Project for the Regional Advancement of Statistics in the Caribbean - PRASC

Transcription

1 Project for the Regional Advancement of Statistics in the Caribbean - PRASC

2 Gender-based Analysis: Understanding the gender gap in labour market outcomes Analysis Workshop - Module 6 2 March 21-24, 2016 Kingstown, Saint Vincent and the Grenadines

3 What is Gender-Based Analysis? An analytical tool for examining the potential impacts of policies, programs, initiatives, and legislation on diverse groups of women and men Takes into account gender and other identity factors Integrates social, economic, and other gender differences into policy analysis 3

4 Identity Factors: 4

5 What are gender statistics? Gender statistics go beyond disaggregation of data by sex. Rather than simply highlighting differences between men and women, gender statistics tells us how and why 5

6 Sex disaggregated statistics versus gender-based analysis Sex-disaggregated statistics Cardiovascular disease (CVD) tends to appear about 10 years later in women than in men Gender-based analysis Cigarette smoking, depression, low income, elevated serum lipid levels, hypertension, obesity and lack of physical activity may also be different for men and women These findings can help target policies and programs to better address patients needs. 6

7 Why are gender statistics important? Gender-based analysis can address major policy issues: Labour force participation, unemployment, workplace decision making and entrepreneurial opportunities Unpaid work and work/family balance Access to assets, childcare, education and health services Gender-based attitudes and violence Social exclusion and treatment of minority groups 7

8 Gender-based Analysis Data Integration Analysis can benefit from the integration of different data sources administrative, population-based and Census surveys. Best practice to integrate data from more than one survey Families, living arrangements, unpaid work Women and health Women and education Economic well-being, paid work Diversity of women: Senior Women, Women with Activity Limitations. 8

9 Analysis: Women in Canada The first edition of Women in Canada was published in 1985, the same year as the United Nations (UN) Third World Conference on Women in Nairobi. It provides an unparalleled compilation of data related to women's family status, education, employment, economic well-being, unpaid work, health, and more. Women in Canada s scope and purpose responded to the UN call for more reliable data to assess improvements in women s status in various sectors. It helps fulfill the Government of Canada s commitment to encourage Gender-Based Analysis (GBA). 9

10 Additional information Statistics Canada Women in Canada: A Genderbased Statistical Report. Cat. No X. Organisation for Economic Co-operation and Development (OECD) How s Life? Measuring Well-being. United Nations Economic Commission for Europe (UNECE) Developing Gender Statistics: A Practical Tool. 10

11 CARICOM GEI Model UN Women and CARICOM Regional Statistics Programme (RSP) to identify an approach to advance the measurement of gender equality The goal: to agree to indicators to monitor gender equality commitments Piloted in four countries Dominica, Grenada, Jamaica and Suriname. 11

12 Proposed Gender Indictors Economic Activity Labour force participation rate, unemployment rate, distribution of employment in various sectors, gender gap in wages, proportion of employed working part-time by sex and age Education Gender parity index of the gross enrolment ratio in primary, secondary, tertiary education Health Unmet need for family planning, maternal mortality ratio, out of pocket health expenditures by sex, adolescent birth rate Public Participation Women s share of government ministerial positions, women s share of managerial positions, share of female policy officers Human Rights Proportion of females subjected to physical/sexual violence in the last 12 months 12

13 Gender-based analysis An illustration using data from the Prince Edward Island case study 13

14 An example of gender-based analysis: the gender wage gap Sex-disaggregated statistics: Hourly wages for men: $22.68 Hourly wages for women: $21.43 Wage gap: $1.25 On average, women earn $1.25 (5.5%) less per hour than men. 14

15 Other differences between men and women are observed Characteristics of employees, Prince Edward Island Men Women % Age group 25 to 34 years old to 44 years old to 54 years old to 64 years old years and over Tenure (current job) Less than 24 months to 60 months months or more Source: Statistics Canada, Labour Force Survey (LFS) 15

16 Other differences between men and women are observed (cont d) Characteristics of employees, Prince Edward Island Men Women % Education Below a university degree University degree Region Charlottetown Rest of PEI Sector of the economy Public sector Private sector Source: Statistics Canada, Labour Force Survey (LFS) 16

17 These differences matter to the interpretation of the gender wage gap Average hourly wages (employees), by selected characteristics, Prince Edward Island Sector of the economy Average hourly wage Public sector $27.26 Private sector $19.03 Tenure (current job) Less than 24 months $ to 60 months $ months or more $ Education Below a university degree $19.66 University degree $28.70 Source: Statistics Canada, Labour Force Survey

18 Multivariate regressions: Overview Multivariate regression techniques are widely used as a data analysis tool at National Statistical Offices y i = α + β x x i + β z z i + u i Common techniques Ordinary Least Squares (OLS) regression models Panel regression techniques Probit and logit regression models (binary dependent variables) Survival analysis techniques 18

19 Multivariate regression as a data analysis tool Multivariate regression analysis: an essential tool for unpacking complex relationships in the data Often used to evaluate the strength of bi-variate relationships by simultaneously taking other relevant factors into account Not controlling for other relevant factors can distort basic relationships in the data 19

20 Multivariate regression: Examples from GBA Research questions: To what extent do the wages of men and women differ? What factors (observable and unobservable) help to explain observable differences in wage rates between men and women? Observable characteristics could include: economic sector, age of worker, tenure, education 20

21 Modelling the wage gap as a function of economic sector (bi-variate analysis) Yellow Wage gap Blue Economic sector Example adapted from: Kennedy, P A Guide to Econometrics. Cambridge: MIT Press. 21

22 Modelling the wage gap as a function of economic sector and job tenure (multivariate analysis) Wage gap Yellow Blue Red Green Brown Orange Economic sector Job tenure Example adapted from: Kennedy, P A Guide to Econometrics. Cambridge: MIT Press. 22

23 Modelling the wage gap as a function of economic sector and job tenure (multivariate analysis) Wage gap Yellow Blue Green Red Brown Orange Education Work experience Example adapted from: Kennedy, P A Guide to Econometrics. Cambridge: MIT Press. 23

24 Multivariate analysis Ordinary least squares (OLS) regression models: Specification A: wages Sex X Specification B: i 1 i 2 i ' i log(wages ) Sex X ' i 1 i 2 i i 24

25 Multivariate analysis Ordinary least squares regression models: Progressively incorporate explanatory variables and analyze how the coefficient associated with sex varies Model 1: X = sex Model 2: X = sex, sector Model 3: X = sex, sector, age Model 4: X = sex, sector, age, tenure Model 5: X = sex, sector, age, tenure, education 25

26 Ordinary Least Squares Regression Results Dependent variable: hourly wages Model A-1 Model A-2 Model A-3 Model A-4 Model A-5 Coefficients Coefficients Coefficients Coefficients Coefficients Intercept * * * * * Sex Men (ref) Women * * * * * Sector of the economy Private sector (ref) Public sector * * * Age group 25 to 34 years old (ref) to 44 years old * * * 45 to 54 years old * * 55 to 64 years old * years and over * * * Tenure (current job) Less than 24 months (ref) to 60 months * * 60 months or more * * Education Below a university degree (ref) - University degree * Adjusted R-square Note: (*) denotes statistically significant at the 5% significance level. Source: Statistics Canada, Labour Force Survey (LFS) 26

27 Ordinary Least Squares Regression Results Dependent variable: log(hourly wages) Model B-1 Model B-2 Model B-3 Model B-4 Model B-5 Coefficient Coefficient Coefficient Coefficient Coefficient Intercept * * * * * Sex Men (ref) Women * * * * * Sector of the economy Private sector (ref) Public sector * * * Age group 25 to 34 years old (ref) to 44 years old 0,099 * * * 45 to 54 years old 0,095 * * 55 to 64 years old -0, * years and over -0,140 * * * Tenure (current job) Less than 24 months (ref) to 60 months * * 60 months or more * * Education Below a university degree (ref) - University degree * Adjusted R-square Note: (*) denotes statistically significant at the 5% significance level. Source: 27 Statistics Canada, Labour Force Survey (LFS)

28 Findings Women in Prince Edward Island earned on average $1.25 (5.5%) less per hour than their men counterparts in December Women were much more likely to work in the public sector than men. Wages are higher and there tends to be more parity between wages of men and women in the public sector. Taking this factor into account, the observed gender wage gap in Prince Edward Island increases. Women are also more likely to be university educated and to have a tenure in their current job of five years or more, factors that should normally be associated with higher wages. 28

29 Additional information OLIVETTI, Claudia and Barbara PETRONGOLO The Evolution of Gender Gaps in Industrialized Countries. Center for Economic Performance, London School of Economics and Political Science. February. 29

30 30