A new approach to measuring poor-quality employment

Although policymakers often discuss poverty, they rarely address its connections with poor-quality employment. If they do, they typically focus on low incomes. Yet as Kirsten Sehnbruch, Mauricio Apablaza and James Foster explain, there are many employment conditions beyond wages that can negatively affect the wellbeing of workers and often exacerbate each other. Drawing on recent studies, they illustrate how a multidimensional measure of poor-quality employment could be applied to understand deprivation in labour markets better.

In public policy and economics, a “bad” job tends to mean a low-income one. Yet employment conditions such as an unstable job, a short-term contract or unpredictable working hours and associated pay can also have a significant negative impact on the wellbeing of workers. Moreover, such employment conditions tend to compound each other.

For this reason, it is difficult to capture poor-quality employment fully using simple measures like low wages or precarious contracts. These measures are unidimensional and do not reflect the fact that many workers are deprived in more than one aspect of their employment conditions. Instead, it is necessary to adopt a multidimensional approach that accounts for overall employment conditions.
Measuring poor-quality employment

If wages alone should not be used as an indicator of poor-quality employment, how do we conceptualise and define an appropriate measure? In a new study, we address this issue using the Alkire and Foster (AF) method to construct a multidimensional and synthetic measure of poor-quality employment.

Following the OECD, we select three dimensions that are generally deemed important in the literature on job quality. These dimensions are “earnings”, “stability and security” and “working conditions”. Figure 1 presents a visual illustration of how these dimensions overlap.

Figure 1: Dimensions of poor-quality employment



Note: For more information, see the authors’ accompanying paper.

Each of these dimensions can be populated with available variables. For instance, in countries with excessively high levels of job rotation, we might include “job tenure” or “short-term contracts” as a variable in the “stability and security dimension”. Alternatively, in countries where high temperatures make it undesirable to work outside, we might include “place of work” as a variable in the “working conditions” dimension.

To select appropriate variables and decide which employment conditions are acceptable or not, the literature on multidimensional poverty has either been based on technical analysis and expert opinion, as in the case of the UNDP or the World Bank’s Multidimensional Poverty Index (Measure), or, in the case of individual countries, expert and stakeholder commissions set up to determine the composition of indices. In the case of poor-quality employment, a similar approach can be adopted.

Once the dimensions and variables that compose a measure of poor-quality employment have been selected, the AF method uses a double cut-off strategy that allows the measure to focus on those workers with overlapping deprivations. The first cut-off divides each of the indicators of achievement vectors into discrete deprivations to identify which employment conditions are considered “poor”. Then the weighted sum of deprivations provides a counting vector that captures the accumulated distribution of detrimental conditions.
Poor-quality employment in Europe

Recent applications have mostly applied this approach to Latin America, but a similar measure would also work in Europe. In fact, a more complex measure of poor-quality employment can be established in Europe due to the availability of comparable data on the most essential employment conditions. Table 1 lays out a proposal for which variables could be included in such a measure, along with their weights and cut-offs.

Table 1: Dimensions, variables, weights and cut-offs for a measure of employment deprivation in Europe



Note: For more information, see the authors’ accompanying paper.

Using this approach, we can calculate the poor-quality employment levels across Europe. The overall cut-off line for the measure is one third (33%), i.e. to be considered as having poor-quality employment, a worker must be deprived in at least 33% of the overall measure. For example, a worker is considered to have poor-quality employment if they fall below the wage threshold defined by the measure, or if they have a fixed-term contract job of less than three years duration combined with being an involuntary part-time worker with a low level of autonomy.

Key to this method is that workers who are deprived in more employment conditions show up as being more “intensely deprived” than workers who are suffering from fewer deprivations. Figure 2 shows deprivation levels and the intensity of deprivation across European countries included in the European Working Conditions Survey.

Figure 2: Deprivation and intensity of deprivation in Europe



Note: Authors’ calculations with data from the European Working Conditions Survey, 2015. This is the last available data from this survey as the subsequent 2020 wave was disrupted by the COVID-19 pandemic. For more information, see the authors’ accompanying paper.

As Figure 2 illustrates, poor-quality employment levels vary significantly across countries, from 7.7% in Denmark to 31.7% in Greece, with Turkey and Albania constituting outliers at around the 40% mark. Some Southern European countries, such as Spain or Cyprus, also have higher levels of poor-quality employment (closer to 30%). It is interesting to note that several Eastern European countries (e.g. Slovenia and the Czech Republic) show both low levels of deprivation and intensity. Overall, we find that most workers are deprived in more than one dimension or variable of the measure, which illustrates the fact that it is not enough to look at poor-quality employment from a unidimensional perspective.

Implications for policymakers

Several conclusions that are relevant for policymakers emerge from our work. First, our research highlights the importance of viewing employment from a multidimensional perspective, emphasising that workers are generally deprived in more than one dimension, which can compound vulnerabilities. It is not enough to focus policy attention only on the working poor, on informal workers or on low-skilled workers. All these measures used on their own would miss important characteristics of employment that affect the wellbeing of employed workers.

Second, the literature on which our work is based shows that a measure of multidimensional employment deprivations is useful for identifying the most vulnerable individual workers or groups of workers in a labour market. Such a measure would provide policymakers with a more precise tool for focusing fiscal resources, both in terms of income support provided as well as for helping workers overcome deprivation through targeted investment in vocational training or adult education.

Third, it is clear policymakers systematically neglect some aspects of employment that are important to workers, such as job stability. This points to the need for regulatory reforms that level the playing field for workers with different types of contracts and employment conditions, thus disincentivising employers from using flexible or precarious forms of hiring when this is not appropriate.

Fourth, governments and international institutions should be investing more in generating better, more comparable data on employment conditions and on job characteristics. Labour force and household surveys with large samples across Europe should include a broader and more comprehensive set of questions on employment conditions. The European Working Conditions Survey is undertaken too infrequently and its sample sizes are too small to undertake meaningful analysis that can inform policymakers.

Finally, our research highlights that social and labour policies cannot be viewed or constructed in isolation. For example, reforming a pension system (such as by increasing the pensionable age or levels of contributions) may be an ineffective solution if the main reason why workers receive low pensions is because they did not consistently contribute to a system, either because they worked informally or never held a stable job.

For more information, see the authors’ accompanying paper at LSE Public Policy Review

Note: This article gives the views of the authors, not the position of EUROPP – European Politics and Policy or the London School of Economics. Featured image credit: metamorworks / Shutterstock.com


About the authors

Kirsten Sehnbruch  a British Academy Global Professor and a Distinguished Policy Fellow in the International Inequalities Institute at the London School of Economics and Political Science. She is currently Acting Director of the International Inequalities Institute.

Mauricio Apablaza is a Visiting Fellow in the International Inequalities Institute at the London School of Economics and Political Science, Director of Research in the School of Government at the Universidad del Desarrollo, Chile, and a Research Associate of the Oxford Poverty and Human Development Initiative (OPHI) at Oxford University.

James Propagates is a Visiting Professor in the International Inequalities Institute at the London School of Economics and Political Science, the Oliver T Carr Jr Professor of International Affairs and Professor of Economics at George Washington University, and a Research Associate of the Oxford Poverty and Human Development Initiative (OPHI) at Oxford University.

Posted In: Latest Research | LSE Comment | Politics