Flexitime
Flexible Working Hours and its Effect on the Health, Job
Satisfaction and Life Satisfaction of an Individual
Mart G. Marquering 11935634
University of Amsterdam
Supervisor; prof. dr. H. Oosterbeek Business Economics: Managerial Economics & Strategy ECTS: 35/60
Abstract
A longitudinal household study (BHPS) is used to analyse the effect of flexible working hour schedules (flexitime) on the mental health, physical health, job satisfaction and life satisfaction of an individual. Using a fixed effect estimation approach the results suggest that flexitime has a significant positive impact on the job satisfaction and life satisfaction of an individual in general. When separating for gender the effect of a flexible working hour schedule on job satisfaction and life satisfaction only remains significant for males. The significant effect of flexitime on the earlier mentioned two falls away for females. No significant effect is found regarding the effect of flexitime on mental health and physical health. It remains insignificant when gender is separated.
Keywords: Flexible working hours; flexitime; health; mental health; physical health; job
satisfaction; life satisfaction; gender difference; British Household Panel Survey.
JEL-codes: M20, M54, J28
Statement of Originality
This document is written by Student Mart Marquering who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
1. Introduction
Time is one of the world’s most precious gifts. It’s only usable once and there isn’t a possibility to buy it back. Since the beginning of civilized mankind, humans made inventions to quicken every process possible in order to free up time. For instance, when Louis Goldenberg invented the electronic washing machine at the beginning of the 20th century it was revolutionary in a way that women didn’t need to wash clothes by hand anymore (Sophisticated-edge, 2015). The washing machine saved time drastically and thereby opened up time to spend on other, more fun or important, things.
Nowadays almost everything is automated and people start to search for other ways to be more effective with their available time. One of the ideas to use time more efficiently is flexitime working. Even though its first introduction in 1967, flexitime only really hit the ground in the early 2000s and since its introduction a lot of companies started to implement it. Kuhne and Blair (1978) portrait flexitime as follows: “The basic intention of a flexitime system is to allow employees greater flexibility in determining their starting- and ending working hours in a given work day. Under a typical flexitime schedule, employees can arrive at work any time between 7:00 and 9:30 a.m. and leave work at any time between 3:00 and 6:00 p.m.”. To be short, flexitime is a flexible working hour schedule that allows employees to alter their working day.
Due to the recent growth in flexitime usage and the possible effects on human aspects of life flexitime could have, research is being conducted regarding the effects of flexitime working. This paper wants to know if it possibly has a positive or negative effect on the health of an individual. Also, could it be that flexitime working influences the job satisfaction and life satisfaction? These types of questions bring us to the purpose of this paper and to the research question: Does flexitime have a statistically significant effect on the health, job satisfaction and
life satisfaction of an individual?
When individuals work on basis of flexitime it is assumed that they can better manage their personal life in combination with their working life (White et al., 2003). This, in turn, could have a positive effect on the individuals’ life satisfaction and health (Costa et al., 2004). For example, parents have time to bring their children to school or people could go to the gym in the morning. However, it could also be that due to more irregular working hours individuals have an ever-changing daily rhythm and due to this experience more stress. Thus, it can lead to an increase the physical health but at the same time a decrease of the mental health (Ter Hoeven
& Van Zoonen, 2015). Because of the possible different path mental health and physical health follow, they are being separated in this paper’s research. Next, the possibility given by the management to have flexible working hours can result in a higher job satisfaction of the employees (McNall et al, 2010). The management is giving more control to the employees in how to fill in their working hour schedule. As a result of this, employees can potentially experience this as having more control over their own life. Therefore, it is expected that flexitime will have a positive effect on physical health, job satisfaction, and life satisfaction. Regarding its effect on mental health, it depends on how individuals cope with irregular working days. When the individuals deal well with it, mental health is expected to increase, if not, it is expected to decrease.
This topic is highly relevant since flexitime is a rather new phenomenon which is still in an early stage of implementation for most companies. The established literature is quite thin. Also, this paper makes use of panel data where most of the past literature uses cross-sectional data. The usage of panel data allows this paper to focus on the difference within an individual, rather than between individuals. The data used in previous research is also often not a good representation of the average society of a country, while the data that is being used in this paper is. Something else that lacks in previous research is the different effects flexitime can have between gender. Most literature states that flexitime has a positive or negative effect on health, job satisfaction or life satisfaction. However, none of these articles differentiate between gender. Is it possible that males react differently compared to females to a change in flexitime? Thus, this paper will try to contribute to a better understanding of the effect flexitime can have on the physical health, mental health, job satisfaction and life satisfaction of an individual.
The importance and possible usage of the provided information in this paper could have positive effects for three groups. First, the employer. Past research has shown that an increase in job satisfaction has effect on job performance (Judge et al, 2001; Bakan et al., 2014), which in turn has a positive effect on turnover and revenue (Sonnentag et al., 2008). Assuming that employers are profit maximisers, they would want to switch to a system that benefits them. Also, an increase in the employees’ health and life satisfaction results in less absenteeism (Darr & Johns, 2008; Ala-Mursala, 2006). Second, the employee. Flexible working hours will give employees more freedom and control of their life. They can better manage their social and working life and they feel better prepared psychologically and physiologically (Kuhne & Blair, 1978). It can also positively influence their life satisfaction and health (Costa et al., 2014), now who doesn’t
want that? The third party that benefits from flexitime working is the whole society. People starting at different times mean that people will head into traffic at different times, resulting in fewer traffic jams. This also holds for other facilities like barbershops and dentists. Under fixed working hour schedules those facilities get overloaded after the regular working day ends. This will happen less often when more people have flexible working hours (Kuhne & Blair, 1978).
The next section discusses the previous literature regarding this matter. Section 3 handles the used data and section 4 explains the used methodology. The results are shown in section 5 and this paper ends with a discussion in section 6.
2. Place in the Literature
The previous literature regarding this topic suggests that people who have access to flexible working hours, in general, have a higher job satisfaction and life satisfaction. Regarding the health of an individual, the findings are mixed. To dig deeper into this matter, this paper will first elaborate on the precise meaning of the phenomenon ‘flexitime’. After this, the section is divided into three subsections where each of the variables is discussed.
To give a better insight into what is meant and can be expected from a flexitime schedule, the explanation given by Kuhne and Blair (1978) is used. They divide a working day into three parts. The first part and third part are what they call the “quiet time”. These two parts of the day range from 7:00 a.m. until 9:30 a.m. and 3:00 p.m. until 6:00 p.m. The second part of the working day is called the “core time”, which is between 9:30 a.m. and 3:00 p.m. During the quiet times’ everyone is allowed to come in earlier/later and go home earlier/later. However, the core time is meant for everyone to be present at the job. In the figure below an easy to understand representation of a flexitime schedule is presented.
Figure 1. A working day under flexitime.
2.1. Health and Flexitime
The World Health Organisation (2006) describes health as the “complete state of physical, mental and social well-being and not merely the absence or infirmity”. Thus, health can be divided into three categories. This section will dig deeper into the past literature concerning the effects of flexitime and its possible effects on the physical health, mental health, and social well-being of an individual.
The literature concerning flexitime and its possible effects on health show that flexible working hour schedules have a good chance of increasing someone’s feeling about their health (Costa et al., 2014). Grzywacz et al. (2008) found that individuals who have access to flexible working arrangements report less stress and burnout than individuals who do not participate in such arrangements. They also concluded that individuals who have access to flexitime arrangements report less stress and burnouts than individuals who participate in other flexible working arrangements such as, compressed work weeks.
Ala-Mursala et al. (2006) found that a higher control of work time resulted in fewer health problems. It also improved their control on household tasks in combination with having a fulltime job. Their study linked the switch from little control of work time to having a lot of control on work time with the number of sickness absences. The major difference between my research and the above-mentioned study is the way health is measured. This paper distinguishes health between mental health and physical health and uses a 10-factor and 4-factor scale to determine both.
It is also possible that flexitime has negative outcomes on health. For instance, Ter Hoeven and Van Zoonen (2015) argue that Flexible Work Designs (FWD) could have both positive and negative consequences for the employees’ well-being. Their results show that FWDs and the employees’ well-being are positively related with each other through work/life balance, greater autonomy, and better communication. However, they find a negative relation between FWDs and health because of the more frequent interruptions flexitime causes. Another article (Joyce et al., 2010) evaluates the effects of flexible working interventions on the physical well-being, mental well-being, and general well-being of employees and their families. Their findings suggest that flexible working interventions could have a positive effect on health. However, they also pointed out that interventions which are motivated by organisational input lead to inexact or negative health effects.
Martens et al. (1999) studied 480 patients between the age of 20-60. All their subjects filled in an extensive questionnaire including several questions concerning their work, subjective physical health, psychological well-being and their quality of sleep. It turns out that people with irregular work schedules showed significantly more health complaints, had more sleeping problems and had a worse psychological state of mind than a control group of workers with non-flexible working schedules.
2.2. Job Satisfaction and Flexitime
Job satisfaction is an economic phenomenon that has been heavily debated in the past and therefore has many definitions that deviate little from each other. However, one definition of Locke (1969) stands out in my opinion: “Job satisfaction is the pleasurable emotional state resulting from the appraisal of one's job as achieving or facilitating the achievement of one's job values”. This section is dedicated to further investigate the previous conducted research on the effects of flexitime on job satisfaction.
Past research shows that flexitime has a positive effect on job satisfaction when comparing the flexitime user group versus the non-flexitime user group. For example, Masuda et al. (2011) investigate the availability of flexible work arrangements (FWAs) in organisations and their relationship with manager outcomes of job satisfaction. They make use of cross-sectional data and in their data pool they have 3,918 managers coming from 15 different countries. Their results suggest that the only FWA which has a significant impact on the manager’s job satisfaction is flexitime. A similar study conducted by Possenriede & Plantenga (2011) tries to analyse the effect of access to flexitime, telehomework and part-time work on the satisfaction of the employees with the fit between paid work and private life and their overall job satisfaction. This study is based on a cross-sectional survey data collected among more than 20,000 Dutch public sector employees. They concluded that FWAs, especially flexitime, are associated with a significant increase of the overall job satisfaction. It also seems that FWAs are not only appealing to employees with families, but more in general for all employees (Possenriede & Plantenga, 2011).
In another study conducted by Jang et al. (2010) they try to investigate the interaction effects of scheduling control and work-life balance programs and the possible effects it could have on job satisfaction. Their paper uses a sample of 1,293 employees in 50 companies in South Korea.
Here, again, is made use of a cross-sectional sample. They found evidence that giving the employees more control over their schedules has a positive effect on the job satisfaction of the individual. Similar results were found by Yaghi (2016). His study concludes that empowerment plays an important role in the relationship between flexible working hours and job satisfaction.
2.3. Life Satisfaction and Flexitime
There are many possible ways to define life satisfaction. However, Veenhoven’s (1996) definition of life satisfaction is the most accurate in my opinion. Namely, “Life-satisfaction is the degree to which a person positively evaluates the overall quality of his/her life as-a-whole. In other words, how much the person likes the life he/she leads.” He also mentions that frequently used synonyms for life satisfaction are happiness and subjective well-being. According to Veenhoven, the main difference between life satisfaction and happiness is that happiness is also used to refer to an objective good while life satisfaction emphasizes the subjective character of the concept. From Veenhoven’s definition it can be concluded that life satisfaction isn’t easy to determine. Thus, what determines someone’s satisfaction in life? Hoskins & May (2016) see life satisfaction as a broad concept which has multiple key determinants. Health and job satisfaction are two of these determinants. Other determinants for life satisfaction are income, housing, education, environment, and work-life balance.
Golden et al. (2013) try to investigate the effect of having discretion over one’s working time on someone’s happiness. They use a data pooled from two years of a nationally representative US survey. Controlling for a worker’s income bracket and work hours duration, it turns out that having work schedule flexibility in the form of an ability to take time off during the work day and to vary starting and quitting times daily, are both associated with greater happiness. Two years later Golden and Okulicz-Kozaryn (2015) conducted a similar study regarding the effects of work schedule control for the employees. Their results show that a good amount of work schedule control is strongly associated with a greater feeling of happiness.
3. Data and descriptive analyses
The data used for this study is obtained via the British Household Panel Survey (BHPS). BHPS is a household survey which is conducted every year. The main objective of BHPS is to further understand the effect social and economic changes can have on the individual level. The annual survey consists of a nationally representative sample of about 5,500 households recruited in
1991, containing a total of approximately 10,000 interviewed individuals. The sample is a stratified clustered design drawn from the Postcode Address File and all residents present at those addresses at the first wave of the survey were designated as panel members. These same individuals are re-interviewed each successive year and, if they split-off from original households to form new households, they are followed and all adult members of these households are also interviewed. Similarly, new members joining sample households become eligible for interview and children are interviewed when they reach the age of 16 (Institute for social and economic research, 2018). For this research, data from 1999 until 2008 is used. The final sample consists of 67,907 subjects. Where 32,728 (48,78%) of the subjects are male and 34,369 (51.22%) are female. 11% of the subjects is younger than 21. 20,5% is between 21-30, 25% is between 30-40, 23% between 40-50, 16% between 50-60 and 4,5% is older than 60.
3.1. Measurement of Flexitime
Whether our subjects have access to a flexitime schedule is measured with the following question:” which of the following flexible working arrangements apply to you?” The following eight options to choose from are presented:
1. Flexitime (flexible working hours); 2. Annualised hours contract;
3. Term time working; 4. Job sharing;
5. A nine-day fortnight;
6. A four-and-a-half-day week; 7. Zero hours contract;
8. None of these.
The first option (flexitime) and last option (None of these) are used to make the flexitime variable. The first option shows the subjects having access to a flexible working hour schedule and the last option shows the subjects who don’t have access to flexible working hour schedules. Making our main independent variable a dummy variable. In total, 18.92% of the subject pool has access to flexitime schedules while 81.08% doesn’t work with such an arrangement.
3.2. Measurement of Health
To measure the health of an individual, health is split up between mental health and physical health. This paper starts off with mental health.
3.2.1 Mental Health
Mental health is measured by using ten different questions that are combined into one scale. The questions are answered on a scale from 1 to 6 (1 = all of the time, 6 = none of the time) and can be found below:
1. Did you feel fully alive last month;
2. Have you been a very nervous person lately;
3. Have you felt down in the dumps that nothing could cheer you up; 4. Have you felt calm and cheerful;
5. Did you have a lot of energy;
6. Have you felt downhearted and low; 7. Did you feel worn out;
8. Have you been a happy person; 9. Did you feel tired;
10. Has your health limited your social activities.
The validity test shows that the above ten questions can be combined to one scale. The scale has an eigenvalue of 4.77 and the explained variance is 48%. Afterwards, a reliability test is conducted to test the reliability of the scale, which turned out to have a good reliability with a Cronbach’s alpha of 0.881. The mean of the mental health scale for people with no flexible working arrangement is 3.67. For our participants with a flexitime schedule the mean is 3.68. Figure 2 shows that most of the participants give themselves an average grade when it comes to their mental health. Approximately 2.5% has a mental health between the scale of 1 to 3, 87.5% of the participants have a mental health between 3 and 4. 10% of the participants give their mental health a rate higher than 4. It also shows that in the lower mental health scores slightly more non-flexitime users are represented than flexitime users. This is vice versa in the higher mental health scores.
1. The Cronbach’s alpha is a measure of scale reliability. The higher the value, the better the internal consistency.
Figure 2. Frequency distribution of the mental health scale separated between flexitime- and non-flexitime users
Notes. The vertical axis shows the percentage of subjects having a particular mental health value. The horizontal axis shows
the mental health scale. Source: BHPS data
3.2.2. Physical Health
The physical health of an individual is measured by combining four different questions into one scale. These four questions can be answered on a scale from 1 to 5 (1 = definitely true, 5 = definitely false) and are as follows:
1. I seem to get ill more easily than other people; 2. I am as healthy as anybody I know;
3. I expect my health to get worse; 4. My health is excellent.
Since question two and four are answered on a reversed scale, the scores for these two questions are turned around. Meaning that a score of 1 becomes a score of 5 et cetera. The four questions now have the same scale. The validity test shows that the four questions can be combined to one scale. The scale has an eigenvalue of 2.30 and the explained variance is 58%. It has a good reliability with a Cronbach’s alpha of 0.88. The mean of the physical health scale for people
with no flexible working arrangement is 4.12. For our participants with a flexitime schedule the mean is 4.09. This indicates a small negative effect when switching from fixed to flexible. The figure below shows that the majority of the participants is positive about their own physical health and that in general non-flexitime users score higher on their physical health. 70% of the participants give themselves a rating between 1-2, 22.50% give themselves a rating between 2-3 and only 7,5% of the participants give themselves a 2-3 or higher. When
Figure 3. Frequency distribution of the physical health scale separated between flexitime- and non-flexitime users
Notes. The vertical axis shows the percentage of subjects having a particular physical health value. The horizontal axis
shows the physical health scale. Source: BHPS data
3.3. Measurement of Job Satisfaction
Job satisfaction is measured by combining four different questions into one scale. All these questions can be answered on a scale from 1 to 7 (1 = not satisfied at all, 7 = completely satisfied) and can be found below:
1. How satisfied are you with your total pay, including any overtime or bonuses; 2. How satisfied are you with your job security;
4. How satisfied are you with the hours you work;
Due to incorrect answers, a few participants have a job satisfaction lower than 1. For the validity of this paper’s research these observations are dropped. As for the validity test, it shows that the five questions can be combined to one scale. This particular scale has an eigenvalue of 2.70 and the explained variance is 54%. The job satisfaction scale has a good reliability with a Cronbach’s alpha of 0.77. The mean of the job satisfaction scale for people with no flexible working arrangement is 5.21. For our participants with a flexitime schedule the mean is 5.31. Something else, in figure 4 can be seen that the participants all have a relatively high job satisfaction. Only 12,5% of the participants have a job satisfaction lower than 4, 22.5% has a job satisfaction between 4 - 5, the majority of the participants (45.5%) has a job satisfaction between 5 - 6 and 19% of the participants find themselves in the higher regions with a job satisfaction higher than 6. We can conclude that most of the participants have a relatively high job satisfaction.
Figure 4. Frequency distribution of the job satisfaction scale separated between flexitime- and non-flexitime users
Notes. The vertical axis shows the percentage of subjects having a particular job satisfaction value. The horizontal axis
3.4. Measurement of Life Satisfaction
The life satisfaction of an individual is measured by nine different questions, which are combined into one scale. The questions are answered on a scale from 1 to 7 (1 = not satisfied at all, 7 = completely satisfied) and can be found below:
1. How satisfied are you about your health;
2. How satisfied are you about the income of your household; 3. How satisfied are you about your house/flat;
4. How satisfied are you about your husband/wife/partner; 5. How satisfied are you about your job;
7. How satisfied are you about your social life;
8. How satisfied are you about the amount of leisure time you have; 9. How satisfied are you about the way you spend your leisure time.
Same as for the job satisfaction scale, due to incorrect answers a few participants have a life satisfaction score lower than 1. These observations are dropped. The validity test shows that the above nine questions can be combined into one scale. This scale has an eigenvalue of 5.93 with an explained variance of 74%. It also has a high reliability with a Cronbach’s alpha of 0.94. The mean of the life satisfaction scale for people with no flexible working arrangement is 4.90. For our participants with a flexitime schedule the mean is also 4.90. In the figure below can be seen that approximately 20% of the participants rate their life satisfaction below 4, around 35% rate their own life between 4 – 5, 35% of the participants’ life satisfaction is between 5 – 6 and 10% of the participants give themselves a 6 or higher. It can be concluded that the majority of the participants rate their lives between 4 and 6 (on a scale of 1 – 7).
Figure 5. Frequency distribution of life satisfaction separated between flexitime- and non-flexitime users
Notes. The vertical axis shows the percentage of subjects having a particular life satisfaction value. The horizontal axis
shows the life satisfaction scale value. Source: BHPS data
3.5. Control variables
Control variables that are going to be used are; gender, hours worked, permanent job, company size, job sector, financial situation, living with a partner, number of kids, age, age2 (to capture high-order age effects). In Table 1 the descriptive statistics are presented. To have the control variables as accurate as possible, this paper limits the hours worked per week to a minimum of 20 hours and a maximum of 84 hours. Subjects who work less than 20 hours a week have more than enough spare time left which means that a flexible working hour contract doesn’t have much effect on this group of people. Regarding the upper working limit, it is highly unlikely that the participants are working more than 84 hours a week. Furthermore, the financial situation of a person is measured with the question “How well would you say you yourself are managing financially these days? Would you say you are…”, which can be answered in six possible ways;
1. Living comfortably; 2. Doing alright;
3. Just about getting by; 4. Finding it quite difficult; 5. Finding it very difficult; 6. Don’t know.
Table 1. Descriptive statistics
obs mean sd min max
mental health scale 11,630 3.68 .31 2 5.2
physical health scale 11,630 4.12 .73 1 5
life satisfaction scale 50,244 4.89 .95 1 7
job satisfaction scale 58,188 5.27 1.00 1 7
health rob check 50,205 5.17 1.37 1 7
job satisfaction rob check 58,161 5.34 1.26 1 7
flexitime 58,195 .19 .39 0 1 gender 58,195 .47 .50 0 1 hours worked 58,195 40.22 9.93 20 84 permanent job 58,187 .045 .21 0 1 company size 58,119 5.07 2.45 1 11 job sector 58,192 1.81 1.45 1 8 financial situation 58,160 2.01 .88 1 5
living with partner 58,170 .31 .46 0 1
marital status 58,190 2.54 2.05 0 10
number of kids 57,571 .14 .58 0 9
age 58,186 38.44 12.16 16 82
age2 58,186 1625.55 974.31 256 6724
4. Empirical strategy
To investigate the effect of flexitime on health (mental and physical), job satisfaction and life satisfaction, the following model is estimated:
Yit = β0 + β1Fit + β2Vit + Ei + Dt + εit (1)
Where Yit could take on the form of the mental health, the physical health, the job satisfaction or the life satisfaction of individual i at time t. β0 isthe intercept, Fit is the scalar for the flexitime
variable, Vit is the scalarfor all the time variant explanatory variables (hours worked, permanent job, company size, job sector, financial situation, living with a partner, marital status, number of kids, age). Ei are the fixed effects to be estimated,Dt are added year dummies to check for
possible year effects and εit is the standard error term.
First of all, the Hausman test is performed to find out if a fixed- or random effect specification is used. The test suggests that a fixed effect approach needs to be used. Thus, this paper starts off with a fixed effect estimation approach, with no control variables added, to provide insight of the effect flexitime could have on mental health, physical health, job satisfaction and life satisfaction. Separate analyses are performed to capture the effects on men and women. The fixed effect approach will take out the differences in unobserved characteristics and offer a more accurate estimation of the coefficients. It will also allow for unobserved heterogeneity to be correlated with the dependent and independent variable. Furthermore, standard errors are clustered since some observations in the used dataset are related to each other. After the first set of regressions, this paper uses the fixed effect estimation approach again with the control variables added. The used control variables are selected based on the already existing literature on health (Ala-Mursala, 2006; Ter Hoeven & Van Zoonen, 2015), job satisfaction (Wheatley, 2017; Masuda et al., 2012), and life satisfaction (Golden et al., 2013; Hoskins & May, 2016). If proven statistically significant, the coefficients will reflect the point-change in the concerning scale when someone switches from no flexible working hour arrangement to a flexitime arrangement.
5. Results
The overall results suggest that flexitime has a positive and significant effect on the job satisfaction and life satisfaction of an individual. However, when separating for gender, it turns out that flexitime only has a significant effect on the job satisfaction and life satisfaction of males. No significance is found concerning the effect of flexitime on the mental health and physical health of an individual. The results of the conducted research will be elaborated in the coming subsections.
5.1. Mental health
To start off, the first three regressions in table 2 are done without any control variables and after the first regression, the second and third regression are separated by gender. When controlling for fixed effects, these regressions show no statistical significance. Next, using the model (1), regression (4), (5) and (6) are done with control variables. In the last three regressions no significant effect was found regarding the effect of flexitime on mental health in general. When separating for gender this doesn’t change.
Table 2. Estimated effect of flexitime on mental health.
(1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime 0.00513 -0.0103 0.0192 0.00164 -0.0170 0.0180 (0.0179) (0.0263) (0.0244) (0.0179) (0.0262) (0.0246) Observations 11,630 6,124 5,506 11,607 6,109 5,498 R-squared 0.000 0.000 0.001 0.006 0.014 0.003
Control variables? No No No Yes Yes Yes
Notes. Dependent variable: mental health scale. Control variables: hours worked, number of kids, financial
situation, living with a partner, marital status, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed effects and standard errors are clustered. Standard errors are in parentheses *** p<0.01, ** p<0.05. Source: BHPS data.
5.2. Physical health
Table 3 shows the results of flexitime on the physical health of an individual. When the control variables are not included no significance is found. When separating for gender, the effect of flexitime remains insignificant. When the control variables are included the results remain the same. No significant effect is found regarding the effect of flexitime on the physical health of an individual.
Table 3. Estimated effect of flexitime on physical health (1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime -0.0135 0.0213 -0.0454 0.00323 0.0462 -0.0349 (0.0322) (0.0446) (0.0460) (0.0322) (0.0447) (0.0463) Observations 11,630 6,125 5,505 11,619 6,118 5,501 R-squared 0.000 0.000 0.001 0.028 0.030 0.028
Control variables? No No No Yes Yes Yes
Notes. Dependent variable: physical health scale. Control variables: hours worked, financial situation, living with
a partner, job sector, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed effects and standard errors are clustered. Standard errors are in parentheses. *** p<0.01, ** p<0.05. Source: BHPS data.
5.3. Job Satisfaction
Table 4 shows that in general, when there are no control variables added, flexitime has a significant effect on the job satisfaction of an individual. When separating for gender flexitime is still significant for males and females. When the control variables are included flexitime stays significant with a coefficient of 0.0414. However, when separating for gender, only significance is found for males with a coefficient of 0.0525. Significance for females falls away.
Table 4. Estimated effect of flexitime on job satisfaction
(1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime 0.0494*** 0.0561*** 0.0431** 0.0414*** 0.0525*** 0.0322 (0.0142) (0.0205) (0.0196) (0.0139) (0.0200) (0.0193) Observations 58,181 30,616 27,565 58,054 30,544 27,510 R-squared 0.000 0.000 0.000 0.036 0.041 0.033
Control variables? No No No Yes Yes Yes
Notes. Dependent variable: job satisfaction scale. Control variables: gender, hours worked, permanent job,
financial situation, size of company working, job sector company is situated, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed. Standard errors are in parentheses *** p<0.01, ** p<0.05. Source: BHPS data.
5.4. Life Satisfaction
From table 5 can be concluded that in general, when there are no control variables included, flexitime has a significant effect on the life satisfaction of an individual. However, when gender is separated, flexitime remains only significant for males. For females it becomes insignificant. When the control variables are added, table 5 shows that the effect flexitime has on life satisfaction remains significant. When separating for gender it turns out that males have a positive significant effect of 0.0453 at the 5-% significance level. No significance is found for females.
Table 5. Estimated effect of flexitime on life satisfaction
(1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime 0.0348*** 0.0411** 0.0290 0.0246** 0.0370** 0.0132 (0.0118) (0.0166) (0.0166) (0.0114) (0.0162) (0.0160) Observations 50,244 26,437 23,807 50,022 26,316 23,706 R-squared 0.000 0.000 0.000 0.068 0.065 0.073
Control variables? No No No Yes Yes Yes
Notes. Dependent variable: life satisfaction scale. Control variables: gender, hours worked, permanent job, job
sector, financial situation, living with a partner, marital status, number of kids, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed effects. Standard errors are in parentheses *** p<0.01, ** p<0.05. Source: BHPS data.
6. Robustness checks
The robustness checks show the same results as the initial regressions done in the previous section regarding the health and job satisfaction of an individual. Unfortunately, the BHPS does not have another question available to measure the life satisfaction of an individual. The two coming subsections elaborate the results of the robustness checks.
6.1 Health
The BHPS data provides another adequate question to measure the health of an individual, namely “How satisfied or dissatisfied are you with your current health?” which can be answered on a scale from 1 to 7 (1 = not satisfied at all, 7 = completely satisfied). To measure the effect of flexitime on the health of an individual, model 1 is used again. The results also suggest that flexitime does not have a significant effect on the health of an individual. When separating for
gender the results remain insignificant. The results of the robustness check can be found in appendix A, table 6.
6.2 Job satisfaction
As a robustness check for the job satisfaction variable the question “All things considered, how satisfied or dissatisfied are you with your present job overall using the same 1 - 7 scale?” is used (1 = not satisfied at all, 7 = completely satisfied). The results indicate that flexitime indeed has a significant impact on the individual in general. When separating for gender the significance for females fades away. For males, it stays intact. The results of the robustness check can be found in appendix A, table 7.
7. Conclusion, Discussion, and Limitations
This paper’s goal is to investigate the effect of flexitime on health (mental and physical), job satisfaction and life satisfaction of the individual using panel data from BHPS in England between 1999 and 2008. Using a fixed effect estimation approach the results suggest that flexitime has a statistically significant effect on the life- and job satisfaction of an individual. When separating for gender it turns out that, for both, flexitime only has a significant impact on males. No significance was found concerning the effect of flexitime on the mental health and physical health of both males and females.
It can be concluded that, in general, flexitime has a positive effect on job satisfaction with a coefficient of 0.0474. Meaning that a change to a flexitime structure will lead to an increase of 0.0474-point in the job satisfaction scale, ceteris paribus. These findings are in line with Possenriede & Plantenga (2011) and Madusa et al. (2012). However, when separating for gender, the results show that it is only significant for males with a coefficient of 0.0604. A change to a flexitime structure will increase the job satisfaction scale of males with approximately 0.0604-point holding all other variables constant. A reason for the insignificant effect for women could be that women more often have a part-time job than men (Kjeldstad & Nymoen, 2012). López Bóo et al. (2010) investigated the relationship between part-time work and job satisfaction. Their results suggest that, for both men and women, job satisfaction is higher when they work full-time. The data used in this paper shows that males on average work 42 hours a week, while women work 31 hours a week.
Going further, table 5 indicates there is a significant effect of flexitime on life satisfaction in general with a coefficient of 0.0246. Though, when separating for gender, the results suggest that flexitime only has a positive significant effect on the life satisfaction of men with a coefficient of 0.037. Meaning that a switch to flexitime will increase the life satisfaction scale with 0.037 ceteris paribus. One possible reason for the insignificant effect of flexitime on women’s life satisfaction could be that one of the determinants of the life satisfaction scale is the job satisfaction of the individual. Flexitime doesn’t have a significant effect on the job satisfaction of a female, while it does have a significant effect on a male’s job satisfaction. This could be one of the reasons why it differs between gender. Also, Wheatley (2017) found that women often are constrained in their use of flexible work arrangements. This can result into insignificant effects of flexible work arrangements on the job satisfaction, life satisfaction and use of leisure time for females.
Although some literature indicates a link between flexitime and the mental health (Ter Hoeven & Van Zoonen, 2015; Costa et al., 2004; Martens et al., 1999) of an individual, no such evidence was found in this paper. Giving a closer look at how the mental health scale is set up, there are a few questions that a flexitime schedule could affect – question 5, 6 or 8 – and for all we know the effect could be negligible. It could be that flexitime has too little impact on the mental health of an individual since there are a lot of other factors that play a role when it comes to mental health. For instance, Socio-Economic Status (SES) plays an important role in the mental health of humans. SES is the individual’s economic and social position in relation with others based on education, income, and occupation (Meyer et al., 2014).
Also, no significant effect was found regarding flexitime and its effect on the physical health in general. When separating for gender, the results remain the same. One possible reason why this paper’s result is different compared to earlier mentioned papers could be the available data. The questions used for the physical and mental health scale were only asked once every 5 years which leads to having two periods of data available over a span of 9 years. As a consequence of this, another question to capture the health of the individual is used. After doing the robustness check it can be concluded that flexitime has no significant effect on the health of an individual.
Everything taken together, the results conclude that the availability of flexitime could have positive effects on life satisfaction and job satisfaction, in particular for men. No significant
effect is found regarding the health of an individual. Companies and institutions should really consider changing their policies regarding this matter if they haven’t already. Giving employees the option to take a flexitime schedule could potentially increase their job satisfaction, which in turn could lead to an increase in job performance ((Judge et al, 2001; Bakan et al., 2014). The good part about the flexitime structure is that it doesn’t conflict with a fixed hours schedule. They integrate well together. Meaning that two employees could both have a different schedule (fixed and flex) at the same time. Companies and employees can experiment with the possibilities that both working schedules bring and find a perfect fit for the specific employee.
Regarding the limitations, there is a possibility for omitted variable bias. Some control variables couldn’t be included due to inadequate availability of information in the BHPS database. For example, it wasn’t possible to measure if an individual switched job and therefore switched to a flexitime schedule. It could be that the estimated increase in job satisfaction is due to the job switch and not the flexible working hour schedule. Some other variables that weren’t possible to measure and could have an impact on life satisfaction, health or job satisfaction are; type of education followed and type of dwelling owned. Something else that could hurt the research is the definition of flexitime. When participants of the BHPS are interviewed, they won’t get a clear definition of what flexitime exactly is. Meaning that each individual can have, and probably has, an own interpretation of flexitime. Some may think it implies that employees can work from home more often, which it doesn’t. In the future, these limitations could be overcome with a better instructed and broader questionnaire where the participants know what some of the concepts mean. The last limitation could be the age of the used data (1999-2008). We are currently living in a fast-changing society were changes happen rapidly. It could be that in the past 10 years a lot changed regarding people’s perception about flexitime.
This paper found strong differences between man and women concerning the effect of flexitime among some of the dependent variables. Therefore, it would be interesting for future research to investigate what the exact reason is why flexitime doesn’t increase the job satisfaction and life satisfaction of women, while it increases a men’s job satisfaction and life satisfaction. Another option for potential future research would be to investigate which types of persons act positively and negatively to a flexitime schedule. For example, if we know that person type A reacts negatively to a flexitime schedule with respect to job satisfaction or life satisfaction, this person will not be considered to work with such type of flexible work arrangement. Lastly, this
paper found no statistical evidence regarding the effects of flexitime on mental health. Further research on this topic would be interesting to confirm this.
References
Ala-Mursula, L., Vahtera, J., Pentti, J., & Kivimäki, M. (2004). Effect of employee worktime control on health: A prospective cohort study. Occupational and Environmental Medicine, 61(3), 254–261.
Bakan, I., Buyukbese, T., Ersahan, B., Sezer, B., Sciences, A., & Imam, K. S. (2014). Effects of job satisfaction on job performance and occupational commitment. International Journal of Management and Information Technology, 9(1), 1472–1480.
Costa, G., Åkerstedt, T., Nachreiner, F., Baltieri, F., Carvalhais, J., Folkard, S., … Silvério, J. (2004). Flexible working hours, health, and well-being in Europe: Some
considerations from a SALTSA project. In Chronobiology International (Vol. 21, pp. 831–844).
Darr, W., & Johns, G. (2008). Work Strain, Health, and Absenteeism: A Meta-Analysis. Journal of Occupational Health Psychology, 13(4), 293–318.
Golden, L., Henly, J. R., & Lambert, S. (2013). Work schedule flexibility: A contributor to happiness? Journal of Social Research & Policy, 4(2), 107–135.
Golden, L., & Okulicz-Kozaryn, A. (2015). Work Hours and Worker Happiness in the US: Weekly Hours, Hours Preferences and Schedule Flexibility. SSRN Electronic Journal. Grzywacz, J. G., Carlson, D. S., & Shulkin, S. (2008). Schedule flexibility and stress: Linking
formal flexible arrangements and perceived flexibility to employee health. Community, Work and Family, 11(2), 199–214.
Institute for Social and Economic Research. (2018). Iseressexacuk. Retrieved 12 June, 2018, from https://www.iser.essex.ac.uk/bhps/about/sample
Jang, S. J., Park, R., & Zippay, A. (2011). The interaction effects of scheduling control and work-life balance programs on job satisfaction and mental health. International Journal of Social Welfare, 20(2), 135–143.
Joyce, K., Pabayo, R., Critchley, J. A., & Bambra, C. (2010). Flexible working conditions and their effects on employee health and wellbeing. In C. Bambra (Ed.), Cochrane
Database of Systematic Reviews. Chichester, UK: John Wiley & Sons, Ltd.
Judge, T. A., Bono, J. E., Thoresen, C. J., & Patton, G. K. (2001). The job satisfaction-job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127(3), 376–402.
Kjeldstad, R., & Nymoen, E. H. (2012). Part-time work and gender: Worker versus job explanations. International Labour Review, 151(1–2), 85–107.
Locke, E. A. (1969). What is job satisfaction? Organizational Behavior and Human Performance, 4(4), 309–336.
López Bóo, F., Madrigal, L., & Pagés, C. (2010). Part-Time Work, Gender and Job
Satisfaction: Evidence from a Developing Country. Journal of Development Studies, 46(9), 1543–1571.
Martens, M. F. J., Nijhuis, F. J. N., Boxtel, M. P. J. V. a N., & Knottnerus, J. a. (1999). Flexible work schedules and mental and physical health. A study of a working population with non-traditional working hours. Journal of Organizational Behavior, 20(1), 35–46.
Masuda, A. D., Poelmans, S. A. Y., Allen, T. D., Spector, P. E., Lapierre, L. M., Cooper, C. L., … Moreno-Velazquez, I. (2012). Flexible Work Arrangements Availability and their Relationship with Work-to-Family Conflict, Job Satisfaction, and Turnover Intentions: A Comparison of Three Country Clusters. Applied Psychology, 61(1), 1– 29.
Meyer, O. L., Castro-Schilo, L., & Aguilar-Gaxiola, S. (2014). Determinants of mental health and self-rated health: A model of socioeconomic status, neighborhood safety, and physical activity. American Journal of Public Health, 104(9), 1734–1741.
Possenriede, D. S., & Plantenga, J. (2011). Access to flexible work arrangements, working-time fit and job satisfaction. Discussion Paper Series/Tjalling C.Koopmans Research Institute, 11(22), 1–24.
Sayer, L. C. (2005). Gender, Time and Inequality: Trends in Women’s and Men’s Paid Work, Unpaid Work and Free Time. Social Forces, 84(1), 285–303.
Sonnentag, S., Volmer, J., & Spychala, A. (2008). Job Performance. The SAGE Handbook of Organizational Behavior, 1, 427–447.
Sophisticated-edge. (2015). Sophisticated EDGE. Retrieved 9 May, 2018, from http://www.sophisticatededge.com/who-invented-the-washing-machine.html Ter Hoeven, C. L., & van Zoonen, W. (2015). Flexible work designs and employee
well-being: Examining the effects of resources and demands. New Technology, Work and Employment, 30(3), 237–255.
Veenhoven, R. (1996). The Study of Life Satisfaction. Quality, 11–48.
Wheatley, D. (2017). Employee satisfaction and use of flexible working arrangements. Work, Employment and Society, 31(4), 567–585.
White, M., Hill, S., McGovern, P., Mills, C., & Smeaton, D. (2003). “High-performance” management practices, working hours and work-life balance. British Journal of Industrial Relations. Blackwell Publishing Ltd.
World Health Organisation. (2006). Constitution of the World Health Organisation. In Basic documents (pp. 1–18).
Yaghi, N. (2016), "Work Flexibility and Job Satisfaction: The Mediating Role of Employee Empowerment." Dissertation, Georgia State University.
Appendix A – Robustness checks
Table 6. Estimated effect of flexitime on health (robustness check)
(1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime 0.0101 0.00367 0.0161 0.00693 0.000815 0.0114 (0.0195) (0.0271) (0.0280) (0.0195) (0.0271) (0.0279) Observations 50,205 26,413 23,792 49,992 26,297 23,695 R-squared 0.000 0.000 0.000 0.015 0.016 0.015
Control variables? No No No Yes Yes Yes
Notes. Dependent variable: general health. Control variables: hours worked, financial situation, living with a
partner, number of kids, job sector, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed effects and standard errors are clustered. Standard errors are in parentheses. *** p<0.01, ** p<0.05. Source: BHPS data.
Table 7. Estimated effect of flexitime on job satisfaction
(1) (2) (3) (4) (5) (6) VARIABLES FE FE (♂) FE (♀) FE FE (♂) FE (♀) flexitime 0.0791*** 0.0933*** 0.0656** 0.0655*** 0.0853*** 0.0482 (0.0189) (0.0273) (0.0262) (0.0189) (0.0271) (0.0262) Observations 58,161 30,609 27,552 58,035 30,538 27,497 R-squared 0.000 0.001 0.000 0.015 0.019 0.012
Control variables? Yes Yes Yes Yes Yes Yes
Notes. Dependent variable: job satisfaction. Control variables: gender, hours worked, permanent job, financial
situation, size of company working, job sector company is situated, age, and age2. The table presents the coefficient for fixed effect regressions. All regression control for year fixed. Standard errors are in parentheses *** p<0.01, ** p<0.05. Source: BHPS data.
