In an era of rapid urbanization and digital transformation, nighttime light data has emerged as a powerful proxy for measuring after-hours work. This study investigates whether the artificial intelligence (AI) pilot cities initiative, launched by the Ministry of Science and Technology in 2019, reduces overtime in Chinese government offices. Drawing on satellite-derived nighttime light data from more than 300 cities, it identifies changes in after-hours work behavior among government employees. Leveraging a quasi-natural experiment and a staggered difference-in-differences model, the findings show that AI implementation in pilot cities led to a 7.82% decline in nighttime overtime, which translates to approximately eight fewer overtime days per year. These estimates remain robust after accounting for potential alternative explanations including pandemic-induced remote work, other policy confounds, city heterogeneity, and selection bias. Overall, this research highlights the potential for AI initiatives to enhance efficiency in urban governance and reduce excessive work hours in public administration.
AbadieAGardeazabalJ (2003) The economic costs of conflict: A case study of the Basque Country. American Economic Review93(1): 113–132.
2.
AcemogluDRestrepoP (2020) The wrong kind of AI? Artificial intelligence and the future of labour demand. Cambridge Journal of Regions Economy and Society13(1): 25–35.
AokiN (2025) En route to “zero bureaucracy and invisible government”: A conjoint analysis of the effects of transforming the state-citizen interface on the favourability of public administration in the digital era. Asia Pacific Journal of Public Administration47(3): 216–240.
5.
ArkhangelskyDAtheySHirshbergDA, et al. (2021) Synthetic difference-in-differences. American Economic Review111(12): 4088–4118.
6.
BaekgaardMMadsenJK (2024) Anticipated administrative burdens: How proximity to upcoming compulsory meetings affect welfare recipients’ experiences of administrative burden. Public Administration102(2): 425–443.
7.
BroadbentBDi PaceFDrechselT, et al. (2024) The Brexit vote, productivity growth, and macroeconomic adjustments in the U.K. The Review of Economic Studies91(4): 2104–2134.
8.
CaiyunC (2023) From national capital to provincial capital: Transformation of Hangzhou’s political spaces in the Early Yuan Dynasty. Chinese Studies in History56(4): 333–361.
9.
CaprottiFCugurulloFCookM, et al. (2024) Why does urban Artificial Intelligence (AI) matter for urban studies? Developing research directions in urban AI research. Urban Geography45(5): 883–894.
10.
ChenQQiJYanG (2025) Placebo tests for difference-in-differences: A practical guide. Journal of Management World41(02): 181–220.
11.
CooperG (2023) Examining science education in ChatGPT: An exploratory study of generative artificial intelligence. Journal of Science Education and Technology32(3): 444–452.
12.
CugurulloFCaprottiFCookM, et al. (2024) The rise of AI urbanism in post-smart cities: A critical commentary on urban artificial intelligence. Urban Studies61(6): 1168–1182.
13.
CugurulloFXuY (2024) When AIs become oracles: Generative artificial intelligence, anticipatory urban governance, and the future of cities. Policy and Society44(1): 98–115.
14.
DengYSunY (2024) How does the usage of artificial intelligence affect felt administrative accountability of street-level bureaucrats? The mediating effect of perceived discretion. Public Management Review27(11): 2795–2815.
15.
EllisonMLeeSSO’RourkeKH (2024) The ends of 27 big depressions. American Economic Review114(1): 134–168.
16.
FloridaRRodríguez-PoseAStorperM (2023) Cities in a post-COVID world. Urban Studies60(8): 1509–1531.
17.
ForouharAChappleKAllenJ, et al. (2025) Assessing downtown recovery rates and determinants in North American cities after the COVID-19 pandemic. Urban Studies62(6): 1209–1231.
18.
GiestSNKlievinkB (2024) More than a digital system: How AI is changing the role of bureaucrats in different organizational contexts. Public Management Review26(2): 379–398.
GomesRCDomingosFD (2025) Herbert Simon’s legacy for public administration. Public Administration103(4): 965–978.
21.
Goodman-BaconA (2021) Difference-in-differences with variation in treatment timing. Journal of Econometrics225(2): 254–277.
22.
HainmuellerJ (2012) Entropy balancing for causal effects: A multivariate reweighting method to produce balanced samples in observational studies. Political Analysis20(1): 25–46.
23.
HenmanP (2020) Improving public services using artificial intelligence: Possibilities, pitfalls, governance. Asia Pacific Journal of Public Administration42(4): 209–221.
24.
JähnenSHelbigM (2023) The dynamics of socio-economic segregation: What role do private schools play?Urban Studies60(4): 734–751.
25.
JobinAIencaMVayenaE (2019) The global landscape of AI ethics guidelines. Nature Machine Intelligence1(9): 389–399.
26.
KrauseGAHongJH (2025) Agency leaders and organizational adaptation to administrative reform: evidence from IT reforms and the timely disbursement of unemployment insurance benefits in the American states. Journal of Public Administration Research and Theory35(4): 489–505.
27.
KrauseRMFeiockRCHawkinsCV (2016) The administrative organization of sustainability within local government. Journal of Public Administration Research and Theory26(1): 113–127.
28.
LeeHJKimMYParkSM, et al. (2020) Public service motivation and innovation in the Korean and Chinese public sectors: Exploring the role of Confucian values and social capital. International Public Management Journal23(4): 496–534.
29.
LiYWYuHB (2025) Changes in the accountability obligation, intensity, and working drive of public employees: Evidence from a survey experiment. Journal of Public Administration Research and Theory. Epub ahead of print 2September2025. DOI: https://doi.org/10.1093/jopart/muaf027
30.
LiuLWangYXuY (2024) A practical guide to counterfactual estimators for causal inference with time-series cross-sectional data. American Journal of Political Science68(1): 160–176.
31.
LongY (2024) On the sphere of applying AI in government approval service analysis according to the data of submitting in seconds and getting result in seconds services in Shenzhen City. In: Proceedings of the 25th annual international conference on digital government research, DGO 2024, 11–14 June 2024, pp. 932–940.
32.
LunkeEB (2025) Car ownership after having children: Exploring the impacts of income and public transport accessibility. Urban Studies62(5): 976–994.
33.
MacrorieRMarvinSWhileA (2021) Robotics and automation in the city: A research agenda. Urban Geography42(2): 197–217.
34.
MoonDKimH (2023) The impacts of working time flexibilization on occupational safety and health: an expert survey. Annals of Occupational and Environmental Medicine35: e20.
35.
MoynihanDHerdPHarveyH (2015) Administrative burden: Learning, psychological, and compliance costs in citizen-state interactions. Journal of Public Administration Research and Theory25(1): 43–69.
36.
MuñozM (2024) Trading nontradables: The implications of Europe’s job-posting policy. The Quarterly Journal of Economics139(1): 235–304.
37.
NoesseltN (2020) City brains and smart urbanization: Regulating ‘sharing economy’ innovation in China. Journal of Chinese Governance5(4): 546–567.
38.
OgasawaraKNakamuraYAleksicB, et al. (2011) Depression associated with alcohol intake and younger age in Japanese office workers: A case-control and a cohort study. Journal of Affective Disorders128(1–2): 33–40.
39.
ParkSLeeDSSonJ (2021) Regulatory reform in the era of new technological development: The role of organizational factors in the public sector. Regulation & Governance15(3): 894–908.
40.
QiuDLiYLLiRQ, et al. (2022) Long working hours, work-related stressors and sleep disturbances among Chinese government employees: A large population-based follow-up study. Sleep Medicine96: 79–86.
41.
RománMOWangZSunQ, et al. (2018) NASA’s Black Marble nighttime lights product suite. Remote Sensing of Environment210: 113–143.
42.
Ruvalcaba-GomezEACifuentes-FauraJ (2023) Analysis of the perception of digital government and artificial intelligence in the public sector in Jalisco, Mexico. International Review of Administrative Sciences89(4): 1203–1222.
43.
SampaioIGBFontesSDBAndradeED, et al. (2024) Enhancing transparency through explainable artificial intelligence: An exploratory analysis on a collusion and corruption scenario in digital government. In: Proceedings of the 25th annual international conference on digital government research, DGO 2024, 11–14 June 2024, pp. 52–62.
44.
SchiffKJSchiffDSAdamsIT, et al. (2025) Institutional factors driving citizen perceptions of AI in government: Evidence from a survey experiment on policing. Public Administration Review85(2): 451–467.
45.
SnidalMMaaouiMHaupertT (2022) Exodus in the American metropolis: Predicting Black population decline in Chicago neighbourhoods. Urban Studies59(15): 3114–3131.
46.
StelbotskyBParkS (2024) Can chatbots represent? Examining the potential for symbolic representation in automated service agents. Public Management Review28(2): 477–497. https://doi.org/10.1080/14719037.2024.2443064
47.
TakahashiKKodamaNAritaK, et al. (2025) Has Japan’s work style reform had the intended effect?Applied Economics57(21): 2757–2780.
48.
TaylorJ (2018) Working extra hours in the Australian Public Service: Organizational drivers and consequences. Review of Public Personnel Administration38(2): 193–217.
49.
TingSKDolleryBVillanoR (2014) Administrative scale economies in local government: An empirical analysis of Sabah municipalities, 2000 to 2009. Urban Studies51(13): 2899–2915.
50.
WeiglLRothTAmardA, et al. (2024) When public values and user-centricity in e-government collide – A systematic review. Government Information Quarterly41(3): 101956.
51.
WhileAHMarvinSKovacicM (2021) Urban robotic experimentation: San Francisco, Tokyo and Dubai. Urban Studies58(4): 769–786.
52.
WuJ (2025a) Digital Jevons paradox in urban data center energy systems. Nature Cities2(8): 677.
53.
WuJ (2025b) Does open government data alleviate the burden on government employees? Evidence from a quasi-experiment in China. Review of Policy Research. Epub ahead of print 14November2025. DOI: https://doi.org/10.1111/ropr.70068
54.
WuJ (2025c) Synergy or conflict: Can e-government policy mixes drive environmental improvements? Evidence from China’s multi-pilot quasi-experiment. Telecommunications Policy49(10): 103069.
55.
XieJWangLLiH, et al. (2025) Measuring overwork in China using daily high-resolution nighttime satellite data. Scientific Data12(1): 417.
56.
XuanYWuK (2023) Overtime work and labor income share: Empirical evidence based on satellite night light. Journal of World Economy46(10): 217–240.
57.
YuNNJinYWangZH (2025) Political cycles of infrastructure investment: How career concerns shape strategic behavior of city leaders?Journal of Chinese Political Science30(1): 31–57.
