Conference Agenda

Okun’s law is one of the best-known empirical regularities in macroeconomics. It addresses the issue of how much a country’s output is ‘lost’ when unemployment exceeds its natural or trend rate. In the existing literature, most of the studies assume a linear long-run relationship between changes in output and unemployment over the business cycle, implying that economic expansions and recessions have the same effect on unemployment. Nevertheless, this negative relationship may take a nonlinear form, in the sense that changes in output may cause asymmetric changes in the unemployment rate. In that context, the objective of this article is to test the linear assumption of Okun's law by deploying a time series analysis of asymmetric adjustments using seasonally adjusted quarterly data from Chile for the period 1996-2019 disaggregated by sex. The analysis is based on the asymmetric error-correction model proposed by Harris and Silverstone (2001), which provides a different approach to describe the short-term dynamics between the variables, but at the same time presents different challenges for its implementation. Hence, the analysis is not only limited to the study of the data, but also to the advantages and disadvantages of procedures that require cointegrated non-stationary time series to establish a long-run equilibrium. Moreover, the relevance of this research lies in the consequences of a misinterpretation of the effects of business cycles on unemployment if asymmetry is ignored. For instance, the existence of asymmetric effects shows different speeds of adjustment towards long-run equilibrium. Therefore, the effectiveness and required ‘size’ of stabilization policy on the real economy will depend on the ‘regime’ in which the Okun’s relationship lies. Additionally, this study contributes to the literature mainly in two aspects. First, there are no other studies on asymmetry in Okun's law using data of Chile, which means a significant gap for policymakers since short-run output and unemployment adjustments to disequilibrium differ according to whether upturns or downturns in the business cycle are considered. And second, incorporating a different econometric approach to test Okun's assumption provides the possibility of analysing the necessary procedural conditions and draw conclusions about the advantages/disadvantages in its implementation. This will lead to a critical analysis of the methods widely used in the literature and the consistency of the results. Finally, the main finding is that changes in output and unemployment in Chile are related in both the long and short-term, but only if an asymmetric approach is taken.

Introduction

The concept of quality of employment (or job quality) has gained international prominence, both for the scientific community and for public authorities, in a global context of an increase in non-standard jobs. In countries where the informal economy is widespread, such as Brazil, traditional indicators such as the unemployment rate, participation rate, and wages are not necessarily the most representative of the dynamics taking place in the labor market.

For these reasons, we analyze socioeconomic conditions through the multidimensional employment quality index proposed by Huneeus et al. (2015), extending their analysis on the evolution and determinants of quality of employment in Brazil that covered the period 2002-2011.

Research question

How has the succession of interlinked crises since 2012 translated into job quality in Brazil?

Data and methodology

The Brazilian Institute of Geography and Statistics (IBGE) has been extending the National Household Sample Survey database with a new continuous database offering micro data on employment and household characteristics (education, work, income, and housing of individuals) disaggregated quarterly and annually. We use it in the construction of our multidimensional employment quality index between 2012 and 2021.

To study the evolution of quality of employment in Brazil between 2012 and 2021, we construct a multidimensional employment quality index of (low) quality of employment that is based on the methodology of the multidimensional poverty measures of Alkire and Foster (2009; 2011). The first dimension concerns the consideration of an appropriate wage and productive work. The indicator chosen is the level of income, which is approximated here by the prevailing hourly wage.

The second dimension relates to social security. The indicator chosen is the formality of employment.

The third dimension concerns job stability and security. The indicator chosen is job stability.

Fortunately, PNADC allows us to retain the dimension of underemployment by asking part-time workers whether their situation is voluntary or involuntary, and how many hours they would like to work if they could (fourth dimension).

To analyse the correlation between employment quality and other worker characteristics, we estimate the probability of observing low-quality employment using a Dynamic Probit with Random Effects.

Preliminary findings

Our preliminary findings highlight that quality of employment continued to improve between 2012 and 2015 and started to deteriorate from the 2016 crisis. Surprisingly, quality of employment improved in 2020 when the Covid-19 crisis hit Brazil hard.

Introduction

ILO’s decent work programme aims at providing full and productive employment by way of facilitating social protection, dialogue and effective representation at work. Most countries are working towards its accomplishment, while some have made significant progress; for others, especially the developing states, challenge of vulnerable employment has made generation of decent work a complex task. ILO, based on status of employment, defines own-account and contributing family workers as vulnerable. However, vulnerability is more widespread than the current definition acknowledges as not all self-employed workers are vulnerable and not all paid-work forms are decent. In this backdrop, the paper aims to develop a composite index for measuring vulnerability in labour markets that addresses the issues of rights at work, job insecurity, social protection and overall decent working environment.

Research Questions

1. Can a technique of measuring vulnerable employment be developed that captures its multifaceted nature and does not generalise a particular status in employment as vulnerable?

2. What is the status of vulnerable workers in the Indian context?

Methodology

The study is empirical in nature and is based on an assessment of the Indian labour market. First, a modified definition of vulnerable employment that recognises its multifaceted nature is proposed, based on which conditions of employment that make work vulnerable are identified. Data is collated from annual periodic labour force surveys furnished by the Indian Government and indicators describing conditions of employment are selected. Several statistical and econometric tools like Principal Component Analysis and Multinomial Logistic Regression are used on the normalised indicators for the development and assessment of the index. The index is further used to identify the factors that lead to vulnerability of a worker and hinders access to decent work.

Findings

Few of the key findings are: (i) self-employment is not always vulnerable but is often driven by the flourishing start-up culture and entrepreneurial spirit in India; (ii) in paid-work, more than two-third workers are working without job and social protection; (iii) lack of skills and training is one of the major reasons why increasing share of workers are in vulnerable work; and, (iv) escaping vulnerability is highly restricted for women, young and older workers.

Contribution

The study presents an innovative measurement technique that not only aids in identifying vulnerability but also provide ways of alleviating vulnerable employment. The results are expected to lend support to policymakers in identifying key areas where decent work deficits exist.