Multivariate analysis results: We have so far tried to establish the association between women's happiness level and various indicators of women's autonomy as well as her socio-economic, cultural and place of residence.
In order to measure the net impact of each autonomy variable on women's happiness we applied Multinomial Logistic Regression technique. The Model-i includes autonomy variables only. In the second step (Model-2) some socio-economic, regional and cultural variables were added to Model-i. The inclusion of these as control variables was necessitated because of their role in explaining women's happiness. This way the results will portray a more exact picture of autonomy variables in explaining women's happiness in life.
Job Status
The majority of the variables show expected results. One exception is Women's labour force participation, which was entered in the model as a dichotomous variable where non-working women are the reference category. Women who work are hypothesised to be more autonomous than those who do not work. It is further conjectured that these autonomous women are happy with their life. Interestingly, multivariate analysis results show an inverse relationship of women's work participation with their happiness level. In other words, women's work participation makes them dissatisfied and unhappy in life (Table 2, Equation 1).
This relationship becomes statistically significant and even stronger in case of the equation 'very happy' relative to 'quite happy' (see Table 2 Equation 2). The direction of relationship and the statistical significance does not change with the addition of socio-economic and cultural variables-hereafter called the 'control variables'-in the Model-2.
The possible explanation for this contrary to expected relationship is that many Pakistani women with their low socio-economic status and poverty experience financial constraints and seek work out of need. They are mostly are involved in low-income menial jobs [Shah (1986); Ali, Siyal, and Sultan (1995); Nasir and Kiani (2005)].
Ironically, the money earned by them is usually taken over by husband or spent to cater for the needs of other household members with almost no money left for themselves that could enhance their status. Moreover, most Pakistani women, besides their gainful work, have to bear the double burden of household and work responsibilities single handedly, which is likely to make them feel unhappy. Due to the above stated reasons, working women report of not having a 'very happy' status in life.
EDUCATION:
Education is another indicator of autonomy, the attainment of which is conjectured to be a source of happiness among women. The results of the multinomial regression analysis (Table 3) show that in general, educational attainment is related to happiness in a woman's life. Interestingly, educational attainment up to primary level brings about a maximum change in their happiness status.
The reason for declining odds of 'very happy' status at higher levels of education could be that the level of consciousness and expectations of those with high education is higher as compared to the ones with primary education, and it is likely that highly educated women do not receive the returns to education as expected. The inclusion of socio-economic, cultural and place of residence variables in the Model-2 has a depressive effect on the magnitude of coefficients at all the three levels of education. Yet the relationship remains statistically significant and in the same direction (Table3, Equation 2). To sum up, one may conclude that attainment of education by women, especially primary level is related to happiness.
FREEDOM OF MOVEMENT:
A woman's happiness in life is also measured here, by the level of her freedom of movement outside home-an indicator of women's autonomy. This variable shows a weak inverse relationship with the dependent variable, ie, women's happiness in life (Table 2, Equation 1 and 2). The addition of control variables in Model-2 (Table 3, Equations 1 and 2) reverses the relationship. However, it remains unimportant as the relationship is statistically not significant.
DECISION-MAKING POWER:Women's decision-making power in the household conforms to the expected outcome, that is, the decision-making power of women is associated with 'very happy status' in their lives relative to 'quite happy status'. The relationship remained statistically highly significant and the value of coefficients and odd ratios remains almost unchanged when controlling variables are added. However, this variable remains statistically insignificant in both the models for its relationship with 'not happy' relative to 'quite happy'.
ACCESS AND CONTROL OF ASSETS:
The degree of women's access to material resources such as land, house/flat, vehicle, jewellery, bank deposit and consumer durables and control over their utilisation is conjectured to provide not only socio-economic status to women, but in turn generate happiness in their lives. This relationship when tested through logistic regression model conforms to this hypothesis.
The relationship is not only statistically highly significant at a less than 1 percent level but the magnitude of these coefficients is also quite large (Table 2, Equation 1). However, interestingly in Equation 2 where as the possession of assets remained an important factor in explaining the relationship between ~'very happy" relative to "quite happy" status of women; the ability and control over asset utilisation loses its statistical significance.
The addition of controlling variables in Table 3 does not bring much change except for the relationship between 'very happy' relative to 'quite happy' status of women (Table 3, Equation 2). This relationship loses the statistical significance for the women who have not only the possession but control over the utilisation of assets.
In Pakistani culture, the supremacy of women is generally not tolerated by men, so all women who possess and have control over these assets have to face more pressures and criticism from men folk. This attitude of men may make these women less happy than those who possess the assets only.
AGE OF WOMEN:
The age of women, entered in the model at interval scale shows an inverse relationship with women's happiness in life. This inverse relationship suggests that as women grow older, mounting pressures of responsibilities result in more anxiety, rendering them relatively discontented and less happy in life. Although the magnitude of the coefficient is small, it is statistically significant (Table 3, Equations 1 and 2).
POVERTY:
In recent years, there has been an increase in the poverty level in Pakistan, from 17 percent in 1987-88 to 33 percent in 2001 [Qureshi and Arif (2001)]. The increase in the levels of poverty undoubtedly affects the lives of people particularly women who are mostly dependent on the earnings of male members of the household. The variable of poverty is included in the model for two reasons: firstly, to measure the relationship of this variable with happiness of women and secondly, to control the effect of poverty so as to test the net affect of autonomy variables.
As expected, the poverty level denoted here in terms of "above poverty line" and "on or below poverty line" emerged as not only a significant variable to explain the happiness level of women but in terms of magnitude of the coefficient, it also brings about a substantial change in the dependent variable ie, 'not happy' relative to 'quite happy' as well as 'very happy' relative to 'quite happy' status of women. The result clearly suggests that financial well-being generates happiness in the lives of women.
REGION OF RESIDENCE:
There is a great divide in urban and rural areas. The region of residence has been included in the analysis, so as to control the impact of the differences in the urban and rural areas. The results of the analysis show that as compared to rural women, urban women are 'quite happy' relative to 'not happy' women. But this relationship is not statistically significant. Likewise, urban life also brings about a 'very happy' status relative to 'quite happy status among women (Table 3, Equation 2). However, the relationship is not statistically significant here as well.
CONFLICT/ABUSES FACED IN THE HOUSE:
Facing conflict and abuse is, unfortunately, common in Pakistani households [Sathar and Kazi (1997)]. Our analysis shows that conflict in the house affects women's happiness. In other words, women who do not face conflict or abuse in the house are more likely to be happy. For example, women who are not facing abuses inside the household are 14 percent less likely to be 'not happy' relative to 'quite happy' as compared to those who are facing abuses. The coefficient is significant at the 10 percent level of significance. Similarly, women not facing abuses are 66 percent more likely to be 'very happy' relative to 'quite happy' compared to those who are facing abuses but the coefficient is statistically not significant.
SICKNESS STATUS:
Although good health may not guarantee happiness but bad health certainly makes one feel unhappy and depressed. Pakistani women who are discriminated against men in almost every sphere of life get sick more often than their male counterparts [Mahmood and Ali (2002)]. Illness affects the wellbeing of the population at large but a woman is affected the most as because she has to do not only the household chores single-handedly but has to take additional responsibilities of bearing and rearing the children.
This is reflective in the results of this study which shows that as compared to sick women, healthy women are 'quite happy' in life. This relationship is very strong as evidenced by odds ratios as well as statistical significance level at less than 1 percent.
But somehow, sickness status of women does not yield a significant effect on the 'very happy' status of women (Table 3, Equation 2). This means that whereas good health guarantees a 'quite happy' status for Pakistani women, it does not necessarily bring about a 'very happy' status for women in Pakistan.
SANITATION:
The provision of potable water and improved sanitation is an important basic right of the population at large. But somehow various national level surveys as cited in of Pakistan, (2000) indicate that coverage and access to water supply facilities in Pakistan range between 50 to 80 percent and sanitation facilities between 40 to 55 percent.
The sanitation variable in this study is formulated on the basis of the availability status of flush or pour flush latrine and water in the house. If these facilities are available inside the house, a good sanitation condition is assumed. Having water and latrine facilities available inside the house is a matter of comfort for women as traditionally, fetching water from a source outside the house is a woman's job. Moreover, she also faces hardship if a toilet facility is not available inside the house. This is reflective in the results of this analysis as sanitation facilities proved to be an important factor in yielding 'quite happy' status for women in Pakistan.
CONCLUSION:
This study has been undertaken to examine whether or not the established autonomy indicators are a source of 'happiness' for Pakistani women. Only two indicators ie, "women's education" and "decision-making authority" proved to be important factors in soliciting 'very happy' status in women's life. Additionally, "possession of assets", also proved to be an important factor in providing 'very happy' status in a woman's life. However, the "possession and utilisation of assets" and "going alone outside the house" were not important indicators of a 'very happy' status in women's life in Pakistan and "Labour force participation" is indicative of unhappiness.
The results of this study show that not all established indicators of autonomy bring about happiness in the lives of Pakistani women. This is because Pakistani society differs from other societies, in particular the western society and hence the concept of 'autonomy' in bringing about 'happiness' in the lives of Pakistani women yields different effects than in other societies.
In order to better understand this relationship, one may have to study the contextual effects of Pakistani society which is influenced by a blend of the South Asian, Arab, and Western cultures.
The contentment, tolerance, happiness, self-denial, and sacrifice are generally the characteristics of people at large. The religious injunctions and rituals followed by a vast majority form a national character. For example fasting in the month of Ramadan, practised by an overwhelming majority, results in staunch self-control and self-denial.
A blend of the above stated influences with the patriarchal and male dominated nature of our society make a Pakistani woman to gain protection, comfort, and happiness living within their families without the need of adopting many western style autonomy indicators. Thus there is a need to focus on the advocacy of only those autonomy variables which lead to happiness in a woman's life. After all, the end goal should be to understand and promote the well-being and happiness of Pakistani women who form a vital part of society.
WOMEN'S AUTONOMY AND HAPPINESS:
MODEL-1
TABLE 2
LOGISTIC REGRESSION COEFFICIENTS AND ODDS RATIOS OF PREDICTOR VARIABLES ON WOMEN 'S HAPPINESS.
==================================================================================
Equation 1 Equation 2
Not Happy Relative to Quite Very Happy Relative to Quite
Happy Happy
==================================================================================
Predictor Variable Coefficients Odds Ratios Coefficients Odds Ratios
==================================================================================
Constant -0.448 -2.307
Job Status
No Job*
Doing Job 0.126 1.134 -0.473*** 0.623
Education
No Education*
1-5 -0.805*** 0.447 0.943*** 2.576
6-10 -0.642*** 0.526 0.647*** 1.91
11+ -0.298** 0.743 0.554*** 1.741
Going out Behaviour
Can't Go Alone*
Can Go Alone 0.098 1.102 -0.036 0.965
Decision-making
Noa
Yes 0.038 1.039 0.666*** 1.947
Assets
No Assets*
Yes but can't Use -0.557*** 0.573 0.575*** 1.777
Yes and can Use -0.79*** 0.454 0.306* 1.358
-2 Likelihood 596.8
Chi Square 293.15
N 3244
==================================================================================
Source : Original data file of PSES 2001.
a Reference category.
Significant at 1 level.
-- Significant at 05 level.
-- Significant at 0.1 level.
-- Ali and Hag
-- Model-2
TABLE 3 Logistic Regression Coefficients and Odd Ratios of Predictor Variables on Women 's Happiness after controlling for Socio-Economic Variables
==================================================================================
Equation I Equation 2
Not Happy Relative to Quite Very Happy Relative to Quite
==================================================================================
Predictor Variable Coefficients Odds Ratios Coefficients Odds Ratios
==================================================================================
Constant -0.515 -2.092
Job Status
No Job*
Doing Job 0.121 1.128 -0.454*** 0.635
Education
No Education*
1-5 -0474* 0.622 0.523** 1.687
6-10 -0.362** .696 0.343** 1.403
11+ -0.096 0.908 0.338** 1.410
Going out Behaviour
Can't Go Alonea
Can Go Alone -0.027 0.974 0.045 1.046
Decision-making
Noa
Yes -0.052 0.95 0.753*** 2.213
Assets
No Assetsa
Yes but can't Use -0.53*** 0.589 0.508*** 1.662
Yes and can Use -0.68*** 0.507 0.221 1.248
Age of the Women
Poverty Status
Below or on Poverty
Linea 0.0204*** 1.021 -0.0224*** 0.978
Above Poverty Line -0.177** 0.837 0.468*** 1.597
Region of Residence
Rurala
Urban -0.139 1.149 0.106 1.112
Abuses Faced
Yest
No -0.156* 0.855 0.505 1.657
Sickness
Yesa
No -0.436*** 0.647 -0.210 0.811
Sanitation
Bada
Good -0.24** 0.786 0.013 0.923
-2 Likelihood 5233.7
Chi-square 426.696
N 3244
==================================================================================
Source : Original data file of PSES 2001.
-- Reference category.
-- Significant at 1 level.
-- Significant at 05 level.
-- Significant at 0.1 level.
WOMEN 'S AUTONOMY AND HAPPINESS
APPENDIX TABLE 1
====================================================================================================================================================
Correlation Matrix
====================================================================================================================================================
Going
Abuses out Decision- Sick- Sanita- Area of Assets Age
Education Faced Alone making Job ness tion Residence Poverty
====================================================================================================================================================
Education 1
Abuses Faced 0.121 1
Going out Alone -0.021 0.015 1
Decision-making 0.001 0.03 0.21 1
Job -0.0587 -0.011 0.113 0.06 1
Sickness 0.042 0.008 -0.084 -0.035 0.009 1
Sanitation 0.329 0.083 -0.057 0.017 -0.153 -.0.001 1
Area Residence 0.335 0.026 -0012 0.038 -0.11 -0.032 0.489 1
Poverty 0.265 0.06 -0.049 0.005 -0.089 -0.041 0.243 0.241 1
Assets 0.216 0.037 0.024 0.142 -0.076 -0.006 0.185 0.161 0.161 1
Age 0.226 -0.002 -0.198 -0.133 -0.021 0.168 0.008 -0.019 -0.013 0.064 1
====================================================================================================================================================
(Concluded) - Courtesy The Pakistan Development Review.
Comments
Comments are closed.