1 Introduction

Tracking the dynamics of the labour market is of utmost importance for both understanding the health of the economy as well as for understanding the welfare implications arising out of the changes in the labour market. It is, therefore, not a surprise that one of the most watched statistics of an economy is the unemployment rate in the country. However, just looking at the aggregate numbers of unemployed may not be enough given the great degree of heterogeneity in the labour market. There is a need to look at details of people in the labour force by— income class, gender, educational attainments, skill levels, and such. Moreover, it is essential to determine what changes are occurring across these groups. To do this, it is necessary to track labour flows frequently over time. The recent Covid-19 crisis and its huge impact on the labour market and the economy has brought this into focus even more. However, this has been difficult for an emerging economy like India. Given the sheer size of the population of India, the demographic structure, the heterogeneity, and the development stage of the country—it is probably more urgent to track the labour market dynamics as frequently as possible as it will have implications not only for India, but for the global economy also. So far, it has been challenging to do that; but with the availability of new data from Centre for Monitoring Indian Economy (CMIE), it is now possible to track the labour market in India three times a year. In this paper, we use the CMIE-Consumer Pyramids Household Survey (CPHS) data to construct a panel and track that every four months. Using that, we create labour flow charts and transition matrices. We then use that to study the impact of Covid-19. Further, using the same data, we do an event study analysis to find if and how different the impact has been across various heterogeneous groups by education levels.

To study labour transitions, a standard technique is to construct Labour Flow Charts and Transition Matrices. In a developed country like the USA, for example, this is done by the Bureau of Labor Statistics every month. The labour flow charts track and map the flow of eligible workers across different employment categories over time (Bleakley et al. 1999). This is done continuously at a regular frequency. The transition matrix uses these labour flow charts and creates a rectangular array describing the probabilities of moving from one state to another in a dynamic system across categories. Each cell in a row of a particular transition matrix represents the probability of moving from the current state to another state during the reference period considered. Each row of the matrix adds to one individually (Bronson & Costa 2009). So far, it was not possible to create these labour flow charts or transition matrices for India due to lack of data at regular intervals. We use a pan-India comprehensive longitudinal panel dataset to generate Labour Flow Charts and the corresponding Transition Matrices for India from January 2019 to December 2021. To the best of our knowledge, this is the first time these have been created for India. This opens up the possibility of studying a large number of questions in the context of the Indian economy, which was not possible so far.

In this paper, in particular, we look at the differential impact of Covid-19 on diverse groups of people in the working-age population. The pandemic engendered a once-in-a-century global crisis that resulted in unprecedented recessions across the globe, resultant job losses and an unprecedented rise in unemployment across economies (Global Economic Prospects June 2021, World Bank 2021). India has not been an exception. The impact of the crisis on the Indian economy has been devastating. The GDP of India nosedived by 23.9% during the first quarter of FY-21 (NSO, Government of India). On the policy front, the government of India reacted early and resorted to one of the world’s most stringent lockdowns to minimize the loss of lives and livelihoods with the onslaught of the pandemic (Anand & Miglani 2020, Chatterjee et al. 2020, Oxford Covid-19 Government Response Stringency Index, Oxford University, Gössling et al. 2020). The Indian government announced a nationwide lockdown beginning March 24th, 2020 (Economic Survey 2020). It was extended twice and continued till the end of May, after which there was some relaxation of lockdown restrictions. The imposition of lockdown and market closure affected the job market status quo, thus leading to mass-scale employment losses. Many papers in the literature comment on the impact of the crisis on the employment scenario globally, including the impact felt in the Indian labour market (Dhingra & Machin 2020; Bertrand et al. 2020; Bhalotia et al. 2020; Deshpande 2020, Lee et al. 2020; Abraham et al. 2021). However, none of the papers are able to generate the labour flows or calculate the transition probabilities, which we do.

Our work augments the literature on the impact of the Covid-19 crisis on the Indian labour market in a meaningful way. We are able to find out the transitions between distinct categories during and after the Covid-19 related restrictions. When the lockdown was imposed in March 2020, we observe a sharp and sudden fall in Employed-to-Employed transition. This is mirrored in an increase in Employed-to-Out of Labour Force transition. However, an important observation is a sharp rise in Employed-to-Data Not Available transition. We also observe how, after the restrictions were lifted, the levels gradually returned to the pre-pandemic levels. The rich dataset also allows us to explore diverse types of heterogeneity. We choose to study how the impact differs across people with various education levels. It is expected that the impact will be different, if not for anything else, because the ability to work from home would vary across education groups. Not surprisingly, all sections were impacted, but the drop in employment was more precipitous for those with primary or secondary education than graduates or post-graduates. Interestingly, the transition to unemployment seems to have been increasing across all categories even before the pandemic hit. While there has been substantial recovery after the restrictions were lifted, the recovery has slowed down for those with primary or secondary education. In this, we do not restrict ourselves to looking at only employment and unemployment; we further look at the type of employment—Full-time versus Part-time, and another category, Data Not Available. The last category includes those who could not be contacted during the survey period of the particular wave. A substantial increase in that during the lockdown period is striking. This probably captures the migration from urban areas to rural areas that ensued during that period (Gerard et al. 2020). We further extend our analysis to look at another dimension of heterogeneity—gender. As expected, the level of employment for men and women differs substantially. We find that there is a higher likelihood of full-time employment for men compared to women for all education categories for all periods. Both men and women, irrespective of their education level, were impacted due to Covid-19, but the impact differed across groups with different education attainment.

The rest of the paper is structured as follows: the next section describes the data structure. Section 3 constructs the labour flow charts and transition matrices. Section 4 discusses the Covid-19 pandemic and labour market implications. An econometric event study analysis is followed in Sect. 5, which assesses the likelihood coefficients for a worker having a full-time job and compares the same ex-ante, during, and ex-post the imposition of Covid-19 induced market restrictions. Here, we show the improved likelihood of having full-time employment for workers with no formal education during the pandemic phase. Finally, Sect. 6 summarizes all the significant findings.

2 Data

Labour force statistics and the corresponding Labor Flow Charts in the USA are prepared using findings from Current Population Survey (CPS) (Abowd & Zellner 1985). CPS is a monthly survey of households conducted by the Bureau of Census for the Bureau of Labor Statistics. It provides a comprehensive body of data on the labour force, employment, unemployment, persons not in the labour force, hours of work, earnings, and other demographic and labour force characteristics. Regular, frequent, and real-time labour market data is crucial to capture the labour market dynamics in the economy. In context of Indian economy, the Periodic Labour Force Survey (PLFS) is the official data source for labour market statistics. Starting 2017, PLFS replaced the Labour Bureau’s Employment and Unemployment surveys (EUS) which had replaced the earlier quinquennial NSS-EUS surveys. First report of PLFS was published in June 2019 for the period 2017–18. Since 2017 there have been three more rounds of PLFS in 2018–19, 2019–20 and 2020–21. However, generating Labour Flow Charts require real-time data points at more frequent intervals. Moreover, we need a panel to track the transitions across different categories. Thus, PLFS does not serve the purpose for our analysis.

To generate Labour Flow Charts and the associated Transition Matrices for India, we use longitudinal panel data from the CMIE’s CPHS which is of higher frequency and is published regularly. It is administered on a panel of over 1,70,000 households across the whole of India thrice a year. The survey is typically conducted face-to-face but owing to the Covid-19 lockdown in India after the third week of March, the face-to-face interview format was replaced with a telephonic one. The response rate during the lockdown was slightly over 60%, compared to over 95% before the lockdown (Vyas 2020). This is the only nationally representative survey that was carried out prior to the lockdown, during the lockdown (partially, as phone survey), as well as in its aftermath (Abraham & Srivastava 2022). The entire sample is surveyed for a period of four months, called a wave. We use nine consecutive CPHS waves for our analysis: Wave 16 (January–April 2019), Wave 17 (May–August 2019), Wave 18 (September–December 2019), Wave 19 (January–April 2020), Wave 20 (May–August 2020), Wave 21 (September–December 2020), Wave 22 (January–April 2021), Wave 23 (May–August 2021), Wave 24 (September–December 2021).Footnote 1

For constructing our charts and matrices, we create a panel of 8,16,549 individuals and track them in all the subsequent waves over 36 months from January 2019 to December 2021. The government of India announced the nationwide lockdown in the month of March 2020. This corresponds to the 19th CPHS wave, i.e., January–April 2019. We group the nine waves into the following three broad categories: Pre-Covid CPHS Waves (16, 17 & 18), During Restrictions CPHS Waves (19, 20), and Post Relaxations CPHS Waves (21, 22, 23 & 24) to conduct our analysis. We use the individual level responses on Type of Employment and Employment Status variables of CPHS to get the employment category of the respondents. All the respondents are categorized into six broad categories, namely Full-time Employment, Part-time Employment, Out of Labour Force, Unemployed, Data Not Available and Others. The systemic organization of labour market comprises full-time employment and part-time employment contracts. Organizations define full-time hours differently and outline their definition within the employment contract or in the company's policies. However, each organization sets a standard for what full-time hour means to them. On the contrary, part-time contracts have fewer working hours and differ from full-time ones regarding job structure, pay and benefits, tax liability, and job security. We do not juxtapose the two kinds of contracts here to get into the comparison metric. Our paper sticks to the organizational definitions of full-time and part-time contracts adopted by the source data body. Adopting the CMIE framework, we recognize an unemployed worker as either Unemployed, willing, and looking for a job or Unemployed, willing but not looking for a job. On the other hand, a respondent who is Unemployed, not willing, and not looking for a job is considered Out of Labour Force. We stick to the CPHS nomenclature of Data Not Available for the associated individuals of the panel who could not be traced in a particular wave. The last category is called Others, comprising the remaining respondents with their Employment Status as employed, but Type of Employment as not applicable.

Table 1 provides the average share of each category under consideration for the entire period of our study. Out of the total panel size of 8,16,549 individuals, around 32% of the total workers fall under Out of Labour Force category during the Pre-Covid CPHS waves. However, during the Covid Restrictions waves, the maximum share of workers belongs to the Data Not Available category. More number of respondents not being traced during course of the pandemic hints at the large-scale labour force reverse migration which was observed in India due to the pandemic induced restrictions and market closure. This observation is also supported by the fall in the share of Full-time and Part-time Employment observed during this period. The share of workers belonging to Employed category, including both Full-time and Part-time Contracts almost halved during this period (a fall of approx. 50%) as compared to the Pre-Covid waves. We observe a fall in the average share of Out of Labour Force category from around 32% to as low as 21% during this period. With lifting of market restrictions and opening of economic activities, our data exhibits a gradual returning of the share of employment and unemployment statistics to pre-pandemic levels. It may be noted here that the share of Part-time Employment has consistently remained very less throughout the entire period of study in context of the Indian labour market.

Table 1 Data Summary
Full size table

3 Labour Flow Charts and Transition Matrices

Using the dataset specified, Labour Flow Charts and Transition Matrices are developed for our study period. Table 2 presents our Labour Flow Charts. Rows in the table depict the transition flows from the category under consideration to all the six categories. Columns placed adjacent to each other show the percentage transition for all the consecutive CPHS waves. For each category, the charts give the percentage share of labour force transition to every other category. Starting from any category, say, Full-time employment, the cumulative percentage share of transition into the same category and all other categories combined sums to 100 per cent. In the similar manner, the transition flows are recorded from Part-time employment, Out of Labour Force category and Data Not Available category here. The columns highlighted in shades give the transition trajectories observed during the period of Covid-19 induced nationwide lockdown and the associated market closure.

Table 2 Labour Flow Charts
Full size table

Full-time transition trend charts in Fig. 1 and Part-time transition trend charts in Fig. 2 represent the transition trajectories from full-time and part-time employment to the other five categories for consecutive waves and the associated months for the study time frame.

Fig. 1
figure 1

Full-time Transition Trend Charts

Full size image
Fig. 2
figure 2

Part-time Transition Trend Charts

Full size image

It may be noted here that from the beginning of Covid-19 lockdown and the subsequent market restrictions, we observe a sharp deviation from the usual trajectories observed during the earlier phase.

Interpreting Table 1 here for the transitions from Full-time employment, we observe around 82% of the total full-time workers retaining their employment status between the first transition bracket of Wave 16 to Wave 17, i.e., January-April 2019 to May–August 2019. A similar trend of around 82% retainment is observed during the next transition from Wave 17 to Wave 18, i.e., May–August 2019 to September–December 2019. We observe declining transition trend for the following two transitions, starting Wave 18 till Wave 20, with retainment falling as low as around 37% during Wave 19 to Wave 20 transition, i.e., January–April 2020 to May–August 2020. This is followed by a gradual recovery starting Wave 21 onwards. The latest transition data available for Wave 23 to Wave 24 corresponding to May–August 2021 to September–December 2021 show a figure of around 80% for retaining full-time contracts between the consecutive waves. Quadrant 1, Fig. 1 draws the V-Shaped trajectory observed for Full-time to Full-time transitions as discussed. All other transition trends can be analyzed in similar pattern from the Labour Flow Charts. It must be noted over here that the labour flows from all categories to Data Not Available category peaked up significantly during the course of the pandemic as shown in the Labour Flow and the Transition Trend Charts above.

Table 3 presents the Transition Matrices for our study period. Expressed as the probability of moving from one state to another in a dynamic setup, the horizontal summation of total probability of transition from a particular category to all other categories always sum to unity. Interpreting Table 3 here, each cell along a horizontal row represents the probability of transition from source category to all other categories over the transition phase. For instance, in the first cell, probability value of 0.8168 gives the probability of retaining a full-time employment contract between Wave 16 and Wave 17 transition, i.e., between January–April 2019 to May–August 2019. Similarly, the probability of moving from a full-time contract to a part-time contract during the same period stands at 0.0011. The highlighted rows represent the transition probabilities observed during the course of the pandemic induced lockdown.

Table 3 Transition Matrices (Probabilities)
Full size table

What we observe is that a person who has a full-time job is most likely to keep that. The probability of transitioning from a full time to a full-time job is high. However, the probability of getting a full-time job from any other status is very low. It is interesting to note that there is a significant fall in the probability of keeping a full-time job during the transition September to December 2019 to January to April 2020 compared to the previous periods. Note that the January to April 2020 wave contains only seven days of the lockdown and much of the survey must have been completed before that, so it is likely that other forces are also at play that reduce the probability of keeping a full-time job. Expectedly, the drop is even more during the next wave which is January–April 2020 to May–August 2020 as the May–August 2020 wave covers most of the lockdown period. Post lifting of the restrictions, the probability has gone up again close to the pre-pandemic level, but has not surpassed it.

To understand these numbers at a more granular level, we take into account the heterogeneity across different groups which differ in educational attainment. We include Education variable of CPHS corresponding to our study period. We create the following five mutually exclusive categories of education for all the individuals: No Formal Education, Primary Education, Secondary Education, Graduate and Postgraduate. Figures 3 and 4 show the transition trend from full-time employment to all the other categories along with the absolute count of full-time workers with different levels of education during a particular transition period considered. The stacked vertical columns give the labour flow transitions between the different categories under consideration.

Fig. 3
figure 3

Full-time Transition Trend Chart with No Formal Education

Full size image
Fig. 4
figure 4

Full-time Transition Trend Chart at Different Levels of Education

Full size image

From Fig. 3 we observe that the total count of full-time workers with no formal education seems to be volatile. There is a gradual decline in the absolute count from the start of the period of this study. This could be a result of gradual slowing down of Indian economy prior to 2018 (Ahmad et al. 2018). We, however, do not get into the details of that in this paper as our focus is to look at the impact of Covid-19. The decline gets steeper, reaching its minimum during the transition period January–April 2020 to May–August 2020. With the pandemic induced lockdown and market restrictions well in place, we observe a greater number of workers without any formal education getting into the labour market. However, the trend begins to fall again with the gradual opening of economy and easing of restrictions. We do not observe a similar pattern in the trend for the educated workforce in Fig. 4. As expected, there is a decline in the absolute count of educated workers in the labour market during the course of the pandemic. With easing of restrictions, more workers entered the market. The trend of Primary Education, Secondary Education, Graduate and Postgraduate education does not differ much in observation. It may be noted over here that there has been some level of flattening of the slope of trend charts for all categories of educated workers over the later stage of opening of the economy, primarily starting from January–April 2021 to May–August 2021 transition. This is a more recent happening. Our paper does not get into the details of exploring this particular observation to keep focus on the impact of Covid-19.

To further investigate this increase in the absolute count of workers without any formal education during the pandemic, in Sect. 5 we carry out an Event Study Analysis exploring the likelihood of a worker having a full-time employment contract at different levels of education and the influence of pandemic on the same.

4 Covid-19 Pandemic and Labour Market Implications

During the last two years, India and the rest of the world faced the blitz of the Covid-19 pandemic. India and other emerging economies on account of their large population density and more contact sensitive nature of the major economic activities, faced the heat of infection more than the advanced economies. The Government of India resorted to one of the world’s most stringent lockdowns to minimize the loss of lives and livelihoods with the onslaught of the pandemic (Oxford Covid-19 Government Response Stringency Index, Oxford University, Gössling et al. 2020). The Indian Government announced a nationwide lockdown beginning March 24th, 2020. It was extended twice and continued till the end of May, after which there was some relaxation of lockdown restrictions. Figure 5 shows the relative position of India and the rest of the world on the strictness parameter of Government’s response to Covid-19 pandemic. India can be seen placed among the countries with very rigorous Covid-19 restriction policies.

Fig. 5
figure 5

Source: Our World in Data, Global Change Data Lab, U.K.

Covid-19 Stringency Index, Oxford University.

Full size image

The imposition of Covid-19 mitigation strategies triggered the temporary and permanent closure of lakhs of businesses and other economic activities, resulting in large-scale job losses and reverse labour migration. In light of the detrimental effects of Covid-19 on the global and the national economy, our paper uses the labour flow charts and the transition matrices that we have created to investigate the impact of the pandemic on labour market.

To emphasize the CPHS waves corresponding to the pandemic period, in Tables 1 and 2, we have highlighted Wave 19 to Wave 20 and Wave 20 to Wave 21, i.e., January–April 2020 to May–August 2020 and May–August 2020 to September–December 2020 transition as the one affected by the imposition of lockdown. We look at the transition trajectories of various category pair of transition, Fig. 1 shows the key labour market implications in the form of a sharp decline in Employment-to-Employment transitions for both Full-time and Part-time contracts. The transition from Employment to Out of Labour Force also doubles for Full-time contracts and increases by more than 50% for Part-time contracts compared to the average value during the pre-Covid period. Looking at transitions from Employment-to-Unemployment, we observe a sharp and significant increase in the transition values for Full-time contracts only. There is no change observed in the transition from Part-time to Unemployment category due to the pandemic disturbances. We do not observe a very sharp movement in the transition from Full-time to Part-time contracts and vice versa. Another significant observation here is the extreme volatility in the transition trajectory for Employment to Data Not Available transitions during the restrictions. The transition from Full-time employment to Data Not Available went up from an average of around 11% prior to the pandemic to as high as 54% during the January–April 2020 to May–August 2020 transition period. The transition figure from Part-time employment to Data Not Available increased from around 13% to around 52% during this period. With the phased opening of the economy post the month of June 2020, we observe the transition figures returning to the pre-Covid figures. This is a robust observation here, hinting on the large-scale job losses and reverse labour migration observed in India during lockdown. As already explained in Sect. 3, we observe a strikingly different trend in the absolute count of workers without any formal education during the pandemic. While the educated workers at distinct levels of education exhibit a fall in their absolute count, those without any formal education entered the labour market during the lockdown period. The next section explores this in detail.

5 Event Study Analysis

Covid-19 has been a once in a century crisis for the entire world. We expect the pandemic to have a significant impact on labour market transition trajectories. We use an event study analysis to understand the dynamics of the labour market outcomes for different groups before and after the event and assess its impact (Pamela Peterson 1989). Following an event study approach (Bussolo et al. 2021; World Bank Group 2021), our paper explores the likelihood of a worker getting into a full-time job with education and gender being the primary independent variables. We assess the impact of the pandemic on the prospects of entering a full-time contract at distinct levels of education for both genders separately. We conduct our analysis to comment on the likelihood of a worker getting a full-time job with skill set and gender being the primary explanatory variables. We follow the standard practise of using highest level of education attained as a proxy for worker’s skill set. We design the following regression framework:

$$Y _{{{\text{is}}}}^{w} = \beta_{1} + \mathop \sum \limits_{k = 2}^{5} \beta_{k} {\text{education}}_{k\;is}^{w} + \beta_{6 } {\text{male}}_{i} + \mathop \sum \limits_{J = 2}^{5} \beta_{5 + J} {\text{education}}_{J\;is}^{w} *{\text{male}}_{i} + \alpha {\text{state}}_{s} + \varepsilon_{{{\text{is}}}}$$
(1)

\({Y }_{\mathrm{is}}^{w}\) here is a dichotomous dependent variable serving as an indicator of Full-time employment. It assumes a value one if the worker i in State s is under Full-time employment in wave w. Else, it assumes a value zero. The education level of the worker,\({\mathrm{education}}_{k \mathrm{is}}^{w}\), is the main predictor variable for each individual wave. \({\mathrm{education}}_{k \mathrm{is}}^{w}\) is a categorical variable simplified into the following mutually exclusive categories:

  1. i.

    \({\mathrm{education}}_{k\mathrm{ is}}^{w}\) = 1 for worker i with no formal education

  2. ii.

    \({\mathrm{education}}_{k\mathrm{ is}}^{w}\) = 2 for worker i with highest qualification as primary education

  3. iii.

    \({\mathrm{education}}_{k\mathrm{ is}}^{w}\) = 3 for worker i with highest qualification as secondary education

  4. iv.

    \({\mathrm{education}}_{k\mathrm{ is}}^{w}\) = 4 for worker i with highest qualification as graduate

  5. v.

    \({\mathrm{education}}_{k \mathrm{is}}^{w}\) = 5 for worker i with highest qualification as postgraduate

\({male}_{i}\) is the gender dummy taking a value 1 for male and 0 for the female gender. \(\varepsilon_{{{\text{is}}}}\) is the error term. Employment Laws being State subject in Indian polity, the standard errors are clustered at the state level. Reference category for the analysis here is thus females with no education. Our model coefficients give the likelihood of a worker having a full-time employment contract in a particular wave. We compare the average likelihood coefficients of having full-time contracts for Pre=Covid CPHS Waves (Wave 16, 17 and 18), During Restrictions CPHS Waves (Wave 19 and 20) and Post Relaxations CPHS Waves (Wave 21, 22, 23 and 24) to visualize the effect of pandemic on the Indian labour market. The likelihood coefficient is analyzed separately for male and female.

Table 4 summarizes the average differential likelihood coefficients for our analysis. Tables 9, 10 and 11 in Appendix can be referred to get the detailed regression output. Analyzing Table 4 here, \({\beta }_{1}\) gives the likelihood coefficient of having a full-time employment contract for the reference category of women with no primary education. \({\beta }_{2}\) onwards till \({\beta }_{5}\) gives the differential likelihood coefficients for women with Primary Education, Secondary Education, Graduate and Postgraduate degrees, respectively, over the reference category, i.e., gives the respective gain in likelihood of having a full-time employment contract for the given increase in education. \({\beta }_{6}\) onwards till \({\beta }_{10}\) gives the differential likelihood for men with no education and the other four categories of education respectively over the reference category of women with no education. The average pre-Covid waves coefficient for the reference category of females with no education is obtained as 0.04, i.e., on an average, a woman without formal education stands a 4% chance of having a full-time job contract. The likelihood goes up by 2.0 percentage points with primary education, 0.4 percentage points for secondary education, 4.0 percentage points for a graduate woman and 14.0 percentage points for a postgraduate woman reference to the base category. We observe a secular increase in the likelihood of fulltime employment for both men and women with the level of education. Our results stand significant and robust controlling for state fixed effects. Positive differential coefficients for men over the reference category implies a higher average likelihood of having full-time contracts for men over women across skill set.

Table 4 Average Differential Likelihood Coefficients
Full size table

Table 5 calculates the average total likelihood coefficients for all the categories of education and no education for men and women separately for Pre-Covid CPHS waves, During Restrictions CPHS waves and Post relaxation CPHS waves period.

Table 5 Total Likelihood Coefficients
Full size table

What we observe from Table 5 is that the likelihood of getting a job goes down for most groups during the pandemic and the lockdown due to that. However, strikingly the likelihood of getting a full-time job goes up for those with no education, for both men and women. In fact, it goes up substantially—by about 50% for women and 44% for men. It also goes up by about 27% for women with post-graduate degree. For all other groups, it either goes down or the change is miniscule. What is also intriguing is that the likelihood of getting a full time job remains elevated for men even after the restrictions are removed, that for women reverts to the level seen before the pandemic. Further investigation is needed to understand why this has happened. Tables 6, 7 and 8 represent the regression outputs for the three phases of study, respectively.

Table 6 Pre-Covid CPHS Waves Analysis
Full size table
Table 7 During Restrictions CPHS waves Analysis
Full size table
Table 8 Post Relaxations CPHS Waves Analysis
Full size table

6 Conclusion

Using CPHS longitudinal panel data, this paper tracks labour market flows and transitions in India for 36 months starting January 2019. The transition matrix and labour flow charts generated in the paper serve as an efficient tool for tracking the labour flow movement in India at regular intervals and open up the possibility of studying a large number of features of the Indian economy, which was not possible earlier. By tracing labour statistics and calculating the probabilities of losing and retaining a job, specifically a Full-time or Part-time job, along with the transitions to Unemployment and Out of Labour Force, this paper provides a comprehensive guide for understanding the Indian labour market and the economy. This will also be useful for evidence based policy design. We apply these constructs to study the impact of pandemic on heterogenous groups in labour market in India. We observe a declining probability of Employment-to-Employment transitions, i.e., higher probability of job loss even before the pandemic. The pandemic reduced that probability drastically even further, which was mirrored in an increase in Employment-to-Out of Labour Force transition during the pandemic. The trajectory shows a similar trend for both full-time and part-time contracts. There is a sharp rise in Employed-to-Data Not Available transition during the pandemic-induced lockdown waves possibly reflecting the large-scale urban to rural migration observed in India. An event study analysis carried out in the paper explains the likelihood of entering labour market as a function of workers’ education and gender, and the impact of the pandemic across various groups. The likelihood coefficients are observed to go up with the educational qualification of the workers. The paper finds men as more likely to get into a full-time contract at all levels of education. Further, the likelihood of getting a job during the pandemic is observed, somewhat unexpectedly, to have improved for workers without formal education. However, it goes down for most other groups in our analysis. The likelihood of getting a full-time job remains elevated for men even after the restrictions are removed, that for women reverts to the level seen before the pandemic.

This paper opens up the possibility of understanding several questions related to the dynamics of the labour market in India in greater depth. This can also become a potent tool in informing policy in India (See Tables 9, 10 and 11 in the Appendix).