. 3) The policy implications need to improve by considering the various stakeholders involved. Response: Thank you so much for your comments. I have revised the policy implications by taking into account the perspectives of all relevant stakeholders. Which is shown as follows (Pages 16): Balancing HB and TI is a complex issue involving multiple stakeholders. Governments, enterprises, and banks each play key roles. Here are recommendations: First, governments should stabilize housing prices by refining tax policies to prevent excessive capital from flowing into the real estate market, which could exacerbate the crowding-out effect of HB on TI. Additionally, governments should enhance policy support and increase funding for TI to improve investment returns and make the sector more attractive. Second, enterprises should prioritize R&D and TI to bolster their core competitiveness. They should avoid over-reliance on real estate investments and instead pursue diversified growth strategies to thrive in the evolving economic landscape. Third, banks should optimize their credit structures by controlling the proportion of real estate loans and increasing financing for technology-driven firms. Offering a variety of financial services support sustainable innovation. 4) I suggest adding the robustness check the test the validity of the findings by both liquidity and crowding out effects. Response: Thank you so much for your comments. To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14-15): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply enhances liquidity, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding-out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding-out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. Reply to the Reviewer 3 Article Title: The Dynamic Relationship Between Housing Boom (HB) and Technological Innovation (TI): An Examination Using the Bootstrap Rolling-Window Method Article Type: Research Article Abstract Evaluation: The abstract clearly presents the main objective of the study, which is to examine the dynamic relationship between the housing boom (HB) and technological innovation (TI). The use of the bootstrap rolling-window method for analysis is appropriately mentioned, and it provides a solid overview of the key findings. Additionally, the article outlines China's economic context and policy recommendations based on the results. However, some aspects, such as the methodological approach and specific empirical findings, could be more briefly summarized to provide a clearer preview of the article's core contribution. General Evaluation: The article effectively investigates the relationship between HB and TI, focusing on China’s economic development and technological advancements. By using an advanced analytical technique and examining data over time, the article provides a meaningful contribution to understanding how these two sectors interact. However, the study could be strengthened by expanding its analysis to include comparisons with other developing economies, offering a broader context for the findings. Methodology Evaluation: The article employs the bootstrap rolling-window method, a robust approach that captures the time-varying relationship between variables. This method is appropriate for the study’s objectives and provides reliable results. However, more detailed information about the data sources, variable selection, and model specifications would improve transparency and reproducibility. Clarifying the underlying assumptions and explaining how the rolling window method was applied could strengthen the methodological clarity. Results and Discussion: The article presents a thorough analysis of the results, explaining the positive and negative effects of the housing boom on technological innovation. The negative impact is attributed to the crowding-out effect of innovation funds due to increased investments in real estate. On the other hand, the positive effect is linked to the rising housing prices, which increase corporate real estate mortgage values and thus enhance financing liquidity. Additionally, the positive relationship between TI and HB is convincingly discussed at the national, regional, and firm levels. The policy implications derived from the study are relevant and practical. The recommendations to balance real estate investment and innovation funding are valuable for policymakers, and the suggestions for market-oriented policies to reduce speculative returns in real estate are well-founded. Constructive Criticism: 1. Literature Review: The literature review section could benefit from a more comprehensive discussion, incorporating additional references and studies related to HB and TI in other contexts. This would provide a broader theoretical foundation for the study. 2. Data Sources and Analysis Details: More information is needed regarding the data sources and the specifics of the analysis. The inclusion of these details would enhance the credibility of the study and allow for a better understanding of the analytical process. 3. Comparative Analysis: While the study focuses on China, comparing the results with other developing economies could provide a more generalized perspective on the relationship between HB and TI. Conclusion: Overall, the article makes a significant contribution to understanding the dynamic relationship between housing markets and technological innovation. The methodology is solid, and the results are well-presented. However, there is room for improvement in the literature review, the transparency of the data and methodology, and the inclusion of a comparative analysis. The article is a valuable contribution to the field, but it would benefit from minor revisions to enhance clarity and expand the analysis. Suggested Revisions: 1. Expand the literature review to include more references. Response: Thank you so much for your comments. We have supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2): 2 Literature review 2.1 The Impact of HB on TI 。。。。 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 2. Provide more details on data sources and the analytical methodology. Response: Thank you so much for your comments. We have further elaborated the data sources as follows (Pages 6): This paper examines the causal relationship between HB and TI using monthly data from 2000M1 to 2024M01. In 2000, China joins the World Trade Organization. This not only enables China's economy to integrate into the global market faster and better, but also enables Chinese enterprises to grow rapidly in global competition, and accelerates TI (Geng et al., 2021). In addition, since 2000, rapid economic development and wealth accumulation have given rise to strong housing demand, resulting in long-term HB (Jiang et al., 2020). There are two main methods to measure TI, input method and output method. Innovation output indicators, especially the number of patent applications, can better reflect the level of TI (Griliches,1990; Cornaggia et al.,2015). So, We choose the monthly number of successful patent applications to reflect the degree of national TI (Su et al.,2022). The data are obtained from the World Intellectual Property Organization. In addition, we use the Monthly Real Estate Climate Index (MRECI), issued by China's National Burea of Statistics, to reflect the degree of HB. MRECI is an index that comprehensively reflects the operation and fluctuation of China's real estate. The MRECI selects 2000 as the base year, setting its growth level at 100. Typically, a level of 100 points is considered the most suitable, with 95 to 105 points indicating a moderate level, below 95 indicating a low level, and above 105 indicating a high level. Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. The data of FC comes from official website, the People’s Bank of China. As for the indicators to measure FC, this paper selects the loan growth of financial institutions. As shown in Table 1. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 7-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 3. Include comparative analysis with other developing economies. Response: Thank you so much for your comments. We consider this is a good idea for in-depth research. We have selected another member of the BRICS nations, India, to conduct a comparative analysis. Due to space limitations, only the main Figures and brief analysis results are presented as follows (Pages 15-16). India shares many similarities with China in terms of economic development, real estate market, population size, TI and participation in globalization, making it an excellent comparative subject for analyzing China alongside other developing economies. This paper examines the bidirectional interaction between HB and TI in India, using monthly data from 2009M01 to 2024M01. We choose number of patent applications to reflect the degree of Indian TI. The data are obtained from the World Intellectual Property Organization (WIPO), We choose Real House Price Index for India (RHPII) to reflect the Indian HB. The data are obtained from Organisation for Economic Co-operation and Development(OECD). Figure 11-14 presents the empirical findings on the bidirectional relationship between HB and TI in India. By analyzing these four figures, we observe that from 2009M01 to 2024M01, HB exerts a significant positive influence on TI without any adverse effects. on the other hand, TI also actively promotes HB. These results inconsistent with findings from China. Some explanations are given: Liquidity effect channel: First, HB enhance the wealth perception of property owners, particularly among India's middle class and tech professionals. This wealth effect boosts their consumption capacity and investment confidence. Tech professionals may reinvest the increased value of their properties into entrepreneurship or tech projects, indirectly supporting TI. Second, in Indian cities with concentrated tech industries, housing prices rise rapidly, and market liquidity is high. Companies can sell properties or secure loans against them to raise funds, thereby increasing investments in TI. Crowding out effect channel: First, The Indian government and society prioritize the tech industry, viewing it as a core driver of economic growth. As a result, even amid rising housing prices, financial and policy resources continue to flow primarily into the tech sector rather than real estate. This prioritization minimizes the occurrence of the crowding-out effect. Second, India is a global hotspot for tech investments, attracting substantial foreign capital. These investments are predominantly directed toward the tech sector rather than real estate. The steady inflow of foreign capital provides ample funding for TI, offsetting any potential crowding-out effects caused by HB. Third, many tech companies, especially in software and internet sectors, operate with a light-asset model, reducing their reliance on real estate. They focus more on talent, technology, and market expansion rather than accumulating fixed assets. This approach limits the crowding-out effect of rising housing prices on tech companies. Figure 11. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI in India. Figure 12. The coefficients for the effect of HB on TI in India. Figure 13. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB in India. Figure 14. The coefficients for the effect of TI on HB in India.Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for AuthorsPeer Review Report
Article Title: The Dynamic Relationship Between Housing Boom (HB) and Technological Innovation (TI): An Examination Using the Bootstrap Rolling-Window Method
Article Type: Research Article
Abstract Evaluation:
The abstract clearly presents the main objective of the study, which is to examine the dynamic relationship between the housing boom (HB) and technological innovation (TI). The use of the bootstrap rolling-window method for analysis is appropriately mentioned, and it provides a solid overview of the key findings. Additionally, the article outlines China's economic context and policy recommendations based on the results. However, some aspects, such as the methodological approach and specific empirical findings, could be more briefly summarized to provide a clearer preview of the article's core contribution.
General Evaluation:
The article effectively investigates the relationship between HB and TI, focusing on China’s economic development and technological advancements. By using an advanced analytical technique and examining data over time, the article provides a meaningful contribution to understanding how these two sectors interact. However, the study could be strengthened by expanding its analysis to include comparisons with other developing economies, offering a broader context for the findings.
Methodology Evaluation:
The article employs the bootstrap rolling-window method, a robust approach that captures the time-varying relationship between variables. This method is appropriate for the study’s objectives and provides reliable results. However, more detailed information about the data sources, variable selection, and model specifications would improve transparency and reproducibility. Clarifying the underlying assumptions and explaining how the rolling window method was applied could strengthen the methodological clarity.
Results and Discussion:
The article presents a thorough analysis of the results, explaining the positive and negative effects of the housing boom on technological innovation. The negative impact is attributed to the crowding-out effect of innovation funds due to increased investments in real estate. On the other hand, the positive effect is linked to the rising housing prices, which increase corporate real estate mortgage values and thus enhance financing liquidity. Additionally, the positive relationship between TI and HB is convincingly discussed at the national, regional, and firm levels.
The policy implications derived from the study are relevant and practical. The recommendations to balance real estate investment and innovation funding are valuable for policymakers, and the suggestions for market-oriented policies to reduce speculative returns in real estate are well-founded.
Constructive Criticism:
- Literature Review: The literature review section could benefit from a more comprehensive discussion, incorporating additional references and studies related to HB and TI in other contexts. This would provide a broader theoretical foundation for the study.
- Data Sources and Analysis Details: More information is needed regarding the data sources and the specifics of the analysis. The inclusion of these details would enhance the credibility of the study and allow for a better understanding of the analytical process.
- Comparative Analysis: While the study focuses on China, comparing the results with other developing economies could provide a more generalized perspective on the relationship between HB and TI.
Conclusion:
Overall, the article makes a significant contribution to understanding the dynamic relationship between housing markets and technological innovation. The methodology is solid, and the results are well-presented. However, there is room for improvement in the literature review, the transparency of the data and methodology, and the inclusion of a comparative analysis.
The article is a valuable contribution to the field, but it would benefit from minor revisions to enhance clarity and expand the analysis.
Suggested Revisions:
- Expand the literature review to include more references.
- Provide more details on data sources and the analytical methodology.
- Include comparative analysis with other developing economies.
Comments on the Quality of English Language
The English could be improved to more clearly express the research.
Author Response
HB+TI Response: Reply to the Reviewer 1 The document analyzes the impact of housing booms (HB) on technological innovation (TI) in China, using advanced causal analysis methods and dynamic models. However, improvements in structure, clarity, and contextualization of the results could significantly enhance the academic and practical impact of the work. 1. The title, “Is Housing Boom a Blessing or Curse for Technological Innovation?”, could be improved by including references to the geographical (China) and methodological context. Response: Thank you so much for your comments. We have revised the article title, and the new title includes the geographical (China) and methodological context of the study, which is shown as follows : The title: Dynamic Causality Between Housing Boom and Technological Innovation in China: A Sub-Sample Rolling-Window Analysis 2. The abstract clearly describes the topic but should be more specific about the Response: Thank you so much for your comments. We have highlighted methodological tools used and the main findings in abstract Section, which is shown as follows (Pages 1): Abstract: This paper employs bootstrap rolling-window tests to investigate the dynamic causal relationship between housing boom (HB) and technological innovation (TI) in China. Through sub-sample analysis, we reveal a dual impact of HB on TI: during periods dominated by the liquidity effect, HB exerts a positive influence on TI, whereas during periods dominated by the crowding out effect, HB negatively affects TI. Furthermore, the study identifies a significant positive effect of TI on HB, suggesting that TI serves as a predictor of real estate development trends. This research not only provides empirical evidence on the bidirectional interaction between HB and TI, but also offers valuable insights for policymakers in balancing the development of the real estate market and TI. 3. The introduction section provides good context but requires a better transition between the discussion of housing booms and technological innovation. The unique contribution of the paper compared to existing literature should be articulated more clearly. Response: Thank you so much for your comments. To create a smoother connection between the discussion of HB and TI, we have reorganized the content of the first paragraph of the introduction and added a transitional sentence. To more clearly articulate the unique contributions of this paper, we have supplemented the content of the third paragraph of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been acknowledged globally. However, the impact of HB on TI remains a topic of debate, lacking a unified consensus. This is due to its complex and variable nature, which depends on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000, China’s real estate sector has experienced significant growth, largely attributed to the rapid economic development and the housing commercialization policy. At present, China have the biggest real estate market in the whole world (Shi et al., 2018). At the same time, China is also the world’s fastest-growing in terms of housing prices (Wang et al., 2024). High housing prices bring high profits and high return on investment (Yang et al., 2024). According to statistics from the National Bureau of Statistics of China (NBSC), the return on investment in China’s real estate industry is twice as high as the return on investment in manufacturing. The high return not only attracts a significant amount of capital from non-real estate enterprises (Li et al., 2021), but also attracts all kinds of financial capital (Wang et al., 2024). The total investment in real estate increases 30 times, from the initial 0.36 trillion yuan in 1998 to 11.09 trillion yuan in 2023. The participation of a large number of non-real estate enterprises further intensifies HB (Meng et al., 2018). However, the main business and technical expertise of non-real estate companies are not in the real estate sector. Therefore, China's HB, as well as the expansion of real estate investment, may have an impact on TI. So, China provides a good case study for examining the relationship between HB and TI. This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, it has largely overlooked interplay between these two phenomena. Our study examines the bidirectional influences between HB and TI, providing a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024a) to capture the dynamic interplay across various time periods. Third, most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 4. The literature reviewed is comprehensive, but it could be synthesized to reduce length and improve readability. However, recent references linking technological innovation to global housing policies are missing. (citations) Response: Thank you so much for your comments. We have synthesized the literature on "The Impact of HB on TI" and supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2-3): 2. Literature review 2.1 The impact of HB on TI The literature extensively explores the impact of HB on TI, revealing both positive and negative impacts. On one hand, HB can enhance corporate financing and investment, thereby supporting TI. Studies like Aghion et al. (2012) highlight that real estate serves as valuable collateral, boosting firms' debt capacity. Similarly, Caballero et al. (2006) note that during bubble growth, high investment levels are sustained through leveraging bubble assets. Chaney et al. (2012) and Corradin et al. (2015) emphasize how rising house prices and housing wealth alleviate credit constraints, enabling innovation investment. Rong et al. (2020) and Gan (2007) further illustrate how real estate value fluctuations directly influence R&D investment, with booms stimulating and busts constraining it. On the other hand, HB can crowd out R&D investment, negatively affecting TI. Saint-Paul (1992) and Battiati (2019) argue that speculative bubbles divert resources from productive investments, including R&D. Chakraborty et al. (2018) find that banks in robust housing markets prioritize mortgage lending over commercial lending, reducing R&D funding. In China, most studies suggest HB crowds out TI funds. Chen et al. (2015) and Chu et al. (2024) observe that rising housing prices encourage real estate investment at the expense of non-real estate sectors, particularly harming less innovative firms. Rong et al. (2016), Yang et al. (2022), and Wang (2017) document how real estate speculation and diversification reduce R&D intensity and innovation output. Yin et al. (2022) and Jia et al. (2021) highlight how housing price increases shift bank credit toward real estate, crowding out manufacturing loans. Wang et al. (2021) and Liang et al. (2024) confirm that the capital relocation effect outweighs the collateral effect, hindering corporate competitiveness and innovation. Chu et al. (2023) note that urban HB lead to investment distortions, thereby weakening overall research and innovation capacity. Rong et al. (2016) find that HB negatively impact the innovation capabilities of manufacturing firms. Li et al. (2023) reveal that HB not only suppress firms' innovation input but also weaken innovation output and the capacity to transform innovation achievements. However, some studies present contrasting views. Mao (2021) and He et al. (2022) find that HB boosts both the quantity and quality of innovation, while Lin et al. (2021) link rising housing prices to increased city innovation and talent attraction. Cao et al. (2015) show that real estate value shocks enhance financing capacity and stimulate innovation. Chen et al. (2024) highlight that HB enhance urban innovation through specific mechanisms, such as attracting and concentrating talent and generating spatial spillover effects that benefit neighboring cities. Other studies suggest a more nuanced relationship. Han et al. (2017) and Miao et al. (2014) argue that while HB increases investment through collateral, it may displace other investments. Yu et al. (2021) and Liu et al. (2024) reveal that housing price increases initially boost urban innovation but eventually hinder it, with spillover effects impacting surrounding cities. Chu et al. (2024) explore the nonlinear relationship between housing prices and corporate innovation, identifying a threshold of 2.82% in housing prices growth rates, beyond which the negative impact on innovation significantly intensifies. 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 5. The methodological section is well-structured but could expand on the explanation of the rolling-window bootstrap method, providing examples of previous studies demonstrating its effectiveness. It would be useful to clarify the criteria for selecting control parameters, such as financing constraints, and to discuss the robustness of the results concerning variations in model parameters. Response: Thank you so much for your comments. We have expanded on the explanation of the rolling-window bootstrap method, Which is shown as follows (Pages 5): 4.3 Bootstrap sub-sample rolling-window causality test To tackle the problem of parameter structural changes, we utilize the sub-sample rolling-window causality test developed by Balcilar et al. (2010). This method involves dividing the entire sample into smaller sub-samples with a fixed window width to test for causality, and then rolling these sub-samples from the beginning to the end of the full sample. The specific steps are as follows: in a time series of length T, set the sub-sample length to f, and define the end of each sub-sample as τ = f, f+1, ..., T, thus constructing T-f sub-samples. Based on the RB-adjusted LR causality test, each sub-sample yields an empirical result for the causality test. By aggregating all observed p-values and LR statistics in chronological order, the results of the rolling window causality test for the sub-samples can be obtained. Equation (4) describes the impact of HB on TI. (4) Here, represents the number of bootstrap repetitions, and denotes the bootstrap estimator derived from the VAR model in Equation (4). Similarly, Equation (5) is used to analyze the impact of TI on HB, where represents the bootstrap estimator obtained from the VAR model in Equation (5). (5) This study employs a 90% confidence interval and uniformly removes the top and bottom 5% of the bootstrap-estimated values to eliminate excessively large or small values, ensuring the accuracy of the test (Su et al., 2024c). We have clarifide the rationale for selecting financial control as a control variable, as detailed below (Pages 6): 5. Data source and descriptive analysis ........Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. ........ To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. 6. The description of the results is detailed but requires greater emphasis on their practical relevance. Graphs and tables could be complemented with more concise explanations. Response: Thank you so much for your comments. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 8-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 7. The discussion is well-developed but lacks a dedicated section on the study's limitations. For instance, reliance on Chinese data might limit its generalizability. Response: Thank you so much for your comments. We have supplemented the comparative analysis content with India. so,we have discussed the study's limitations as follows (Pages 16-17): The causal relationship between HB and TI not only shows a time-varying structure but also varies across regions, populations, and other dimensions. This study focuses on temporal variability using data from China, and conducted a comparative analysis with India. Account for cultural, economic, and institutional differences, which may limit the generalizability of the findings to other contexts. Future research should prioritize investigating these variations to uncover nuanced dynamics and enhance broader applicability. 8. Some repetitions could be eliminated to improve the cohesion of the text. Response: Thank you so much for your comments. We have eliminated Some repetitions as follows: ......... HB mitigates the risk associated with real estate investment while enhancing returns, thereby stimulating investment in the real estate industry. Consequently, this leads to a decrease in investment in non-real estate industries, resulting in a crowding out effect on investments of TI. ......... because of high returns and low risks of real estate. Then, from the perspective of financing, the crowding out effect is formed again ........ In essence, the relationship between HB and TI is intricate and challenging to delineate, resulting in diverse innovation activities and outcomes. ........ Our observations indicate that HB positively influences TI during certain periods, aligning with the liquidity effect. Conversely, during other periods, TI experiences effects from HB, consistent with the crowding out effect. Additionally, fluctuations in TI lead to corresponding changes in HB direction during some impactful periods, indicating that TI can stimulate HB. Concentrate into: Our observations indicate that HB positively and negative influences TI, TI only stimulate HB. 9. Recommended literature: Resampling techniques for real estate appraisals: Testing the bootstrap approach / Del Giudice, Vincenzo; Salvo, Francesca; De Paola, Pierfrancesco. - In: SUSTAINABILITY. - ISSN 2071-1050. - 10:9(2018), p. 3085. [10.3390/su10093085] Rogoff, K., & Yang, Y. (2024). CHINA’S REAL ESTATE CHALLENGE. FINANCE & DEVELOPMENT MAGAZINE，28-32. Atkinson, R. D. (2024). China Is Rapidly Becoming A Leading Innovator in Advanced Industries. Information Technology and Innovation Foundation: Washington, DC, USA. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34-58. Li, J., Lyu, P., & Jin, C. (2023). The Impact of Housing Prices on Regional Innovation Capacity: Evidence from China. Sustainability, 15(15), 11868. Response: Thank you so much for your recommended literature. These recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited them as references. Reply to the Reviewer 2 I have several suggestions for improving this work. 1) I suggest improving the introduction section by highlighting the research gaps, significance of the research, aims and novelties, and research questions. Response: Thank you so much for your comments. In order to more clearly articulate and highlight the research gaps, the significance of the research, the aims and novelties, as well as the research questions, I have rewritten the first and third paragraphs of the introduction. Which is shown as follows (Pages 1-2): 1. Introduction The primary aim of this research is to explore the bidirectional interaction between housing boom (HB) and technological innovation (TI) in China. HB profoundly influence the process of TI through channels such as capital flows, resource allocation, and market expectations (Ortiz-Villajos, 2018; Goel et al., 2022 ). TI, characterized by its high risk and extended return period, necessitates substantial and sustained R&D investments from enterprises (Xin et al., 2019). The dynamics of HB can significantly influence enterprise R&D investments, thereby impacting TI (Kuang et al., 2020). On one hand, HB may prompt corporations to reallocate investments from TI to the real estate sector in pursuit of profit maximization. This reallocation can result in a crowding-out effect on TI (Chen et al., 2015; Ning et al., 2024). On the other hand, the escalation in housing prices associated with HB can enhance the mortgage value of real estate, thereby alleviating corporate financing constraints. This, in turn, enables enterprises to secure more bank credit, potentially increasing R&D investments and generating a liquidity effect that benefits TI (Miao et al., 2014). When the liquidity effect predominates, the impact of HB on TI is positive, acting as a boon for TI. Conversely, when the crowding-out effect is more pronounced, the impact turns negative, rendering HB a bane for TI. The influence of HB on macroeconomic fluctuations (Liu et al., 2013) and financial risks (Tajik et al., 2015; Peng et al., 2017) has been extensively documented and acknowledged globally. However, the effect of HB on TI remains a subject of debate, lacking a unified consensus due to its complex and variable nature, which hinges on the relative strength of multiple effects. Consequently, HB and TI are intricately linked, presenting a significant and intriguing area of research that remains underexplored and misunderstood. This study aims to bridge this knowledge gap, shedding light on the nuanced relationship between HB and TI. Since 2000 ...... This study makes several contributions to the field. First, while existing research has primarily focused on the one-way impact of HB on TI, or vice versa, it has largely overlooked the intricate and multifaceted interplay between these two phenomena. Our study breaks new ground by examining the bidirectional influences between HB and TI, offering a more holistic perspective that captures their complex interdependencies. Second, earlier literature predominantly operates under linear assumptions, with limited exploration of dynamic and nonlinear perspectives. As a result, the nuanced and often overlooked dynamics between HB and TI remain understudied. To address this gap, we employ an advanced sub-sample technique (Su et al., 2024) to capture the dynamic interplay between HB and TI across various time periods. This methodological innovation enables a more refined understanding of their relationship, shedding light on previously uncharted aspects of this critical economic interplay. Third, while most existing research focuses on micro-level dynamics within enterprises and cities, our study adopts a macro, national perspective to examine the impact of HB on TI. This broader approach provides valuable insights for policymakers seeking to design strategies that foster TI while maintaining equilibrium in the real estate market. Fourth, previous studies have largely concentrated on data intervals before 2015, leaving a significant gap in understanding the effects of the new HB on TI. Our research addresses this limitation by extending the data coverage from 2000 to 2024. This extended timeframe encompasses pivotal events such as China's economic new normal, the COVID-19 pandemic, the Sino-US trade war, and the collapse of Silicon Valley Bank. By doing so, our study offers a comprehensive view of the HB-TI relationship, providing fresh insights into how these phenomena interact in the face of contemporary challenges. 2) I recommend adding more studies related to housing like "The real estate industry in Turkey: a time series analysis. The Service Industries Journal, 2021, 41(5-6), 427-439." to the literature review section. Response: Thank you so much for your recommended literature. This recommended literature have been highly inspiring and helpful for my writing. I have actively drawn on the perspectives from these sources and cited it as reference. 3) The policy implications need to improve by considering the various stakeholders involved. Response: Thank you so much for your comments. I have revised the policy implications by taking into account the perspectives of all relevant stakeholders. Which is shown as follows (Pages 16): Balancing HB and TI is a complex issue involving multiple stakeholders. Governments, enterprises, and banks each play key roles. Here are recommendations: First, governments should stabilize housing prices by refining tax policies to prevent excessive capital from flowing into the real estate market, which could exacerbate the crowding-out effect of HB on TI. Additionally, governments should enhance policy support and increase funding for TI to improve investment returns and make the sector more attractive. Second, enterprises should prioritize R&D and TI to bolster their core competitiveness. They should avoid over-reliance on real estate investments and instead pursue diversified growth strategies to thrive in the evolving economic landscape. Third, banks should optimize their credit structures by controlling the proportion of real estate loans and increasing financing for technology-driven firms. Offering a variety of financial services support sustainable innovation. 4) I suggest adding the robustness check the test the validity of the findings by both liquidity and crowding out effects. Response: Thank you so much for your comments. To ensure the robustness of the above quantitative results, this study replaces financing constraints (FC) with the money supply (M2) and economic policy uncertainty (EPU) ,as a control variable,conducted a new test. Which is shown as follows (Pages 14-15): To enhance the robustness of the quantitative findings, this study substitutes the initial control variable-financing constraints (FC), with the money supply (M2) and economic policy uncertainty (EPU) as alternative control variables, and performs a fresh round of testing. M2 significantly influences HB and TI through the liquidity effect channel. A larger money supply enhances liquidity, making it easier for businesses and individuals to obtain loans, thereby stimulating investment and consumption. This further drives housing price growth and corporate R&D investment, creating a liquidity effect on HB and TI. Therefore, M2 is added as a control variable. EPU significantly affects HB and TI through the crowding-out effect channel. When EPU rises, funds may choose to enter the real estate sector to seek short-term, stable, and substantial returns, rather than investing in R&D, which requires large inputs, long payback periods, and higher risks. This results in a crowding-out effect of HB on TI. Hence, EPU is included as another control variable. Figure 7-10 presents the evaluation results using M2 and EPU as control variable. we have observed that although this study replaces the control variable with M2 and EPU, the outcomes are comparable to those obtained from the prior research, providing proof of the quantitative analyses’s robustness. Figure 7. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 8. The coefficients for the effect of HB on TI. Figure 9. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB. Figure 10. The coefficients for the effect of TI on HB. Reply to the Reviewer 3 Article Title: The Dynamic Relationship Between Housing Boom (HB) and Technological Innovation (TI): An Examination Using the Bootstrap Rolling-Window Method Article Type: Research Article Abstract Evaluation: The abstract clearly presents the main objective of the study, which is to examine the dynamic relationship between the housing boom (HB) and technological innovation (TI). The use of the bootstrap rolling-window method for analysis is appropriately mentioned, and it provides a solid overview of the key findings. Additionally, the article outlines China's economic context and policy recommendations based on the results. However, some aspects, such as the methodological approach and specific empirical findings, could be more briefly summarized to provide a clearer preview of the article's core contribution. General Evaluation: The article effectively investigates the relationship between HB and TI, focusing on China’s economic development and technological advancements. By using an advanced analytical technique and examining data over time, the article provides a meaningful contribution to understanding how these two sectors interact. However, the study could be strengthened by expanding its analysis to include comparisons with other developing economies, offering a broader context for the findings. Methodology Evaluation: The article employs the bootstrap rolling-window method, a robust approach that captures the time-varying relationship between variables. This method is appropriate for the study’s objectives and provides reliable results. However, more detailed information about the data sources, variable selection, and model specifications would improve transparency and reproducibility. Clarifying the underlying assumptions and explaining how the rolling window method was applied could strengthen the methodological clarity. Results and Discussion: The article presents a thorough analysis of the results, explaining the positive and negative effects of the housing boom on technological innovation. The negative impact is attributed to the crowding-out effect of innovation funds due to increased investments in real estate. On the other hand, the positive effect is linked to the rising housing prices, which increase corporate real estate mortgage values and thus enhance financing liquidity. Additionally, the positive relationship between TI and HB is convincingly discussed at the national, regional, and firm levels. The policy implications derived from the study are relevant and practical. The recommendations to balance real estate investment and innovation funding are valuable for policymakers, and the suggestions for market-oriented policies to reduce speculative returns in real estate are well-founded. Constructive Criticism: 1. Literature Review: The literature review section could benefit from a more comprehensive discussion, incorporating additional references and studies related to HB and TI in other contexts. This would provide a broader theoretical foundation for the study. 2. Data Sources and Analysis Details: More information is needed regarding the data sources and the specifics of the analysis. The inclusion of these details would enhance the credibility of the study and allow for a better understanding of the analytical process. 3. Comparative Analysis: While the study focuses on China, comparing the results with other developing economies could provide a more generalized perspective on the relationship between HB and TI. Conclusion: Overall, the article makes a significant contribution to understanding the dynamic relationship between housing markets and technological innovation. The methodology is solid, and the results are well-presented. However, there is room for improvement in the literature review, the transparency of the data and methodology, and the inclusion of a comparative analysis. The article is a valuable contribution to the field, but it would benefit from minor revisions to enhance clarity and expand the analysis. Suggested Revisions: 1. Expand the literature review to include more references. Response: Thank you so much for your comments. We have supplemented it with additional literature on "The Impact of TI on HB". Which is shown as follows (Pages 2): 2 Literature review 2.1 The Impact of HB on TI 。。。。 2.2 The impact of TI on HB The impact of TI on HB garners significant attention and is considered positive. Beracha et al. (2023) find that innovation positively influences HB in the United States. Hirano et al. (2024) observe that asset price bubbles often emerge within broader historical trends driven by shifts in industrial structure due to TI. Quercia et al. (2002) note that high-tech activity significantly boosts housing prices, affecting moderate-income households. In China, scholars like Zhou & Liu (2024) find that population agglomeration, income growth, and TI significantly enhance HB. Dong & Zhu (2022) emphasize the positive impact of innovation factor aggregation on HB, while Wang & Yang (2022) highlight how improvements in regional innovation ecosystems improve residents' home-buying capacity and attract talent, further driving HB. Yang et al. (2020) conclude that enhanced TI capabilities significantly increase housing prices nationally. Gu & Jie (2024) and Zhang et al. (2023) also underscore the positive effects of talent concentration and urban innovation vitality on housing prices. In summary, while the existing body of research is extensive, there remain notable gaps that warrant further exploration. First, prior studies predominantly rely on linear assumptions, with limited investigation into dynamic and nonlinear perspectives. Second, much of the existing literature focuses on unidirectional effects, lacking in-depth examination of bidirectional interactions. Third, most research samples are concentrated at the enterprise and city levels, with insufficient attention paid to the national-level impact of housing markets on TI. These identified gaps offer promising avenues for future research, presenting valuable opportunities for this paper to delve deeper into unexplored dimensions and significantly expand the current understanding of these complex relationships. References for the newly added content. 1. Rong, Z., Wang, W., & Gong, Q. (2016). Housing price appreciation, investment opportunity, and firm innovation: Evidence from China. Journal of Housing Economics, 33, 34–58. 2. Li, J., Lyu, P., & Jin, C. (2023). The impact of housing prices on regional innovation capacity: Evidence from China. Sustainability, 15(15), 11868. 3.Chen, Z., Li, M., & Zhang, M. (2024). The Effect of Housing Prices on Urban Innovation Capability: New Evidence From 246 Chinese Cities. American Journal of Economics and Sociology, 83(5), ahead-of-print. 4. Chu, Z., Chen, X., Cheng, M. et al. (2024). Booming house prices: friend or foe of innovative firms? Journal of Technology Transfer, 49, 642–659. 5. Chu, M., Pan, L., Guo, M. et al. (2023). Has high housing prices affected urban green development?: Evidence from China. Journal of Housing and the Built Environment, 38, 2185–2206. 6. Beracha, E., He, Z., & Wintoki, M. B. (2022). On the relation between innovation and housing prices–A metro level analysis of the US market. Journal of Real Estate Finance and Economics, 65, 622–648. 7. Hirano, T., & Toda, A. A. (2024). Bubble economics. Journal of Mathematical Economics, 111, 102944. 8. Quercia, R. G., Stegman, M. A., & Davis, W. R. (2002). Does a high-tech boom worsen housing problems for working families? Housing Policy Debate, 13(2), 393–415. 9. Zhou, X., & Liu, S. A. (2024). How Does Digital Infrastructure Development Affect Housing Prices? A Quasi-Natural Experiment Based on the "Broadband China" Program. Housing Policy Debate, ahead-of-print. 10. Dong, F., & Zhu, L. (2022). Spatial Correlation between Innovation Aggregation and Housing Prices. International Conference on Construction and Real Estate Management (ICCREM), 2022, 581–587. 11. Wang, G., & Yang, H. (2022). Research on the relationship between the purchasing ability of regional residents and the gathering of scientific and technological talents---The threshold effect test based on the coupling of innovation ecology. Studies in Science of Science, 40(6), 1001–1013. 12. Yang, M. W., Sun, B. Y., & Zhao, Z. L. (2020). Sci-technological innovation ability, regional heterogeneity and housing price in China: An empirical study on 31 provinces in China. Journal of Chongqing University (Social Science Edition), 26(3), 50–65. 13. Gu, H., & Jie, Y. (2024). Escaping from “dream city”? Housing price, talent, and urban innovation in China. Habitat International, 145, 103015. 14. Zhang, J., Zhou, J., Qian, L., & Zhang, D. (2023). The inter-relationships among mobility, housing prices and innovation: evidence from China’s cities. International Journal of Strategic Property Management, 27(4), 233–245. 2. Provide more details on data sources and the analytical methodology. Response: Thank you so much for your comments. We have further elaborated the data sources as follows (Pages 6): This paper examines the causal relationship between HB and TI using monthly data from 2000M1 to 2024M01. In 2000, China joins the World Trade Organization. This not only enables China's economy to integrate into the global market faster and better, but also enables Chinese enterprises to grow rapidly in global competition, and accelerates TI (Geng et al., 2021). In addition, since 2000, rapid economic development and wealth accumulation have given rise to strong housing demand, resulting in long-term HB (Jiang et al., 2020). There are two main methods to measure TI, input method and output method. Innovation output indicators, especially the number of patent applications, can better reflect the level of TI (Griliches,1990; Cornaggia et al.,2015). So, We choose the monthly number of successful patent applications to reflect the degree of national TI (Su et al.,2022). The data are obtained from the World Intellectual Property Organization. In addition, we use the Monthly Real Estate Climate Index (MRECI), issued by China's National Burea of Statistics, to reflect the degree of HB. MRECI is an index that comprehensively reflects the operation and fluctuation of China's real estate. The MRECI selects 2000 as the base year, setting its growth level at 100. Typically, a level of 100 points is considered the most suitable, with 95 to 105 points indicating a moderate level, below 95 indicating a low level, and above 105 indicating a high level. Besides, financing constraints (FC) may affect TI, which is mainly due to the reduction of R&D investment caused by FC (Filipe et al., 2012; Alessandra et al., 2014; Po-Hsuan et al., 2014; Bronwyn et al., 2016; Khan et al., 2021; Rathnayake et al., 2022; Cecere et al., 2020; Ding et al., 2022). When banks shrink the scale of commercial credit, debt financing for house buyer will be limited, As a result, the housing demand been restricted (Favilukis et al., 2017). Moreover, tightening of bank lending standards, will deteriorate real estate developers’ liquidity, thus reducing real estate investment (Zhang et al.,2024). So, FC significant impact HB from the supply and demand sides. Therefore, banks will impact HB, TI and other economic activities by providing loans. As the interrelation between HB and TI may be influenced by FC (Zhao et al., 2016; Jia et al., 2021), we take it a control variable. The data of FC comes from official website, the People’s Bank of China. As for the indicators to measure FC, this paper selects the loan growth of financial institutions. As shown in Table 1. We have supplemented some analysis processes and added some Figure explanations, as detailed below (Pages 7-9): Table 1 presents descriptive statistics. The mean values for HB, TI, and FC are 100.620, 2229.080, and 782607.300, respectively. HB, TI, and FC exhibit considerable variation in their maximum and minimum values, indicating high volatility. The skewness of HB shows negative, while the skewness of TI and FC display positive. The kurtosis values for HB, TI, and FC are below 3, indicating a platykurtic distribution. Additionally, the Jarque-Bera test for three variables are significant at the 1% level, suggesting a non-normal distribution. Therefore, applying the traditional Granger causality test may not be appropriate. Thus, this paper employs the RB method to solve the problem of the potentially non-normal distributions in the variables. The ADF (Dickey and Fuller 1981), PP (Phillips and Perron 1988) and KPSS (Kwiatkowski et al. 1992) methods are selected to test the unit roots in HB, TI and FC, to check whether the series are stationary. The results are displayed in Table 2. The first differences of HB, TI, and FC reject the null hypothesis of a unit root at the 1% level, whereas the original series do not. This indicates that the original series achieve stationarity after first differencing. Therefore, this study employs the first differences of these three variables for analysis. Table 1. Descriptive statistics. MeanMedianMaximumMinimumStandard DeviationSkewnessKurtosisJarque-Bera HB 100.62 101.27 109.14 92.39 4.021 -0.371 2.165 15.032*** TI 2229.08 1462.00 9369.000 20.000 2280.303 0.918 2.745 41.353*** FC 782607.30 555253.10 2425048 93838.200 669314.800 0.859 2.539 38.065*** Notes: *** indicates significance at the 1% level. Table 2. Unit root test. ADFPPKPSS Original Series HB -1.682 (4) -1.263 [5] 0.814 [8]** TI-1.310 (4)-1.418 [2]0.764 [6]*** FC-1.121 (4)-1.173 [5]1.305 [4]*** First Difference HB -6.414 (4)*** -8.434 [7]*** 0.453 [4] TI-13.484 (4)***-6.561 [5]***0.372 [9] FC-15.514 (4)***-12.087 [6]***0.316 [3] Notes: The values in parentheses indicate the lag orders selected for optimisation based on the SIC criterion. The numbers in brackets represent the bandwidths chosen by the Newey-West method. ** and *** are the significance at 5% and 1% levels. This paper also utilizes the Johansen cointegration test to examine the long-term cointegration relationship between HB and TI. The results, presented in Table 3, reject the null hypothesis of no cointegration or at most one cointegration relationship at the 1% significance level. This confirms the presence of a cointegration relationship between the variables. Table 3. The Johansen cointegration test. Hypothesis Statistic value Critical value p-value None 185.116 11.473 0.000*** At most 1 68.615 2.742 0.000*** Notes: *** denotes significance at the 1% level. ............. 5.Empirical results ......... Figure 3 presents p-values for the hypothesis that HB does not Granger cause TI. The hypothesis is rejected when the values are lower than 0.1, and causalities exist. Figure 4 illustrates the direction of influence from HB to TI. When the blue line exceeds zero, there is a positive influence, and oppositely exists a negative one. By combining these two figures, we observe that during the periods 2009M09-2009M12, 2012M01-2012M03, 2019M02-2020M02, and 2023M05-2024M01, HB has a positive impact on TI. Conversely, during the periods 2003M02-2004M12 and 2014M05-2014M12, HB has a negative impact on TI. Figure 3. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI. Figure 4. The coefficients for the effect of HB on TI. 3. Include comparative analysis with other developing economies. Response: Thank you so much for your comments. We consider this is a good idea for in-depth research. We have selected another member of the BRICS nations, India, to conduct a comparative analysis. Due to space limitations, only the main Figures and brief analysis results are presented as follows (Pages 15-16). India shares many similarities with China in terms of economic development, real estate market, population size, TI and participation in globalization, making it an excellent comparative subject for analyzing China alongside other developing economies. This paper examines the bidirectional interaction between HB and TI in India, using monthly data from 2009M01 to 2024M01. We choose number of patent applications to reflect the degree of Indian TI. The data are obtained from the World Intellectual Property Organization (WIPO), We choose Real House Price Index for India (RHPII) to reflect the Indian HB. The data are obtained from Organisation for Economic Co-operation and Development(OECD). Figure 11-14 presents the empirical findings on the bidirectional relationship between HB and TI in India. By analyzing these four figures, we observe that from 2009M01 to 2024M01, HB exerts a significant positive influence on TI without any adverse effects. on the other hand, TI also actively promotes HB. These results inconsistent with findings from China. Some explanations are given: Liquidity effect channel: First, HB enhance the wealth perception of property owners, particularly among India's middle class and tech professionals. This wealth effect boosts their consumption capacity and investment confidence. Tech professionals may reinvest the increased value of their properties into entrepreneurship or tech projects, indirectly supporting TI. Second, in Indian cities with concentrated tech industries, housing prices rise rapidly, and market liquidity is high. Companies can sell properties or secure loans against them to raise funds, thereby increasing investments in TI. Crowding out effect channel: First, The Indian government and society prioritize the tech industry, viewing it as a core driver of economic growth. As a result, even amid rising housing prices, financial and policy resources continue to flow primarily into the tech sector rather than real estate. This prioritization minimizes the occurrence of the crowding-out effect. Second, India is a global hotspot for tech investments, attracting substantial foreign capital. These investments are predominantly directed toward the tech sector rather than real estate. The steady inflow of foreign capital provides ample funding for TI, offsetting any potential crowding-out effects caused by HB. Third, many tech companies, especially in software and internet sectors, operate with a light-asset model, reducing their reliance on real estate. They focus more on talent, technology, and market expansion rather than accumulating fixed assets. This approach limits the crowding-out effect of rising housing prices on tech companies. Figure 11. p-values of the rolling-window estimation examining the null that HB is not Granger cause of TI in India. Figure 12. The coefficients for the effect of HB on TI in India. Figure 13. p-values of the rolling-window estimation examining the null that TI is not Granger cause of HB in India. Figure 14. The coefficients for the effect of TI on HB in India.
Author Response File: Author Response.docx
