The changes in job-housing balance during the Covid-19 period in China

Abstract

By using three continuous years of national-scale cellphone signaling data from Jan. 2019 to Dec. 2021, this study adds fresh evidence for job-housing balance changes at the Quxian level during the COVID-19 period in China. The findings show that according to the resident-balance index and worker-balance index, the job-housing balance jumped when the number of COVID-19 confirmed cases reached its peak in February 2020, with an average of 94.4 % which is the highest level during these three years. The study also found that the Quxian-level job-housing balance has generally improved steadily in the two years of the pandemic. In addition, the results highlighted the huge gaps between females and males in the job-housing balance, but the gender disparities in job-housing balance were reduced to a minimum during the pandemic lockdown. In addition, by comparison analysis of the changes in resident-balance index and worker-balance index during this unprecedented crisis, this study found that for Quxians with high economic vitality, worker-balance index increased greater than resident-balance index, but for Quxians with low economic vitality, the reverse happened. Our findings provide a better understanding of the job-housing relationship during public health crises that can support the urban management in the future policymaking.

1. Introduction

Job-housing balance refers to a relatively balanced or equal relationship between the number of jobs and the housing within a given geographical unit. Considerable evidence suggests that improving the job-housing balance will reduce traffic congestion and environmental pollution (Zhao, Lü, & Linden, 2009), improve urban space and land use efficiency, promote building a healthy and livable city (Ta, Chai, Zhang, & Sun, 2017), and provide other lasting benefits for sustainable city development.

In recent years, the COVID-19 pandemic and the resulting economic fallout have brought significant changes to people's employment and choice of residence location, and thus the job-housing balance. For example, the COVID-19 pandemic has brought many job losses (Hensvik, Le Barbanchon, & Rathelot, 2021) and great disturbance to the housing and rental market (Sequera et al., 2022). In particular, the COVID-19 pandemic has brought great disruption to people who commute long distances and across districts (Harrington & Hadjiconstantinou, 2022). Some people changed to working from home to avoid the risk of being infected with COVID-19 (Galanti, Guidetti, Mazzei, Zappalà, & Toscano, 2021). Some people relocated from communities with high population densities to others with low population densities (Liu & Su, 2021). There were also social inequities in relation to housing and jobs during the crisis (Bushman & Mehdipanah, 2022). For example, some renters were evicted or fell behind on rent (Benfer et al., 2021). To sum up, it is widely believed that changes in the job-housing balance occurred during the COVID-19 pandemic.

However, more evidence from different contexts is needed for this. This study adds fresh evidence by examining job-housing balance changes during the COVID-19 period in China. By using three continuous years of national-scale cellphone signaling data from Jan. 2019 to Dec. 2021, we calculated three indicators to assess the job-housing balance: the job-housing ratio, resident-balance index, and worker-balance index (for details see Section 2) at the Quxian level (urban district or rural county), which is an administrative unit that is lower than a municipal government but higher than a rural township government. The Quxian was the main administrative boundary to control human mobility during COVID-19 in China (the current number of Quxians in China is 2844).

2. Methodology and data

2.1. Indicators for assessing the job-housing relationship

Scholars have long focused on the job-housing relationship, and present many indicators for measuring the job-housing balance from different perspectives (Wang, Zhou, Rong, Liu, & Wang, 2022; Zhang, He, & Zhao, 2018; Zhao, Lü, & De Roo, 2011). Among them, the Job-housing ratio, Resident-balance index, and Worker-balance index are the commonly used indicators to measure the job-housing relationship in a given geographical unit.

• (a)Job-Housing Ratio (JHR) refers to the employment distribution relative to the housing distribution within a given geographic area. When these distributions are approximately equal, this area is considered job-housing balanced. The formula is expressed as follows:

$${\mathrm{JHR}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{J}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{H}}_{\mathrm{i},\mathrm{t}}}$$ (1)

where ${\mathrm{JHR}}_{\mathrm{i},\mathrm{t}}$ represents the job-housing ratio of districts and counties i in month t. ${\mathrm{J}}_{\mathrm{i},\mathrm{t}}$ represents the number of workers in district and county i in month t. ${\mathrm{H}}_{\mathrm{i},\mathrm{t}}$ represents the number of residences in Quxian i in month t.

• (b)Resident-balance index (RBI) refers to the percentages of people who are living and working in a given geographic area to the total number of residents in this geographic area. The closer the index values are to 100 %, the higher the resident-based job-housing balance. The formula is expressed as follows:

$${\mathrm{RBI}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{M}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{H}}_{\mathrm{i},\mathrm{t}}}\times 100\%$$ (2)

where ${\mathrm{RBI}}_{\mathrm{i},\mathrm{t}}$ represents the worker-balance index, ${\mathrm{M}}_{\mathrm{i},\mathrm{t}}$ represents the number of people living and working in Quxian i in month t, and ${\mathrm{H}}_{\mathrm{i},\mathrm{t}}$ represents the total number of residents in Quxian i in month t.

• (c)Worker-balance index (WBI) refers to the percentages of the number of people who are living and working in a given geographic area to the total number of workers in this geographic area. The closer the index values are to 100 %, the higher the worker-based job-housing balance. The formula is expressed as follows:

$${\mathrm{WBI}}_{\mathrm{i},\mathrm{t}}=\frac{{\mathrm{M}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{J}}_{\mathrm{i},\mathrm{t}}}\times 100\%$$ (3)

where ${\mathrm{WBI}}_{\mathrm{i},\mathrm{t}}$ represents the worker-balance index, ${\mathrm{M}}_{\mathrm{i},\mathrm{t}}$ represents the number of people living and working in Quxian i in month t, and ${\mathrm{J}}_{\mathrm{i},\mathrm{t}}$ represents the total number of workers in Quxian i in month t.

2.2. Data sources

To enable a long-term, comprehensive, and consistent investigation of the Quxian-level changes in the job-housing relationship before, during, and after the COVID-19 pandemic in the whole country of China, we used cellphone signaling data from Jan. 2019 to Dec. 2021 to measure the job-housing relationship, covering one year before the pandemic and two years of the pandemic. The data were provided by the China Unicom company, which had a market share of 19 % in China in 2021. The historical daily number of confirmed cases came from the official website of the World Health Organization. The night light data used in our study were corrected by Zhong, Qingwu, and Guie (Zhong, Qingwu, & Guie, 2022) and are openly available. By using the Zonal Statistics tool in ArcGIS, the average nighttime light brightness value in each Quxian was calculated.

3. Analysis

3.1. Overall changes in job-housing balance in China from 2019 to 2021

Based on the job-housing ratio indicator, the calculations show that on average, the monthly Quxian-level job-housing ratio in China was in the range of 1.00 to 1.02 from Jan. 2019 to Dec. 2021, which means the job-housing ratio remained statistically unchanged during the COVID-19 period (see the purple line in Fig. 1a). However, the resident-balance index (see the green line in Fig. 1a) and worker-balance index (see the blue line in Fig. 1a) improved following the outbreak of the pandemic, and reached a peak with an average of 94.4 % during Feb. 2020, when confirmed COVID-19 cases peaked (see the orange dotted line in Fig. 1a). It might be owing to the increases in the rates of staying at home, remote work and working at home, and lockdown-caused mobility restrictions after the pandemic (Chen, Zhang, & Zhou, 2023; Liu et al., 2021). Additionally, after confirmed cases fell sharply in Mar. 2020, the resident-balance index and worker-balance index dropped to slightly higher than their pre-pandemic levels, and they remained there for the subsequent two years.

Fig. 1.

The overall changes in job-housing balance in China before and after the pandemic. (a) The monthly changes in job-housing balance in China from 2019 to 2021. The gender gaps in (b) job-housing ratio, (c) resident-balance index, and (d) worker-balance index over time from 2019 to 2021. Note: The month of the Chinese New Year, namely Feb. 2019, Jan. 2020, and Feb. 2021 is not shown in these figures, because most people return to their hometowns during these periods, leading to low accuracy in the job and housing location identification from the cellphone data.

Analysis of gender disparities data in the job-housing relationship showed that the job-housing ratio of males was higher than that of females (see Fig. 1b). In addition, females had a relatively higher balance than males in the resident-balance index (see Fig. 1c) and worker-balance index (see Fig. 1d) in general, which might be explained by the average commuting radius of males being larger than that of females. Interestingly, we also found that the gender gaps in the job-housing balance reduced to a minimum (see Fig. 1b, c, and d) during Feb. 2020, which might be the result of males and females facing similar situations (e.g. staying at home, working from home) during the pandemic lockdown.

3.2. The spatial heterogeneity of changes in job-housing balance during the pandemic

In view of the Job-housing balance was mutated during Feb. 2020, this section compares the spatial heterogeneity of changes in the job-housing balance between the pre-pandemic period in Dec. 2019 and the pandemic in Feb. 2020 for the 2728 Quxian spatial units in China. The job-housing ratio of most Quxians (over 95 %) was in the range of 0.9 to 1.1 before and during the pandemic, which means that from the job-housing ratio lens, Quxian-level areas were acting as self-sufficient spatial units in general (see the purple line in Fig. 1a). However, different periods (before and during the pandemic) had different balance types (Fig. 2a, b). Before the pandemic, Quxians with ratios below and above 1.00 in the job-housing ratio accounted for 39.5 % and 60.5 %, respectively, compared with 59.6 for below 1.00 % and 40.2 % for above 1.00 during the pandemic.

Fig. 2.

The spatial heterogeneity of job-housing balance changes before (Dec. 2019) and during (Feb. 2020) the pandemic. The spatial distribution of job-housing ratio/resident-balance index/worker-balance index in Dec. 2019 (a/d/g), in Feb. 2020 (b/e/h), and the range of changes between these two periods (c/f/i).

For the resident-balance index and worker-balance index, more than half of the Quxians reached a balance of more than 95 % during the pandemic. Specifically, 12.9 % of the Quxians had the highest resident-balance index (over 95 %) (Fig. 2d) before the pandemic, and this reached 58.4 % during the pandemic (Fig. 2e). For the worker-balance index, there was a similar trend, increasing from 11.1 % of the Quxians having the highest balance index before the pandemic (Fig. 2d) to 58.5 % during the pandemic (Fig. 2e).

The majority of Quxians increased their resident-balance index scores during the pandemic, whereas in areas located along the eastern coastal areas, the job-housing balance reduced during the pandemic (see the light blue plaque in the lower right corner of Fig. 2f), which means that workers in these areas experienced larger job losses than other areas during the crisis. This is partly because in the eastern coastal areas with the largest number of migrant workers in China, the COVID-19 pandemic was a whole country outbreak during the Chinese New Year travel rush, leading to many migrant workers not returning to work from their hometowns.

3.3. Relationship between job-housing balance change and local economic activity

Using the nighttime light data (Fig. 3a) detected by satellites, which has been widely used as a proxy for the economic activity of a geographic spatial unit (Gibson, Olivia, & Boe-Gibson, 2020), we revealed how the job-housing balance changes differ with local economic activity. As shown in Fig. 3c, d, e, and f, Quxians with higher night light brightness have lower ratios in the resident-balance index and worker-balance index, which is in line with the previous findings that job-housing separation and imbalance are significant in economically developed areas (Blumenberg & King, 2021).

Fig. 3.

Relationship between night light brightness and job-housing balance. (a) Spatial distribution of nighttime light brightness value in Dec. 2019. (b) Relationship between resident-balance index, worker-balance index, and the night light brightness value. Quadrant analysis of the resident-balance index/worker-balance index in nighttime light brightness of each Quxian in Dec. 2019 (c/f), in Feb. 2020 (d/g), and the range of changes between these two periods (e/h).

Fig. 3e shows the resident-balance index of most Quxians increased during the pandemic, which means that some people changed their workplace to their Quxian of residence during the pandemic. Similarly, Fig. 3h shows that the worker-balance index of most Quxians increased during the pandemic, which indicates that some people moved to live in proximity to their jobs during the pandemic. Safety concerns and factory closures due to the pandemic may be the factors driving people to change their workplaces or residences.

A comparative analysis between the resident-balance index and worker-balance index change degree (see Fig. 3b) shows that Quxians with the highest night light brightness (such as the Huangpu district of Shanghai, Dongcheng district of Beijing, Heping district of Tianjin, and Huli district of Xiamen) have worker-balance index increases larger than resident-balance index increases (the circles with greater night lights brightness are more likely to lie above the 45-degree line in Fig. 3b), indicating that people in economically active areas are more likely to move to live in proximity to their workplace during the pandemic. This means that the place of residence is more easily changed and that the stickiness of the job is stronger than residence stickiness for economically developed areas. Correspondingly, in the smaller Quxians, the resident-balance index increased more than the worker-balance index, as migrant workers who originally went out to work before the pandemic may have worked in their hometowns during the pandemic.

4. Conclusion

By using the three continuous years of cellphone signaling data from Jan. 2019 to Dec. 2021, we have provided a comprehensive understanding of the changes in the job-housing relationship at the Quxian-level in China during the COVID-19 pandemic. The key points of the findings are as follows: first, according to the resident-balance index and worker-balance index indicators, the job-housing balance jumped in Feb. 2020 when the number of COVID-19 confirmed cases reached its peak, and it stood at 94.4 % on average in Feb. 2020 which is the highest level during these three years. Also, we found that the Quxian-level job-housing balance improved steadily in the two years of the pandemic in general, particularly for Quxians with high economic vitality, in which it increased more than in other areas. In addition, when gender disparities were considered, the results highlighted the significant gaps between females and males in the job-housing balance, but the gender disparities in job-housing balance were reduced to a minimum during Feb. 2020. Moreover, through the comparative analysis of the changes in the resident-balance index and the worker-balance index during this unprecedented crisis, we found that for Quxians with high economic vitality, worker-balance index increased more than resident-balance index, but for Quxians with low economic vitality, the reverse happened. It seems to indicate that the stickiness of the job is stronger/smaller than residence stickiness in economically active/inactive areas.

Concerning the implications for future policymaking and studies, first, this study suggests that the job-housing balance has improved and its gender gaps have lessened due to the pandemic. In this sense, we might infer that stronger preferences for working from home (Mayer & Boston, 2022) have likely contributed to the improvement in job-housing balance and increased gender fairness in job-housing balance, particularly for developed countries and computer-based employees (Ford et al., 2021). Second, as government policies have been seen as the main factors influencing the local job-housing balance (Zhao et al., 2011), how the job-housing balance changes differed with local COVID-19 conditions and containment policies needs further study. Third, as the job-housing housing relationship relies strongly on the observation of geographic units, a robustness test of the results could be undertaken by a comparative analysis of different spatial-scale units to explore the changes in job-housing balance during the COVID-19 period.

Funding acknowledgement

This study was supported by National Natural Science Foundation of China (Grant numbers: 41925003, 42130402), and Guangdong Provincial Natural Science Foundation (Grant number: 2022A1515010696).

CRediT authorship contribution statement

Zhenjun Li: Supervision, Writing – review & editing, Data curation. Pengjun Zhao: Conceptualization, Funding acquisition. Ling Yu: Methodology, Writing – original draft, Visualization. Xiaodong Hai: Conceptualization, Validation. Yongheng Feng: Conceptualization, Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.