Labour supply in conditions of high unemployment: Evidence from a transition and post-conflict economy – The case of Kosovo

Abstract

In this paper, we investigate the determinants of labour supply in Kosovo. Our approach takes into account demand-side restrictions of the labour market, which are typical in such a high unemployment economy, and pays particular attention to gender differences. We find that males’ labour supply is positively affected by their potential wages, especially in urban areas, while females’ labour supply is negatively affected by household labour income. We find significant regional differences in the labour supply of females, but not for males, suggesting that females are less mobile and unable to take advantage of more distant employment opportunities. In addition, we assess the appropriateness of the ILO guidelines in determining an individual’s labour force status in such environments and find that 18% of the inactive individuals are statistically a distinctive group (discouraged workers). We investigated the determinants engaging in job search and found that males and the more educated show a stronger attachment to the labour market. Our advice is for national statistics to report both narrow (ILO definition) and broad (that accounts for the discouraged workers) unemployment rates. This is the first systematic study of these issues in this post-socialist and post-conflict economy.

Keywords:

• labour supply
• discouraged workers
• unemployment
• transition economies
• Kosovo

JEL classifications:

• J21
• J24
• J64
• P39

Notes

1. Models have also been developed which analyze the dynamics of labour supply over the lifecycle. These models assume that a person maximizes her utility taking into account lifecycle earnings and other related determinants of the labour force participation.

2. Later we investigate empirically if this assumption applies to all of the inactive population.

3. Some 0.5% of the sample did not report their education level. They are included in the reference category of less than upper secondary education. As Cameron and Trivedi (2005) suggest, we could have applied listwise deletion for these observations, since they represent a small proportion of the total sample. But, given that primary and lower secondary education is compulsory then it is very likely that these individuals have completed at least this level of education. This is likely to be the case since most of those individuals with missing data on their education level are above 50.

4. A better way to test for this would have been by including a variable for the number of children, but this information cannot be constructed from our data-set.

5. We constructed an interactive dummy (not reported in Table 2) between marital status and female of age 25–40 to further examine this issue, but found no significant coefficient on this dummy.

6. The OECD (1995) finds that in 1993 they constituted 5.7% of the labour force in Italy, 3.1% in Sweden, 3% in Australia etc. Jones and Riddell (1999) find for Canada that the discouraged workers constitute 25–35% of the unemployed.

7. Due to not having many observations in each subcategory (in particular for N_d) we do not include interactive dummies between the residence and other explanatory variables. We rely only on the residence dummy that allows for different intercepts.