Paper for LoWER-conference ‘Combining Home, Work and Education’, 26-27 October 2001, Braga, Portugal Amsterdam, 10 August, 2001 Cees NIEROP University of Amsterdam Roetersstraat 11 1018 WB Amsterdam The Netherlands Tel. +31(0)20 – 525 4053 Email: [email protected]What are the Consequences of Non-Standard Work Arrangements for the quality of life?
41
Embed
What are the Consequences of Non-Standard Work Arrangements …worlddatabaseofhappiness.eur.nl/hap_bib/freetexts/nier… · · 2018-01-233 I. Introduction Non-standard work arrangements
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
In brief, working with a non-standard work-arrangement is � also in Germany in the
mid nineties � above all something for women. It can predominantly be found in low-
skilled/low status/low paid jobs. The incidence of non-standard work is especially
underrepresented in the middle range of the income distribution. It gets slightly higher at the
upper end of the distribution. Consequently, the concentration of non-standard work-
arrangements is relatively high for professionals and in the Education-sector. These results are
in line with other research in the field (See Smith 1997; Kalleberg 2000 and Nolan et al.
2001).
However, these results are - at best - only indications of a lower quality of life of
employees with flexible work arrangements. The fact that flexible work-arrangements are
concentrated in these � unpleasant � parts of the economy is, however, not definite evidence
that flexible work-arrangements cause a lower quality of life of the employees involved. For
assessing the quality of life and judging flexible work arrangements, we need much more
information. We need to know firstly, the importance of income and status for life-quality.
We need to know among others how one values leisure-time, what the flexible labour is like,
whether it is outspokenly monotonous or not, we need to know how far flexible work-
8
arrangements are an individual choice or not, and we need to know how well a flexible work-
arrangement fits in the broader life-program of the individual. Above all, we need to define
quality of life. We need to answer the question whether quality of life should be measured in
terms of happiness and satisfaction or in terms of development or self-realisation. This is the
subject of the next section.
IV. QUALITY OF LIFE
In assessing quality of life, there are in political philosophy two dominant strands:
utilitarianism and liberalism. Of both strands, there are different versions. In the present
paper, only the basic ideas of both strands will be addressed (see for an overview of both
strands: Kymlicka 1995 and Rachels 1993).
Utilitarians define quality of life in terms of utillity. The exact definition of utility is,
however, a matter of dispute. Bentham and other early utilitarians, like John Stuart Mill, saw
utility or welfare as a desirable or agreeable state of consciousness. Bentham defined utility or
happiness as the balance of pleasure over pain. Pleasure should and will be seeked and pain
will be avoided. The experience or the sensation of pleasure is the chief good for mankind. Or
as Mill puts it: �The Utilitarian doctrine is that happiness is desirable, as an end; all other
things being desirable as means to that end� (Mill quoted in Rachels 1993: 103).
According to modern views however, pleasure and pain are much too shallow to
represent the full range of conscious states which determine an individual�s overall utility.
According to Dworkin, the concepts of �enjoyment� and �dissatisfaction� capture the range of
important conscious states much better. Therefore, the balance of enjoyment over
dissatisfaction is a more proper definition of utility. So, for increasing their utility, individuals
will seek enjoyment and avoid dissatisfaction. The better one succeeds, the happier one is and
the higher is, according to this theory, his or her quality of life. In this theory are only the
feelings of happiness important for the quality of life. How happiness should be measured is,
however, within and outside the utilitarian strand a subject of dispute (see for an overview
Kymlicka 1995, Dworkin 1981; ). One of the questions is, whether feelings of happiness
should be measured or �success� in a number of areas. Pending further inquiry, the present
paper attempts to measure feelings of happiness or satisfaction and leaves aside the success-
question.
9
The utilitarian account of �quality of life� has been the most dominant in the field of
philosophy, economics and public policy until the beginning of the seventies. Since then the
liberal account is gaining field. In the liberal account two concepts are of utmost importance:
human dignity and � related � neutrality.
The concept of human dignity, originally founded by Kant, requires that human beings
may never be used as a mean to an end. Or in Kant�s terminology: human beings are ends in
themselves. This is in sharp contrast to animals: �animals are merely means to an end. That
end is Man�, according to Kant. This is so because, firstly, human beings have desires and
goals in relation to their projects. Kant considers animals to be unable to have self-conscious
desires and goals. The second reason human beings have dignity is because they are rational
agents. That is: they are capable of making their own decisions, setting their own goals, and
guiding their own conduct by reason alone. Whilst human beings are rational agents with
rational plans and desires, they are ends in themselves. This means they may never be allowed
to be used as a mean for other ends than his own. This statement is strongly related to the
second distinctive feature of liberalism: neutrality.
Neutrality requires that we may never manipulate people, or use people to achieve our
purposes, no matter how good these purposes may be. If human beings are capable of setting
their own goals and have to be treated as ends in themselves, they should be free to pursue
their own goals and projects and the state should not reward or penalise particular projects or
conceptions of the good life. The extent to which one lives according to the own conception
of the good � happily or unhappily � determines in liberalism human dignity. And human
dignity determines instead of happiness or well-being, in liberalism the quality of life. The
questions what this dignified life consists of and which means are needed for a dignified life,
has in recent decades very persuasively been addressed by among others John Rawls and
Amartya Sen.
Building on Kant�s ideas, Rawls states that the chief human good is realising his
�higher-order-interests�. The higher-order interests are the interests concerning realising the
needs for forming, revising, and pursuing a conception of the good life (1993: 72). For
realising these higher-order interests, Rawls states that, �primary goods� are needed. The
�primary goods� are the basic liberties, income and wealth and the bases of self-respect. These
primary goods are so called �all purpose means�, things that men presumed to want whatever
else they want (1971: 260). Hence, the more primary goods one possesses, the more is one
capable of forming, revising and pursuing an own conception of the good life, which is the
chief human good. However, given the Kantian concepts of dignity and neutrality, the extent
10
to which individuals succeed in advancing their way of life, is for a liberal state not the
ultimate end. The end is to maximise the extent to which individuals are capable to succeed in
advancing their way of life. Therefore for measuring the quality of life, one should measure
the capabilities for forming, revising and pursuing a conception of the good life. These
capabilities concern, according to Rawls, the primary goods: the basic liberties, income,
wealth and the bases of self-respect.
Sen�s view of capabilities for life-quality is more broader. He sees living as a
combination of various doings and beings. These doings and beings represent part of the
states of a person � in particular the various things that he or she manages to do or be in
leading a life. Sen calls these doings and beings �functionings�. And a capability �reflects the
alternative combinations of functionings the person can achieve, and from which he or she
can choose one collection� (Sen 1993: 31, italics added - CN). According to Sen, life-quality
should be assessed in the concepts mentioned above: capabilities and functionings.
Some functionings are very basic, such as being adequately nourished, being in good
health, well sheltered, etc. For these functionings one may need for instance � among others �
Rawls�s primary goods income and wealth. (Or in Sen�s terminology: one should have the
capabitly to get enough food in order to be well nourished ( = functioning)).
Besides the basic functionings, there are also more complex functionings, such as
achieving self-respect, being socially integrated, having control over own life and � in
reference to Adam Smith � appearing in public without shame (Sen 1993). The extent to
which one has the capability to function in the ways mentioned above, determines according
to Sen the quality of life.
In the present paper, the consequences of flexible work-arrangements for the quality of
life will be assessed along the lines mentioned above. In the next section, the utilitarian and
liberal accounts of the quality of life will be further developed with the help of the variables
of the GSOEP-dataset.
V. DESCRIPTION OF THE KEY VARIABLES
The utilitarian account
In this section, an attempt will be made to operationalise the utilitarian, Rawlsian and Sennian
account of life-quality in order to determine the relationship between flexibible work-
11
arrangements and the quality of life. The three notions are all highly abstract and by way of
operationalising them, one loses much of the meaning of the concepts. However, for assessing
the consequences of for example policy or labour-market developments, measuring � even in
a poor way � the impact on the quality of life is from utmost importance.
In the GSOEP-dataset, there are two kinds of variables which attempt to measure the
well-being or happiness. Variables which measure life satisfaction and variables which
measure stress. The variables which measure life-satisfaction can be distinguished in variables
which measure general satisfaction and variables which measure domain satisfaction. With
general satisfaction is meant the satisfaction with life as a whole. In the GSOEP-dataset, two
variables measure the general satisfaction. From a scale from 1 (extremely dissatisfied) to 10
(extremely satisfied), the interviewees answer the questions: �How satisfied are you with your
life today?�, and �How happy do you think you will be in five years from now?�
Besides these general satisfaction variables, there are nine domain satisfaction-
variables in the dataset. These variables measure the satisfaction with specific domains of life
(Van Praag 2000: 2). Like the general satisfaction-variables, the domain satisfaction variables
have a scale from 1 to 10. In the dataset, the interviewee�s satisfaction with (1) health, (2)
work, (3) housework, (4) household income, (5) dwelling, (6) amount of leisure time, (7)
goods and services, (8) standard of living, and (9) environment has been measured.
In addition to the two mentioned general-satisfaction variable, I add one self-
constructed variable: the aggregate of the nine domain-satisfaction variables mentioned
above. All variables will be divided by their mean in order to make the effects comparable to
each other. In the next section, I will examine to what extent flexible work-arrangements
explain the variation of the general- and domain-satisfaction-variables.
Besides the mentioned variables which measure satisfaction, there are also in the
GSOEP-dataset several variables which measure stress. From a scale from 1 (applies fully) to
4 (does not apply), the interviewees have been asked to what extent the following statements
apply:
• Have control over own life
• Plans are successful
• Confident about future
• Not lonely
12
• enjoys work
• Able to cope with things
I have aggregated the scores on these variables and divided by the mean. This constructed
variable appears in the next sections under the name �control over life�.
Furthermore, in measuring stress, I take into account three variables which measure
from a scale from 1 (big worries) to 3 (no worries) to what extent one is worried about (1)
finances, (2) dwelling, and (3) job security. The variables will be aggregated and divided by
the mean to standardise the scores and make them comparable with the satisfaction-variables.
I call this constructed variable �the amount of worries�. In the next section, it wil be examined
to what extent flexible work-arrangements produce the feelings mentioned above and the
aggregate of the control- and worries-variables: stress.
The liberal account
In the liberal account, the focus is on human dignity and autonomy rather than on subjective
feelings of utility, well-being or satisfaction. Rawls stated that for human dignity, one should
have the capabilities to form, revise and pursue an own conception of the good life. The
primary goods reflect, according to Rawls, these capabilities. The primary goods were: the
basic liberties, income and wealth and the bases of self-respect. The more primary goods one
has, the more is one capable to live according to the own conception of the good, which is the
chief human good.
In this paper, I assume that all interviewees have the same basic liberties. The focus
will be on income and wealth, and the bases of self-respect. The primary goods income and
wealth are relatively easy to operationalise for the research. In the present paper, we use the
post-government household income as a proxy for the primary good �income� and the variable
�estimated value of the income from dividend and interest� as a proxy for wealth. How the
�bases of self-respect should be operationalised is far more complex. Pending further inquiry,
I decided to use the �control over life� variable mentioned in section IV to construct the
primary good �bases of self-respect�. This vector appears in section VI under the name
primary goods.
The proxies of the primary goods will be taken together. This metric will be an
indication of the extent to which one is capable to pursue an own conception of the good. The
13
question whether flexible work-arrangements increase or decrease this metric, will be
addressed in the next section.
Sen�s theory focuses on the extent to which one is capable to function as a human
being. For functioning as a human being, one should at least have the capabilities to be
adequately nourished, adequately sheltered, the capability of escaping avoidable morbidity
and premature mortality. In the present paper, income will serve as a proxy for these
capabilities. I assume that one should have at least two third of the median of the net
household income for elementary human functioning. And, in addition one should at least
have a fair state of health. This has been measured in the GSOEP from a scale from 1 (none)
to 3 (strong).
What exactly the capabilities for the more complex functionings are, is far more
difficult to determine. In order to operationalise Sen�s account of quality of life, I decided to
use the actual functioning in a range of areas of life as proxies for the capabilities. These are
indeed poor proxies because it leaves aside the choice element. That is: it leaves aside
whether for instance one chooses not to participate in local politics or to attend to cultural
events. Hence, if one does not participate in politics, the conclusion could be wrong that this
person lacks the capability. Nevertheless, for participating in local politics one should at least
have the capability. By observing the actual functioning, one captures at least some of the
capabilities. Pending further inquiry, I decided to measure the more complex capabilities in
the way mentioned above.
The GSOEP-dataset consists of several variables which measure from a scale from 1 (every
week) to 4 (never) the extent to which one attends to social events. The following variables
will be used for measuring the extent one is socially integrated:
• attend cultural events
• attend cinema, pop, jazz concerts
• participate in sports
• attend social gatherings
• help out friends, relatives, acquaintances
• participate in local politics
• attend church or other religious events
The construction of the variables mentioned above will appear in the next sections under the
name degree of social integration.
14
The following variables will be used to measure the extent to which one is capable to
develop other activities than working:
• Hours running errands
• Hours on housework
• Hours on childcare
• Hours on repairs/yard work
• Hours on other activities
• Hours on training, education
This construction will be called leisure.
All variables taken together � the basic functionings, the social functionings, and the
individual functionings � reflect Sen�s account of the quality of life. The name of this variable
is capabilities. In sections VII the relationship between the liberal accounts of life-quality and
non-standard work-arrangements will be estimated.
VI. RESULTS FOR THE UTLITARIAN ACCOUNT OF LIFE-QUALITY
In this section the question will be addressed whether working with a non-standard work-
arrangement has severe implications for the well-being or satisfaction of the employees
involved. Five different variables will be used for estimating the impact. Firstly, the variables
which measure satisfaction: the aggregate of the mentioned domain satisfaction-variables, the
satisfaction with life today � variable and the satisfaction-with-life-in-the-past-five-years �
variable. Furthermore, an attempt will be made to measure feelings of stress, control over life
and the amount of worries along the lines described in the former section.
Table 2 shows by work-arrangement the means of the three variables which measure
life-satisfaction and the means of the variables which measure stress (overall mean = 1.0).
15
Table 2: means well-being by work-arrangementQuality of
life-measure
Type of workarrangement
Standardizedaggregatedsatisfaction
standaridized satisfactionwith lifetoday
standardizedsatisfactionwith life infive years
standardizedoverallcontrol
Overallworries
standardizedstress
Parttime andtemporary
Mean 1.01 0.99 1.00 1.00 0.93 0.96
N 97 97 97 90 97 90Std.Deviation
0.23 0.24 0.31 0.13 0.28 0.17
Parttime andpermanent
Mean 1.10 1.02 1.01 0.99 1.03 1.01
N 939 939 939 924 939 924Std.Deviation
0.23 0.24 0.28 0.14 0.23 0.15
Fulltime andtemporary
Mean 0.94 0.94 0.95 0.98 0.87 0.93
N 227 227 227 227 227 227Std.Deviation
0.25 0.26 0.31 0.14 0.28 0.18
Self-employed
Mean 0.95 0.99 1.04 1.01 0.94 0.98
N 594 594 594 587 594 587Std.Deviation
0.26 0.25 0.28 0.14 0.30 0.19
Fulltime andpermanent
Mean 0.99 1.01 1.00 1.01 1.02 1.01
N 3861 3861 3861 3830 3861 3830Std.Deviation
0.25 0.24 0.29 0.14 0.23 0.15
Total Mean 1.01 1.00 1.00 1.00 1.01 1.00N 5719 5719 5719 5658 5719 5658Std.Deviation
0.25 0.24 0.29 0.14 0.24 0.16
(Source: own computions on GSOEP)
In table 2 we see that employees with full-time temporary contracts score on all
measures lower than employees with permanent contracts. These differences are all
statistically significant. Furthermore, table 2 shows that part-time employees with permanent
contracts score significant higher than full-time employees with permanent contracts on the
aggregated satisfaction (t-statistic=12.08). They score, however, significantly lower on the
�control over life�-variable (t-statistic=-3.09).
In section III, we saw that part-time and/or temporary employees were primary located
in the low-skill, low paid and low status jobs. In order to measure the impact of only the type
of work-arrangement on the quality of life, one should control for these labour-market
characteristics which influence the quality of life. Therefore, I will use the regression
technique. I will estimate the following model:
µ = α + βX + βZ (1)
16
where µ is subsequently the aggregated satisfaction, the satisfaction with life today, the
expected satisfaction with life the next five years, the control over life, the amount of worries,
and the amount of stress, α is the constant, X the type of working-arrangement and Z includes
dummies for gender, age, children, industry, profession, income and federal state.
In the following tables the α and the β�s are estimated in the whole sample and in different
sub-samples. Table 3 reports the estimates for the whole sample.
Table 3: Regression on satisfaction-measuresaggregated happiness satisfaction with life today satisfaction with life in five
yearscontrol worries stress
B T B T B T B T B T B T(Constant) 1.38*** 10.95 1.41*** 11.71 1.35*** 9.38 1.18*** 16.54 1.26 11.05 1.22*** 16.09Reference:full-time,permanent,maleparttime andtemporary
B T B T B T B T B T B Tone child -0.02** -1.98 0.00 -0.23 0.01 0.75 0.01 1.57 -0.01 -1.17 0.00 -0.15two children -0.05*** -3.31 -0.01 -0.72 0.01 0.77 -0.01 -0.97 -0.01 -0.73 -0.01 -1.01threechildren
B T B T B T B T B T B TBremen -0.08* -1.76 0.11** 2.55 0.13** 2.51 0.04 1.48 0.07* 1.78 0.06** 2.04Hesse 0.00 -0.09 -0.03* -1.70 -0.05** -2.54 -0.01 -1.08 -0.01 -0.56 -0.01 -0.93Rheinland,Pfalz, Saar
SEE 0.21 0.21 0.14 0.14 (Source: own computations on GSOEP)
Table 4 panel A to C shows that particularly men dislike working part-time and/or
temporary. They score significantly less on the aggregate of the various domains, they are less
satisfied with life today, and they have more worries than women. Women are surprisingly
happy with their part-time work-arrangements: the effects of working part-time are positive
for the aggregated satisfaction and the future satisfaction. Furthermore, women working part-
time seem to be less worried than men.
Table 5 reports the estimates by education. The question arises whether there are any
significant differences between high educated or low-educated employees. Whilst it could be
expected that the high educated are to a lesser degree forced to work part-time and/or
temporary. Hence, they work more often voluntary with non-standard work-arrangements.
Therefore, the expectation is that the high-educated employees are more satisfied with their
non-standard work-arrangements. Whether this statement holds, could be answered with the
help of table 5.
24
Table 5: regression on welfare measures by education
Panel Aaggregated happiness satisfaction with life todayless than high school more than high school less than high school more than high schoolB T B T B T B T
Panel Bsatisfaction with life in five years controlless than high school more than high school less than high school more than high schoolB T B T B T B T
Primary goods Primary goods + leisure Degree of social integration Basic and socialfunctionings
Capabilities (incl leisure)
B T B T B T B T B Tone child 0.00 0.02 0.10*** 4.62 0.02*** 2.82 0.04*** 7.89 0.12*** 8.31two children 0.00 -0.06 0.16*** 6.23 0.04*** 4.11 0.06*** 10.28 0.20*** 11.26Threechildren
Primary goods Primary goods + leisure Degree of social integration Basic and socialfunctionings
Capabilities (incl leisure)
B T B T B T B T B TBremen 0.02 0.11 -0.09 -1.12 -0.01 -0.27 -0.01 -0.80 -0.11** -1.97Hesse 0.07 1.42 0.04 1.19 -0.02* -1.80 -0.01 -1.22 0.00 -0.13Rheinland,Pfalz, Saar
Panel BDegree of social integration Basic + social functioningsless than high school more than high school Less than high school more than high schoolB T B T B T B T