Urbanization effects on job search decision�

Yudai Higashi

Okayama University

​

October 8, 2021

The 11th ARSA

1

Background

• Population decline and labor shortages
• Some countries have recently been facing population decline, and others are also predicted to do so in the near future
• Labor shortages are expected to constitute a substantial issue for these countries
• One of the solutions lies in utilizing a potential labor force
• Job search in local labor markets
• What the individuals consider to search for a job is the conditions of labor markets as well as their own personal situations
• The labor markets which the individuals consider could be local rather than nationwide
• One of the factors characterizing the conditions of local labor markets is their size, namely a level of urban agglomeration

2

Literature

• Agglomeration economies in job search and matching
• Individuals’ likelihood of finding a job is high (de Blasio and Di Addario, 2005; Phimister, 2005; Di Addario, 2011)
• Because of lower search costs and higher expectations to benefit from matching with an attractive job from plenty of offers
• Improving matching quality (Wheeler, 2001, 2008; Bleakley and Lin, 2012; Andini et al., 2013; Abel and Deitz, 2015)
• By reducing the gap between the skills of workers and the productivity of firms
• Heterogenous individual attributes
• The agglomeration effect on the likelihood of matching with a job is different across
• Gender (Phimister, 2005; Andini et al., 2013)
• Ages and skills (Andini et al., 2013)

3

Motivation

• Research question
1. Whether and how the urban agglomeration affects the decision of a non-working individual to search for a job?
2. Are the agglomeration effects different across heterogenous individual attributes, such as gender, education, and age?

​

• Contributions to the literature
• Studying the decision of non-working individuals to search for a job
• This is a step intervening before the success of individuals in finding jobs or earning wages occurs
• The investigation into this step is significant since it is a cost to the economy if individuals with enough ability to work do not wish to engage in productive activities

4

Theoretical background

• Non-working individual’s decision problem
• Bellman equation regarding a value of search (Mortensen, 1986)

​

​

​

​

​

​

​

​

• A non-working individual compares “searching and accepting the offer wage” or “doing nothing” to maximize his/her welfare

5

Present value of searching for a job

Interest rate

Value of leisure

Search cost

distribution of the wage offers

Offer arrival rate

best of random offer wage

Condition of deciding to search

6

Interest rate

reservation wage

Value of leisure

= opportunity cost of job search

Search cost

= out-of-pocket cost

of job search

distribution of the wage offers

Offer arrival rate

best of random offer wage

Benefit

Cost

Present value of searching for a job

Extending the model

7

Extending the model

8

Extending the model

9

Our model

• New participation condition
• incorporating the agglomeration effect and the heterogeneity of individual attributes

​

​

​

​

​

• The level of agglomeration weighted by individual attributes seems to determine whether the participation condition is satisfied
• Whether and how agglomeration affects the job search decision depends on the difference between the offer wage and the value of leisure/household production determined by individual attributes

​

10

6(4) types of individuals

11

Empirical model

12

Individual and household characteristics

Survey year dummies

Regional block dummies

Identification

13

Data

• Individual microdata
• Source: Japanese Employment Status Survey (8 waves, 1982–2017)
• Our sample contains individuals who answer not working
• Number of obs.=1,004,967 in the pooled data
• Regional agglomeration data (employment density)
• Source: Establishment and Enterprise Census (6 waves, 1981–2006) and the Economic Census for Business Frame (2009 and 2014)
• Regional unit:
• Urban Employment Area (UEA), based on commuting rates across municipalities, proposed by Kanemoto and Tokuoka (2002)
• A good proxy of the local labor markets in Japan
• Number of regions=222 UEAs

14

Descriptive evidence

15

Descriptive evidence

16

Results: whole sample

17

Results:�Gender, education, and age

18

agglomeration-search type

always-search type

(compared with other ages of low educated men)

always-search type

(compared with less educated men)

never-search type

(compared with men)

agglomeration-stop-search type

Offer wage

• less educated<highly educated
• men>women

value of leisure/household production

• men<women

Source of negative effect

• Characteristics of less educated women aged 30s

​

19

Labor force participation rates by age group

Female mean age of first marriage and childbirth

• Married women who raise their children are on average 30s years old
• They seem to spend their time on housework and childcare rather than working in the labor market

Results:�Marital status and presence of children

20

lower value of leisure/household production

higher value of leisure/household production

(agglomeration-search type -> never-search type -> agglomeration-stop-search type)

always-search type

(compared with less educated men)

higher value of leisure/household production

(always-search type -> never-search type -> never-search type)

Heterogeneous effects

• Gender gap
• the husbands and wives tend to specialize in the paid work in the labor market and household production (e.g., housework and childcare), respectively
• Life events (i.e., marriage and childbirth) makes the female value of leisure/household production high, especially in urban areas
• Due to supply shortages of childcare facilities in urban areas

​

• Low and highly educated women
• (one possible interpretation is that) highly educated women tend to marry highly educated men who earn high wages, therefore, the wives specialize in household production more

​

21

Conclusion

• Main findings
• Less educated men benefit from agglomeration, while those with high education and women do not
• Due to offer wage
• Life events, such as marriage and childbirth, remarkably change the female response of job search decisions to the agglomeration level
• Policy implications to utilize potential labor force
• To utilize inactive men:
• reinforcement of public employment referral to decrease their out-of-pocket cost of job search in rural areas
• Vocational training for them in rural areas to increase their offer wage
• To utilize inactive women:
• solving shortages of childcare facilities there to reduce their value of leisure/household production

22