What are We Missing? Quantitative Analysis of Time-Related Underemployment
Onur Yavaş, Bilal CoşanThis research aims to analyse the variables that influence time-related underemployment in Türkiye. The sample was derived from the 2022 Household Labour Force Statistics microdataset from the Turkish Statistical Institute (TURKSTAT). First, descriptive statistics were used to analyse the distribution of time-related underemployment across the sub-sample groups. For the difference analysis, the Mann–Whitney U and Kruskal– Wallis-H tests were applied. Finally, binary logistic regression analysis was used to identify the determinants of time-related underemployment. All the independent variables, which were categorised as individual, social, and work-related, were found to result in statistically significant differences between the sub-sample groups. Furthermore, the findings indicate that gender, age, level of education, marital status, proximity, region of residence, Social Security Institution (SSI) registration, and workplace situation variables are significant predictors of time-related underemployment. In contrast, the analysis revealed that household size did not emerge as a significant predictor. When considered collectively, these independent variables account for 19.2% of the variance in time-related underemployment risk.
Bizim Neyimiz Eksik? Zamana Bağlı Eksik İstihdama Yönelik Nicel Bir Analiz
Onur Yavaş, Bilal CoşanBu araştırmanın amacı Türkiye’de zamana bağlı eksik istihdamı etkileyen değişkenleri analiz etmektir. Çalışmanın örneklemi, 2022 yılına ait Türkiye İstatistik Kurumu (TÜİK) Hanehalkı İşgücü İstatistikleri mikro veri setinden elde edilmiştir. İlk olarak betimleyici istatistikler ile zamana bağlı eksik istihdamın alt örneklem grupları arasındaki dağılımı incelenmiştir. Mann-Whitney U ve Kruskal Wallis-H testleri ise farklılık analizleri için kullanılmıştır. Son olarak, ikili (binary) lojistik regresyon analizi aracılığıyla zamana bağlı eksik istihdamın belirleyicileri tespit edilmiştir. Bireysel, sosyal ve iş değişkenleri kategorilerinde yer alan bağımsız değişkenlerin tamamının alt örneklem grupları arasında istatistiki açıdan anlamlı farklılaşmalara yol açtığı görülmüştür. Ayrıca cinsiyet, yaş, eğitim düzeyi, medeni durum, yakınlık, yaşanılan bölge, Sosyal Güvenlik Kurumu’na (SGK) kayıtlılık ve işyeri durumu değişkenlerinin zamana bağlı eksik istihdam riskini açıkladığı sonucuna ulaşılmıştır. Buna karşılık hanehalkı büyüklüğünün istatistiki açıdan anlamlı olmadığı saptanmıştır. Belirtilen bağımsız değişkenler, bir arada değerlendirildiğinde zamana bağlı eksik istihdam riskini %19,2 düzeyinde açıklamaktadır.
In the last quarter of the 20th century, the structure of sectors and employment has changed and advances in technology have increased unemployment rates. On the other hand, changes in the production and organisational structures of workplaces, as well as the internationalisation of competition due to globalisation, have led to an increase in flexible forms of employment such as temporary and part-time work (Görmüş, 2019, p. 68).
When discussing the labour market’s failure, unemployment is often the first issue that comes to mind, but underemployment is also a significant issue when considering the productivity loss it causes. Underemployment is undoubtedly a significant issue in terms of the wasteful use of resources, even though it does not have the same profound macroeconomic implications as unemployment (Dikmen, 2021, p. 212).
In practically all developing countries, underemployment and unemployment in urban areas are still prevalent. In this regard, underemployment and unemployment are significant issues in Türkiye. People cannot withstand unemployment for a long time because it comes at a heavy cost. To end unemployment, unemployed people might need to take advantage of their current job chances. People are being used inefficiently in this situation (Küçükali and Özmen, 2019, p. 142). In other words, underemployment causes a loss of social welfare. This problem is more common in agriculture economies, where informal employment is widespread, and in medium- and small-sized underdeveloped or developing countries where businesses dominate the economy. In this respect, it can be stated that underemployment is an important problem area in Türkiye (Ünal and Gönülaçan, 2019, p. 130).
Underemployment affects young people, women, and low-skilled workers globally. These individuals want full-time, permanent jobs but are often forced to accept temporary or part-time jobs. In general, the average education level of the labour force is increasing worldwide. However, despite the high educational attainment, the labour market often does not meet the needs of higher education. One consequence of this is the problem of overqualification, skills mismatch, or underemployment. Similarly, one of the major problems facing Türkiye is the inability of educated youth to find a job. In addition, another important problem area is that young people work in fields different from their qualifications or are employed on a temporary and part-time basis rather than full-time. As in other countries of the world, underemployment, which is a broadly defined element of unemployment in Türkiye, is an important social policy problem that points to the loss of productivity (Emeç, Üçdoğruk Birecikli and Kümbül Güler, 2020, p. 139-140).
Time-related underemployment is an important issue that requires examination. The purpose of this research is to analyse the variables influencing time-related underemployment in Türkiye. The sample was derived from the microdataset for 2022 Household Labour Force Statistics from TURKSTAT. First, those who were employed and working during the reference week were chosen from among the 628,936 participants. Then, individuals working less than 40 h per week were included in the sample. In the third stage, people except for the heads of households and their spouses and children were eliminated. Finally, those who worked during the reference week but had 0 working hours were excluded from the sample. The final sample size consisted of 37.958 participants.
In this research, time-related underemployment was the dependent variable. This variable was measured on the basis of respondents’ answers to the questions on employment, willingness to work more time and availability to work more time. Three categories were used to group the independent variables. Individual variables include gender, age, level of education, and marital status. Social variables included proximity, household size, and region of residence, whereas work-related variables included SSI registration and workplace situation. The research aims to answer two basic questions: (1) Are there statistically significant differences in time-related underemployment between the sub-sample groups? (2) To what extent do the independent variables used in the research explain the probability of being in time-related underemployment?
The article is divided into three main sections. The first highlights the conceptual discussion of time-related underemployment and emphasises the issue’s importance in the literature. Following that, TURKSTAT data are used to present the current situation of affairs in Türkiye. In the second part, after the method and sample used were introduced, the analyses were carried out. The distribution of time-related underemployment among subsample groups is analysed using descriptive statistics. Mann-Whitney U and Kruskal– Wallis-H tests, which are non-parametric analysis techniques, were utilised for difference analyses since it was found by the normality test (Kolmogorov-Smirnov) that the data were not normally distributed. Finally, binary logistic regression analysis was used to analyse the variables that explain time-related underemployment. In the last section, the results are summed up and contrasted with the results of the previous studies in the literature.
The results of the analyses that were carried out can be summed up. The difference analysis findings indicate that every variable used causes statistically significant differences in time-related underemployment. Most the sub-sample groupings show a considerable differentiation when pairwise comparisons between them are considered. The dependent variable’s change is explained by every independent variable, with the exception of household size, according to the findings of the binary logistic regression analysis. Together, individual, social, and work-related variables account for 19.2% of the risk of time-related underemployment.