Guest editorial

Citation

Beyer, R., Ghate, C. and Rama, M. (2019), "Guest editorial", Indian Growth and Development Review, Vol. 12 No. 2, pp. 149-152. https://doi.org/10.1108/IGDR-11-2019-126

Even though South Asia has been the fastest growing region in the world in recent years, there is a genuine concern whether this growth has been able to generate productive employment in these countries. For example, in India, the National Sample Survey Office (NSSO) data reveals that employment growth from 1993-1994 to 2017-2018 averaged 0.00975 per cent per annum; lower than the 0.185 per cent growth of the labor force[1]. The growth of regular wage employment has been even smaller, and formal employment remains the exception rather than the norm. Similar trends have been observed in other countries in South Asia.

Jobs matter on their own, and not just as a source of aggregate output. In developing countries, most household income is derived from work so that sluggish employment growth can affect income distribution. Jobs are also a source of identity and agency, potentially supporting the social compact. In patriarchal societies, paid employment can also help empower women, giving them greater voice at home and shifting household expenditures more toward children and human capital investments.

This centrality of jobs warrants a deeper understanding between the (possible) mismatch between economic growth and employment in the South Asia region. To address this issue, the Indian Statistical Institute – Delhi and the World Bank organized a workshop entitled “Jobless Growth?” in March 2018. The objective of the workshop was to try and answer the question – is jobless growth in South Asia a myth or a reality?

The workshop also examined the sectoral composition and patterns of economic growth, and their implications for job creation and destruction. Attention was given to the changing composition of employment (male vs. female, regular vs. casual), to the relationship between job creation and firm characteristics such as their size or age, and to its relationship with location characteristics including city size and business environment. The workshop discussed trends in labor productivity and their implications for job creation. Last but not least, there was a focus on the challenges of measuring employment (what is really a “job”?) and on changes in labor force participation, especially among women.

Subsequent to the workshop, five papers were selected to form part of a special issue on jobelss growth in South Asia in the Indian Growth and Development Review. While not doing justice to the richness of the discussion at the workshop, these five papers nonetheless provide an informative view of the status of academic research on employment and growth in the region.

The issue of measuring employment came out as particularly important, because in South Asia, a large part of the workforce is engaged in low productivity and low-income work. And disentangling underemployment, unemployment and inactivity is especially challenging when farming and casual work are prevalent. An important goal is therefore to develop a data architecture that is cognizant of the distinct characteristics of the region’s employment landscape.

Unfortunately, the available data sets do not capture jobs, but rather employment status. Three types of data sources can be tapped in this respect. The first group comprises population censuses and households surveys, the second includes censuses and surveys of establishments and the third builds on administrative data.

In India, for instance, the most prominent data source from the first group is the NSSO Quinquennial Employment-Unemployment Survey. This survey asks you whether you are in the labor force or not. It then asks one of three activity status: usual status (which has a long reference period of 365 days) – within this it asks for principal status, and then for subsidiary activity; current weekly activity status; and current daily status. The main problem with this dataset is that it doesn’t generate data at a high enough frequency. Recently, the survey has been discontinued. There is also a labor bureau unemployment- employment survey, although the last survey was done in 2016-2017, and which has also been discontinued. Both the NSSO Quinquennial survey and the Labor Bureau Survey have been recently replaced by the Period Labor Force Survey (PLFS) which has annual estimates for employment status, although there is quarterly data for urban areas.

In the second group of data sources, the ASI (Annual Survey of Industries) covers registered manufacturing firms, whereas the NSSO survey on unincorporated non-agricultural enterprises covers the unorganized sector. In a broad sense, the latter covers that part of manufacturing that is left out of the ASI. In enterprise surveys, the definition of employment is different. Here the definition is based on total man days worked divided by the total number of working days. In comparison, in household surveys, it corresponds to an economic activity during a particular reference period.

More recently, administrative data (i.e. payroll data) is being increasingly used to measure joblessness. The key question is whether a new entry into a payroll database necessarily means a new job created. There have been lots of revisions in the data, which leads to volatility.

To address this diversity of sources and definitions, the first article in this issue, by Robert C.M. Beyer, Milagros Chocce and Martin Rama, entitled “Employment in South Asia: A new dataset” presents a new dataset of comparable employment indicators for South Asian countries, constructed from more than 60 primary data sources from 2001 to 2017. The authors curate primary data from population censuses and household surveys in a consistent way across countries. They illustrate the usefulness of their dataset by first describing the structure of employment and its changes over time, and then analysing the dynamic relationship between economic growth and employment growth in South Asia.

An important distinction in this paper is between the short- and long-term effects of economic growth. In the short term, economic growth boosts labor demand, leading to the creation of more jobs and possibly driving wages up. A rich literature has analyzed this short-term effect, known as Okun’s Law, especially in advanced economies where the focus has been on the resulting changes in the unemployment rate. In the longer-term, however, higher labor earnings also allow children to stay longer in school and may result in women withdrawing from the labor force, which amounts to a reduction in labor supply. Combining both effects, Beyer et al. estimate that in South Asia one percentage point of growth of gross domestic product has led on average to a 0.34 per cent increase in employment.

The relationship between economic growth and employment growth is further explored in the paper by Vinoj Abraham and the paper by Suresh Chand Aggarwal and Bishwanath Goldar. In a broad sense, both papers aim to disentangle the contribution of structural transformation. Indeed, the relative weight of different sectors of activity evolves as countries become richer. The share of agricultural output declines, that of services increases and most often there is also industrialization, at least up to a point. If these sectors of activity are characterized by different intensities in the use of labor, the change in the economic structure may either amplify or dampen the mechanical contribution that could be expected from growth alone.

Using data from India, Vinoj Abraham in a paper entitled “Jobless growth through the lens of structural transformation” tries to understand why the elasticity of employment generation to GDP growth in India has declined between 1993-94 and 2011-2012. He finds that a rise in GDP per-capita despite a fall in the employment rate indicates the increasing prominence of labour productivity in GDP per capita growth. During this time period, low productivity sectors experienced slack in both employment growth within sectors and labour absorption from other sectors and, labour absorption was the highest in mid-productivity sectors, while employment growth was the highest in high productivity sectors. The rise in output per worker across all sectors, barring the construction sector, occurred along with a decline in employment in agriculture and agro-based industries. The decrease in employment in these sectors had two features: the declining share of persons that are in the workforce and a decline in employment growth itself.

Using the India KLEMS data set (2017), Goldar and Aggarwal in a paper entitled “Structure and growth of employment: evidence from India KLEMS data” analyze the structure and trend in employment in the Indian economy during 1980-1981 to 2015-2016. The authors explore several possible causes of the slow growth rate of employment in India. They find there has been a decline in the employment elasticity from 0.41 during 1980-1993 to just 0.10 during 2003-2015. As a policy prescription, they suggest that since long run GDP growth in India is typically pegged at 8 per cent in many policy documents, it is essential that aggregate employment intensity be increased by focusing on those sectors where the long term employment elasticity is high such as in construction, manufacturing, and services. They also suggest that since GDP growth in India has been mainly domestic demand driven and may continue to be so for many more years, policy and reform focus has to be on those domestic sectors which could be the drivers of employment growth and GDP growth in the future.

While macroeconomists analyse the drivers of employment at the aggregate or sectoral level, labor economists focus on its microeconomic determinants. Whether individuals are willing to work depends on their characteristics, including their educational attainment, their gender and age. Characteristics of their households matter as well. For example, the choices are not the same if the household owns agricultural land, or if it has other sources of income. Importantly, the characteristics of the locations where households live play a decisive role. The opportunities to work are different in rural versus urban areas, or in dynamic regions versus remote areas.

The paper by Izza Aftab and Umair Mazher entitled “Determining the Odds of Employment in Pakistan” analyzes employment trends in Pakistan using microeconomic data from the Pakistan Social and Living Standard Measurement Survey (PSLM) for the years 2010, 2012 and 2014. The survey contains individual information on employment status, health and education, and household information on sanitation, economic situation and asset ownership. Importantly, the district the household lives in is identified, which allows the authors to conduct their research differentiating the outcomes by province. The estimated role of individual and household characteristics is consistent with findings of the literature for other countries. But the results also show that the province households live in dramatically affects the employment odds of their members.

Economic research focuses not only on describing and explaining real phenomena, such as employment growth: it also aims to understand how economic policy can make a difference. Governments intervene in multiple ways to affect employment outcomes. Some interventions, such as a fiscal stimulus or easier access to credit, are outside the labor market. Others, including minimum wages or labor standards, try to influence the workings of the labor market. And finally, others involve the direct creation of employment, as in public works programs such as the National Rural Employment Guarantee Scheme (NREGS).

As discussed above, economic policies that foster economic growth should result in employment growth. There is less of a consensus in relation to labor market regulations, some of which may correct market imperfections and boost employment, whereas others may increase labor costs and discourage labor demand. As for public works programs and other forms of employment targeting, they could be expected to unambiguously increase aggregate employment. However, this is a result that cannot be taken for granted in a dual labor market.

The last paper in this issue, by Chetan Ghate and Debojyoti Mazumder entitled “Employment Targeting in a Frictional Labor Market” analyzes the issue in the context of a theoretical model characterized by search and matching frictions. In a first stage, they consider an economy without an informal sector. In that case, publicly-provided employment either reduces or fosters aggregate employment depending on whether it increases or reduces private sector wages. The possibility that a public works program could crowd-out private sector employment may have some empirical support in India’s case, where it has been argued that NREGS operates as a de facto minimum wage for rural areas.

Ghate and Mazumder further develop their model to include an informal sector. In this environment, they show that public sector intervention in the labor market can lead to an increase in the size of the informal sector. Because the informal sector is characterized by a high turnover rate and the absence of unemployment benefits, public works programs could have a perverse effect on labor market outcomes.

Note

1. The employment number corresponds to UPSS (Usual Principal and Subsidiary Status) which is commonly used for employment figures. The authors are grateful to Vinoj Abraham for helping them with this calculation.