Chapter 4 Gender Inequalities Beyond Averages: Between Social Norms

Chapter 4

Gender inequalities beyond averages: Between social norms and power imbalances

4. Gender inequalities beyond averages: Between social norms and power imbalances

Gender disparities remain among the most persistent forms of inequality across all countries.1 Given that these disadvantages affect half the world’s people, gender inequality is arguably one of the greatest barriers to human development. All too often, women and girls are discriminated against in health, in education, at home and in the labour market—with negative repercussions for their freedoms.

Progress in reducing gender inequality over Development Report’s Gender Inequality the 20th century was remarkable in basic Index—a measure of women’s empowerment achievements in health and education and par- in health, education and economic status— ticipation in markets and politics (!gure 4.1).2 shows that overall progress in gender inequali- Much of this progress was celebrated with the ty has been slowing in recent years.5 Beijing Platform for Action during the 1995 Consider two developments. First, gender Fourth World Conference on Women.3 But gaps are deeper than originally thought. Time as the event’s 25th anniversary approaches magazine’s 2017 Person of the Year was “the in 2020, many challenges to equality remain, silence breakers,” women who denounced particularly for enhanced capabilities that alter abuse. Accomplished women were unprotect- power relations and enhance agency. ed against persistent sexual abuse. #e silence #e world is not on track to achieve gender breakers were also given voice by the #MeToo The world is not on equality by 2030. Based on current trends, movement, which uncovered abuse and vul- track to achieve it would take 202 years to close the gender nerability for women, well beyond what is gender equality by 2030 gap in economic opportunity.4 #e Human covered in o$cial statistics. In Latin America,

FIGURE 4.1

Remarkable progress in basic capabilities, much less in enhanced capabilities

Enhanced Agency capabilities and change

Social Tradeoffs/ norms power imbalances

Subsistence and Basic participation capabilities

Source: Human Development Report Ofﬁce.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 147 too, the #NiUnaMenos movement has shed structural barriers to equality. #e tradeo&s light on femicides and violence against women are in%uenced strongly by social norms and from Argentina to Mexico.6 by a structure of mutually reinforcing gender Second, there are troubling signs of di$- gaps. #ese norms and gaps are not directly culties and reversals on the path towards gen- observable, so they are o'en overlooked and der equality—for female heads of state and not systematically studied. government and for women’s participation in the labour market, even where there is a buoyant economy and gender parity in access Gender inequality in to education.7 And there are signs of a back- the 21st century lash. In several countries the gender equality agenda is being portrayed as part of “gender Gender inequality is intrinsically linked to hu- ideology.”8 man development, and it exhibits the same dy- In other words, precisely when awareness is namics of convergence in basic capabilities and increasing more needs to be done to achieve divergence in enhanced capabilities. Overall, gender equality, the path becomes steeper. #is it is still the case—as Martha Nussbaum has chapter explores why progress is slowing, identi- pointed out—that “women in much of the fying today’s active barriers that pose challenges world lack support for fundamental functions for future prospects for equality, which include of a human life.”9 #is is evident in the Gender personal and public beliefs as well as practices Inequality Index and its components—re%ect- that generate biases against gender equality. It ing gaps in reproductive health, empowerment stresses that gender inequality re%ects intrinsic and the labour market. No place in the world imbalances in power—something well known has gender equality. In Sub-Saharan Africa 1 in to women’s movements and feminist experts— every 180 women giving birth dies (more than and documents two trends: 20 times the rate in developed countries), and • Gender inequalities are intense, widespread adult women are less educated, have less access and behind the unequal distribution of hu- to labour markets than men in most regions man development progress across levels of and lack access to political power (table 4.1). socioeconomic development. • Gender inequality tends to be more intense Gender inequality as a human in areas of greater individual empowerment development shortfall and social power. #is implies that progress Gender inequality is is easier for more basic capabilities and harder Gender inequality is correlated with a loss in correlated with a loss for more enhanced capabilities (chapter 1). human development due to inequality (fig- in human development #e !rst trend indicates the urgency in ad- ure 4.2). No country has reached low inequal- dressing gender inequality to promote basic ity in human development without restricting due to inequality human rights and development. #e second the loss coming from gender inequality. raises a red %ag about future progress. Progress Investing in women’s equality and li'ing both at the basics is necessary for gender equality, their living standards and their empowerment but it is not enough. are central to the human development agenda. Social norms and gender-speci!c tradeo&s “Human development, if not engendered, are key barriers to gender equality. Social and is endangered,” concluded the pioneer 1995 cultural norms o'en foster behaviours that Human Development Report, based on similar perpetuate inequalities, while power concen- evidence.10 trations create imbalances and lead to capture Today looks di&erent from 1995. #e 1995 by powerful groups such as dominant, patri- Human Development Report noted sizeable archal elites. Both a&ect all forms of gender gender disparities, larger than today’s, but doc- inequality, from violence against women to umented substantial progress over the preced- the glass ceiling in business and politics. In ing two decades, particularly in education and addition, gender-speci!c tradeo&s burden the health, where the prospect of equality was complex choices women encounter in work, visible. #e conclusion: “#ese impressions are family and social life—resulting in cumulative cause for hope, not pessimism, for the future.”11

148 | HUMAN DEVELOPMENT REPORT 2019 TABLE 4.1

Gender Inequality Index: Regional dashboard

Maternal mortality Adolescent Share of ratio birth rate seats in Gender (deaths per (births per parliament Population with at least Labour force Inequality 100,000 live 1,000 women (% held by some secondary education participation rate Index births) ages 15–19) women) (% ages 25 and older) (% ages 15 and older) Female Male Female Male Region 2018 2015 2015–2020 2018 2010–2018 2010–2018 2018 2018

Arab States 0.531 148.2 46.6 18.3 45.9 54.9 20.4 73.8 East Asia and the Pacific 0.310 61.7 22.0 20.3 68.8 76.2 59.7 77.0 Europe and Central Asia 0.276 24.8 27.8 21.2 78.1 85.8 45.2 70.1 Latin America and the Caribbean 0.383 67.6 63.2 31.0 59.7 59.3 51.8 77.2 South Asia 0.510 175.7 26.1 17.1 39.9 60.8 25.9 78.8 Sub-Saharan Africa 0.573 550.2 104.7 23.5 28.8 39.8 63.5 72.9

Source: Human Development Report Ofﬁce (see Statistical table 5)

FIGURE 4.2 intense, and progress towards gender equality is slowing (!gure 4.3). #e space for gains based Gender inequality is correlated with a loss in on current strategies may be eroding, and un- human development due to inequality less the active barriers posed by biased beliefs Loss in human development due and practices that sustain persistent gender to gender inequality (percent) inequalities are addressed, progress towards 90 equality will be far harder in the foreseeable future.

60 Gender inequality and empowerment: Catching up in the basics, widening gaps in enhanced capabilities 30 Accumulating capabilities requires achievements of different natures. As chapter 1 0 discussed, progress in human development is 0 10 20 30 40 50 linked to expanding substantive freedoms, ca- Inequality in Human Development Index distribution (percent) pabilities and functionings from basic to more enhanced. Progress towards equality tends to On the positive side Note: Countries mapped by their Gender Inequality Index performance relative be faster for basic capabilities and harder for to their performance on the Inequality-adjusted Human Development Index. The women are catching higher the loss due to gender inequality, the greater the inequality in human enhanced capabilities. Gender equality–related development. up in basic areas of Source: Human Development Report Ofﬁce. capabilities follow a similar pattern. On the positive side women are catching up development. But in basic areas of development. Legal barriers progress has been Today, the prospects are di&erent. #e past to gender equality have been removed in most two decades have seen remarkable progress in countries: Women can vote and be elected, uneven as women education, almost reaching parity in average they have access to education, and they can par- pull away from basic primary enrolment, and in health, reducing the ticipate in the economy without formal restric- areas into enhanced global maternal mortality ratio by 45 percent tions. But progress has been uneven as women since 2000.12 But gains in other dimensions pull away from basic areas into enhanced ones, ones, where gaps of women’s empowerment have not been as where gaps tend to be wider. tend to be wider

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 149 FIGURE 4.3

Progress towards gender equality is slowing

Gender Inequality Index (mean value) 0.500

0.400

0.300

0.000 1995 2000 2005 2010 2011 2012 2013 2014 2015 2016 2017 2018

Source: Human Development Report Ofﬁce (see Statistical table 5).

#ese patterns can be interpreted as re%ect- portfolios in a&airs such as transport, econom- Women make greater ing the distribution of individual empower- ics or !nance. Certain disciplines are typically and faster progress ment and social power: Women make greater associated with feminine or masculine charac- where their individual and faster progress where their individual em- teristics, as also happens in education and the powerment or social power is lower (basic capa- labour market. empowerment or bilities). But they face a glass ceiling where they Economic participation also shows a gra- social power is lower have greater responsibility, political leadership dient (see figure 4.4, right panel). When (basic capabilities). and social payo&s in markets, social life and empowerment is basic and precarious, women politics (enhanced capabilities) (!gure 4.4). are over-represented, as for contributing fam- But they face a This view of gradients in empowerment is ily workers (typically not receiving monetary glass ceiling where closely linked to the seminal literature on basic payment). #en, as economic power increases they have greater and strategic needs coming from gender plan- from employee to employer, and from employ- responsibility, political ning (box 4.1). er to top entertainer and billionaire, the gender Take access to political participation (see gap widens. leadership and social !gure 4.4, le' panel). Women and men vote in Empowerment gradients appear even for payoffs in markets, elections at similar rates. So there is parity in a uniform set of companies, as with the gen- social life and politics entry-level political participation, where power der leadership gap in S&P 500 companies. is very di&used. But when more concentrated Although women’s overall employment by (enhanced capabilities) political power is at stake, women appear these companies might be close to parity, severely under-represented. The higher the women are under-represented in more senior power and responsibility, the wider the gender positions. gap—and for heads of state and government it In developing countries most women who is almost 90 percent. receive pay for work are in the informal sec- Similar gradients occur even for women who tor. Countries with high female informal reach higher power. Only 24 percent of na- work rates include Uganda, Paraguay, Mexico tional parliamentarians were women in 2019,13 and Colombia (!gure 4.5), where more than and their portfolios were unevenly distributed. 50 percent of women are protected by minimal Women most commonly held portfolios in regulations; have few or no bene!ts; lack voice, environment, natural resources and energy, social security and decent work conditions; and followed by social sectors, such as social af- are vulnerable to low salaries and possible job fairs, education and family. Fewer women had loss.

150 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.4

The greater the empowerment, the wider the gender gap

Global gender gap in politics Global gender gap per type of employment (gap with respect to parity, percent) (gap with respect to parity, percent) 100 100

80 80 60 60 40

40 20

0 20 -20 0 -40 Votea Lower house Upper house Speakers of Head of Contributing Own Employees Employers Top 100 Top 500 or single parliament government family account entertainers billionaires house workers

Basic Enhanced Basic Enhanced a. Assumes an equal proportion of men and women in the voting population. Source: Human Development Report Ofﬁce calculations based on data from the World Values Survey, the Inter-Parliamentary Union, ILO (2019b) and Forbes (2019).

BOX 4.1

Practical and strategic gender interests and needs

The notion of practical and strategic gender interests and relations, such as a law condemning gender-based and needs (pioneered by Caroline Moser),1 which in- violence, equal access to credit, equal inheritance and forms much of the gender policy analysis framework, others. Addressing these should alter gender power is connected to the conception of basic and enhanced relations. Sometimes practical and strategic needs capabilities and achievements in this Report. As ar- coincide—for example, the practical need for child ticulated in gender social policy analyses,2 practical care coincides with the strategic need to get a job out- gender needs refer to the needs of women and men side the home.3 The difference is comparable to that to make everyday life easier, such as access to water, between basic and enhanced capabilities discussed better transportation, child care facilities and so on. in this Report. Transformative changes that can bring Addressing these will not directly challenge gender about normative and structural shifts are the strongest power relations but may remove important obstacles predictors of practical and strategic interventions ex- to women’s economic empowerment. Strategic gender panding women’s agency and empowerment for gen- needs refer to needs for society to shift in gender roles der equality.

Notes 1. Molyneux 1985; Moser 1989. 2. Moser 1989. 3. SIDA 2015.

Today, women are the most quali!ed in women’s reproductive roles (see Dashboard 2 history, and newer generations of women in the statistical annex), revealing one of the have reached parity in enrolment in primary moving targets discussed in chapter 1. Some education.14 But it now seems that this is not represent a natural part of the process of de- enough for achieving parity in adulthood. velopment—the constant need to push new #e transition from the education system to boundaries to achieve more. Others represent the world of paid work is marked by a gen- the response of deeply rooted social norms to der equality discontinuity, associated with preserve the underlying structure of power.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 151 FIGURE 4.5

The percentage of informal employment in nonagricultural employment in developing countries is generally higher for women than for men

Informal employment in nonagricultural Male Female employment, 2017 (percent) 90

0 Uganda Paraguay Mexico Colombia Costa Thailand Chile South Turkey Russian Serbia Ukraine Moldova Greece Rica Africa Federation (Rep. of)

Source: ILO 2019b.

Are social norms and power in many countries still cannot reach their full imbalances shifting? potential.19 Beliefs about what others do and what others Gender inequality has long been associated with think a person in some reference group should persistent discriminatory social norms prescrib- do, maintained by social approval and disap- Gender inequality has ing social roles and power relations between proval, o'en guide actions in social settings.20 long been associated men and women in society.15 Social norms held So it is useful to measure the beliefs and atti- with persistent by individuals and their reference groups are tudes that create biases and prejudices towards values, beliefs, attitudes and practices that assert women’s empowerment in society. discriminatory social preferred power dynamics for interactions be- Social norms cover several aspects of an in- norms prescribing tween individuals and institutions.16 As broader dividual’s identity—gender, age, ethnicity, reli- social roles and power constructs, norms are operationalized through gion, ability and so on—that are heterogeneous beliefs, attitudes and practices.17 and multidimensional. Discriminatory social relations between men People’s expectations of individuals’ roles norms and stereotypes reinforce gendered iden- and women in society in households, communities, workplaces and tities and determine power relations that con- societies can determine a group’s functioning. strain women’s and men’s behaviour in ways Women o'en face strong conventional societal that lead to inequality. Norms in%uence expec- expectations to be caregivers and home makers; tations for masculine and feminine behaviour men similarly are expected to be breadwin- considered socially acceptable or looked down ners.18 Embedded in these social norms are on. So, they directly a&ect individuals’ choices, longstanding patterns of exclusion from house- freedoms and capabilities. hold and community decisionmaking that Social norms also re%ect regularities among limit women’s opportunities and choices. So, groups of individuals. Rules of behaviour are despite convergence on some outcome indica- set according to standards of behaviour or tors—such as access to education at all levels ideals attached to a group’s sense of identity.21 and access to health care—women and girls Individuals have multiple social identities and

152 | HUMAN DEVELOPMENT REPORT 2019 behave according to identity-related ideals, freedom, and beliefs about social censure and they also expect others sharing a common and reproach create barriers for individuals identity to behave according to these ideals. who transgress. For gender roles these beliefs Norms of behaviour related to these ideals can be particularly important in determining a&ect people’s perception of themselves and the freedoms and power relations with other others, thus engendering a sense of belong- identities—compounded when overlapping ing to particular identity groups. #e beliefs and intersecting with those of age, race and people hold about appropriate behaviour class hierarchies (box 4.2). often determine the range of choices and How prevalent are biases from social norms? preferences that they exercise—in that con- How are they evolving? How do they a&ect text norms can determine autonomy and gender equality? #ese are di$cult questions,

BOX 4.2

Overlapping and intersecting identities

When gender identities overlap with other identities, norms and stereotypes of exclusion can be associated Overlapping identities they combine and intersect to generate distinct prejudic- with different identities. For instance, regarding medi- must be considered in es and discriminatory practices that violate individuals’ an years of education completed in Angola and United equal rights in society. Intersectionality is the complex, Republic of Tanzania, an important gap distinguishes research and policy cumulative way the effects of different forms of discrim- women in the highest wealth quintile from those in the analysis because ination combine, overlap or intersect—and are amplified second or lowest quintile (see figure). If the differences different social norms when put together.1 A sociological term, intersectionality are not explicitly considered, public programmes may refers to the interconnected nature of social categories leave women in the lowest quintiles behind. and stereotypes of such as race, class, gender, age, ethnicity, ability and Moreover, individuals’ different social identities exclusion can be residence status, regarded as creating overlapping and can profoundly influence their beliefs and experiences interdependent systems of discrimination or disadvan- about gender. People who identify with multiple minori- associated with tage. It emerges from the literature on civil legal rights. ty groups, such as racial minority women, can easily be different identities It recognizes that policies can exclude people who face excluded and overlooked by policies. But the invisibil- overlapping discrimination unique to them. ity produced by interacting identities can also protect Overlapping identities must be considered in re- vulnerable individuals by making them less prototypical search and policy analysis because different social targets of common forms of bias and exclusion.2

How gaps in median years of education distinguish rich from poor in Angola and United Republic of Tanzania, 2015

9.2

7.0 6.5 6.5 6.4 6.6 6.0 6.2 4.9 4.4

1.4

Total Lowest Second Middle Fourth Highest Total Lowest Second Middle Fourth Highest 15–49 quintile quintile 15–49 quintile quintile

Angola United Republic of Tanzania

Note: Lowest quintile refers to the poorest 20 percent; highest quintile refers to the wealthiest 20 percent. Source: Demographic and Health Surveys.

Notes 1. IWDA 2018. 2. Biernat and Sesko 2013; Miller 2016; Purdie-Vaughns and Eibach 2008.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 153 mainly because social norms and attitudes are (!gure 4.9). At the other extreme, indicating a hard to observe, interpret and measure. But backlash, the share of men with no bias fell in using data from the World Values Survey wave 5 Sweden, Germany, India and Mexico. (2005–2009) and wave 6 (2010–2014), a social The share of women with no gender so- norms index can be constructed to capture how cial norms bias increased the most in the social beliefs can obstruct gender equality along Netherlands, Chile and Australia. But most multiple dimensions (!gure 4.6 and box 4.3). countries in the sample showed a backlash, led by Sweden, India, South Africa and Romania Widespread biases and backlash (see !gure 4.9).

The multidimensional gender social norms Gender inequality and social norms count index and high-intensity index (see box 4.3) show widespread biases in gender The multidimensional gender social norms The multidimensional social norms. According to the count index, indices appear linked to gender inequality, as only 14 percent of women and 10 percent of might be expected. In countries with higher bi- gender social norms men worldwide have no gender social norm ases (measured through the multidimensional indices appear linked bias (!gure 4.7). Women are skewed towards gender social norms indices), overall inequality to gender inequality. less bias against gender equality and women’s (measured by the Gender Inequality Index) is empowerment. Men are concentrated in the higher (!gure 4.10). Similarly, the indices are In countries with middle of the distribution, with 52 percent positively related to the Gender Inequality higher biases, overall having two to four gender social norms biases. Index in time spent on unpaid domestic chores inequality is higher #e high-intensity index shows that more than and care work. half the world’s people have a high-intensity Biases in social norms also show a gradient. bias against gender equality and women’s #e political and economic dimensions of the empowerment. multidimensional gender social norms index Both indices provide evidence of a stagnation indicate biases for basic women’s achievement or a backlash from 2005–2009 to 2010–2014. and against more enhanced women’s achieve- #e share of both women and men worldwide ment (!gure 4.11). Overall, the biases appear with no gender social norms bias fell (!gure 4.8). more intense for more enhanced forms of Progress in the share of men with no gender women’s participation. #e proportion of peo- social norms bias was largest in Chile, Australia, ple favouring men over women for high-level the United States and the Netherlands political and economic leadership positions is

FIGURE 4.6

How social beliefs can obstruct gender and women’s empowerment

Dimensions Political Educational Economic Physical integrity

Men make better Women have University is more Men should have Men make better Proxy for Proxy for Indicators political leaders the same rights important for a man more right to a business executives intimate reproductive than women do as men than for a woman job than women than women do partner violence rights

Dimension index Political empowerment Educational empowerment Economic empowerment Physical integrity index index index index

Multidimensional gender social norms index

Source: Mukhopadhyay, Rivera and Tapia 2019.

154 | HUMAN DEVELOPMENT REPORT 2019 BOX 4.3

The multidimensional gender social norms index—measuring biases, prejudices and beliefs

Research prepared for this Report proposed the multidimensional gender so- includes 77 countries and territories accounting for 81 percent of the world cial norms index to capture how social beliefs can obstruct gender equality population. The second set consists of only countries with data for both along multiple dimensions. The index comprises four dimensions—political, wave 5 and wave 6. This set includes 32 countries and territories accounting educational, economic and physical integrity—and is constructed based for 59 percent of the world population. on responses to seven questions from the World Values Survey, which are used to create seven indicators (see figure 4.5 in the main text). The an- Definition of bias for the indicators of the multidimensional gender swer choices vary by indicator. For indicators for which the answer choices social norms index are strongly agree, agree, disagree and strongly disagree, the index defines Dimension Indicator Choices Bias definition individuals with a bias as those who answer strongly agree and agree. For Men make better Strongly agree, Strongly agree the political indicator on women’s rights, for which the answer is given on political leaders agree, disagree, and agree a numerical scale from 1 to 10, the index defines individuals with a bias as than women do strongly disagree Political those who choose a rating of 7 or lower. For the physical integrity indicators, Women have the 1, not essential, Intermediate for which the answer also ranges from 1 to 10, the index defines individuals same rights as to 10, essential form: 1–7 with a bias using a proxy variable for intimate partner violence and one for men reproductive rights. University is Strongly agree, more important Strongly agree Educational agree, disagree, Aggregation for a man than and agree strongly disagree For each indicator a variable takes the value of 1 when an individual has a for a woman bias and 0 when the individual does not. Two methods of aggregation are Men should have Agree, neither, Strongly agree then used in reporting results on the index. more right to a disagree and agree The first consists of a simple count (equivalent to the union approach), job than women where the indicators are simply summed and therefore have the same Economic Men make Strongly agree, weight. This result has a minimum of 0 and a maximum of 7: better business agree, disagree, Agree The calculation is a simple addition of dichotomic variables, but it com- executives than strongly disagree plicates the disaggregation and analysis by dimension and indicator. women do To address this, the second method follows the Alkire–Foster methodol- Proxy for 1, never, to 10, Strongest form: ogy,1 which counts the different gender social norm biases that an individual intimate partner always 2–10 faces at the same time (following the intersection approach). These dimen- violence Physical integrity sions are analysed to determine who has a bias on each indicator. This result Proxy for 1, never, to 10, counts only people with high-intensity bias. reproductive Weakest form: 1 always The methods are applied to two sets of countries. The first set consists rights of countries with data for either wave 5 (2005–2009) or wave 6 (2010–2014) of the World Values Survey and uses the latest data available. This set Source: Mukhopadhyay, Rivera and Tapia 2019.

Note 1. Alkire and Foster 2011. Source: Mukhopadhyay, Rivera and Tapia 2019. higher than the proportion of people favouring potential and are much more frequent among men over women in access to basic political men than women.22 rights or paid employment. Gradients in biases are likely to a&ect elec- Several theories linked to social norms could tions and economic and family decisions, mak- account for these di&erences. One suggests an ing gender equality more di$cult to reach when inability to discern between con!dence and higher levels of empowerment are at stake. competence. If people misinterpret con!dence as a sign of competence, they can mistakenly What causes change—and what believe that men are better leaders than women determines the nature of change? when men are simply more con!dent. In other words, for leadership the only advantage that How can practices and behaviours either men have over women is that manifestations change or sustain traditional gender roles? of overcon!dence, o'en masked as charisma or Norms can change as economies develop, by charm, are commonly mistaken for leadership changes in communications technology, by

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 155 FIGURE 4.7 would miss a deeper understanding of social change.24 Only 14 percent of women and 10 percent of men Consider the subtle differences between worldwide have no gender social norms biases descriptive and injunctive norms.25 Descriptive Percent of surveyed population responding norms are beliefs about what is considered a with biases towards gender equality normal practice in a social group or an area. and women’s empowerment Injunctive norms state what people in a com-

Female Male munity should do. #is distinction is important 50 52 for practice, as it can lead to an understanding of why some aspects of gender norms and relations shi' faster than others.26 24 18 21 The family sets norms, and experiences 14 from childhood create an unconscious gen- 10 12 der bias.27 Parents’ attitudes towards gender in%uence children through mid-adolescence, No gender 1 bias 2–4 biases More than 28 biases 5 biases and children at school perceive gender roles. Parenting practices and behaviours are thus Note: Balanced panel of 77 countries and territories with data from wave 6 among the predictors of an individual’s (2010–2014) of the World Values Survey, accounting for 81 percent of the world population. gendered behaviours and expectations. For Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World instance, children tend to mimic (in attitudes Norms can change Values Survey. and actions) how their parents share paid and as economies unpaid work.29 develop, by changes new laws, policies or programmes, by social and Parenting experiences may, however, in- in communications political activism and by exposure to new ideas %uence and change adults’ social norms and and practices through formal and informal established gender roles. In the “mighty girl technology, by new channels (education, role models and media).23 e&ect,” fathers raise their awareness of gender laws, policies or Policymakers o'en focus on the tangible— disadvantages when they are rearing daugh- programmes, by social on laws, policies, spending commitments, ters.30 Parenting a school-age girl makes it easi- public statements and so forth. #is is driven er for men to put themselves in their daughter’s and political activism partly by the desire to measure impact (and shoes, empathize with girls facing traditional and by exposure thus prove effectiveness), by frustration gender norms and embrace nontraditional to new ideas and with the vagueness of “talking shops” ar- ones that would not place their daughters at a 31 practices through guing about rights and norms and by sheer disadvantage to men in the labour market. impatience with the slow pace of change. Adolescence is another key stage for gender formal and informal Yet neglecting the invisible power of norms socialization, particularly for boys.32 Young channels (education, FIGURE 4.8 role models and media) The share of both women and men worldwide with no gender social norms bias fell between 2005–2009 and 2010–2014

Percent of surveyed population responding 2005–2009 with biases towards gender equality and women’s empowerment 2010–2014

Indicated bias in one or Female 4 fewer questions from the Male World Values Survey

Indicated bias in two or Female more questions from the World Values Survey Male

Note: Balanced panel of 32 countries and territories with data from both wave 5 (2005–2009) and wave 6 (2010–2014) of the World Values Survey, accounting for 59 percent of the world population. Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey.

156 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.9

Progress in the share of men with no gender social norm bias from 2005–2009 to 2010–2014 was largest in Chile, Australia, the United States and the Netherlands, while most countries showed a backlash in the share of women with no gender social norms bias

Men Women

Chile Netherlands Australia Chile United States Australia Netherlands China Argentina Slovenia Poland Japan Thailand Ukraine Japan Trinidad and Tobago Trinidad and Tobago Thailand Spain United States Korea (Republic of) Jordan China Germany Georgia Russian Federation Romania Mexico Morocco Malaysia Jordan Cyprus Ghana Spain Cyprus Uruguay Malaysia Brazil Rwanda Ghana Russian Federation Rwanda Uruguay Argentina South Africa Poland Brazil Korea (Republic of) Ukraine Morocco Turkey Turkey Slovenia Georgia Mexico Romania India South Africa Germany India Sweden Sweden

–0.1 –0.05 0 0.05 0.1 0.15 –0.1 0 0.1 0.2 Mean change (value) Mean change (value)

FIGURE 4.10

Countries with higher social norms biases tend to have higher gender inequality

Gender Time spent on unpaid domestic chores Inequality Index (value) (ratio between women and men) 12 0.8

0.6

0.4 6

0.2 4

2 0 0 0.2 0.4 0.6 0 0.2 0.4 0.6 Social Norms Index (value) Social Norms Index (value)

Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey and Dashboard 2 in the statistical annex.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 157 FIGURE 4.11

Biases in social norms show a gradient

Biases against women’s capabilities (percent) Men make better political leaders than women do Men make better business executives 49.6 Men should have than women do Women should have more rights to a job the same political than women 42.1 rights as men 32.2 29.1

Basic Enhanced Basic Enhanced Politics The economy

Note: Balanced panel of 77 countries and territories with data from wave 6 (2011–2014) of the World Values Survey, accounting for 81 percent of the world population. Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey.

adolescents in di&erent cultural settings com- or think di&erently.36 Because of intertwined monly endorse norms that perpetuate gender social dynamics,37 challenging discriminatory inequalities, and parents and peers are central norms that impede gender equality and wom- in shaping such attitudes. Some of the endorsed en’s empowerment requires acting on more masculinity norms relate to physical toughness than one factor at a time. (showing higher tolerance for pain, engaging in !ghts, competing in sports), autonomy (being !nancially independent, protecting and pro- Restricted choices and power viding for families), emotional stoicism (not imbalances over the lifecycle “acting like girls” or showing vulnerabilities, Powerlessness dealing with problems on their own) and het- Gender inequality within households and manifests itself erosexual prowess (having sex with many girls, communities is characterized by inequality exercising control over girls in relationships) across multiple dimensions, with a vicious as an inability to (box 4.4).33 cycle of powerlessness, stigmatization, discrim- participate in or Social convention refers to how compliance ination, exclusion and material deprivation all inﬂuence decisions with gender social norms is internalized in reinforcing each other. Powerlessness manifests that profoundly affect individual values reinforced by rewards or itself in many ways, but at its core is an inabil- sanctions. Rewards use social or psychological ity to participate in or in%uence decisions that one’s own life, while approvals, while sanctions can range from ex- profoundly a&ect one’s own life, while more more powerful actors clusion from the community to violence or le- powerful actors make decisions despite neither make decisions despite gal action. Stigma can limit what is considered understanding the situation of the vulnerable normal or acceptable and be used to enforce nor having their interests at heart. Human neither understanding stereotypes and social norms about appropri- development is about expanding substantive the situation of the ate behaviours. A social norm will be stickiest freedoms and choices. This section presents vulnerable nor having when individuals have the most to gain from evidence of restricted or even tragic choices complying with it and the most to lose from women face.38 their interests at heart challenging it.34 Social norms have enough Examples of restricted choices can be iden- power to keep women from claiming their legal ti!ed in a lifecycle approach. Some represent rights due to pressure to conform to societal blatant limits to basic freedoms and human expectations.35 rights; others represent subtle manifestations of Social norms can also prevail when individ- gender biases. #e disparities of childhood and uals lack the information or knowledge to act adolescence are ampli!ed when women reach

158 | HUMAN DEVELOPMENT REPORT 2019 BOX 4.4

The man box

Engaging men and boys is a critical piece of advancing behaviours of the gender roles restrict men to act in a cer- the gender equality agenda. Gender equality implies tain way that preserves existing power structures. In 2019 changing and transforming the way individuals express Promundo along with Unilever estimated the economic and experience power in their lives, relationships and impacts of the man box in Mexico, the United Kingdom communities. Reaching equality, women and men will and the United States, considering bullying, violence, de- have the same agency to make choices and participate pression, suicide, binge drinking and trafﬁc accidents as in society. While women and girls bear the brunt of costs of restricting men to masculine behaviours.2 Two of gender inequalities, men and boys are also affected by the most damaging consequences for men are related to traditional conceptions of gender. their mental health: Men are less likely to seek mental Gender is a social construct of attributes or roles health services than women are, and men are more likely associated with being male or female. What it means to die by suicide than women are. Besides the ethical and to be a man or a woman is learned and internalized social gains of gender equality, men as individuals can based on experiences and messages over the course of beneﬁt from expressing freely, from having more options a lifetime, normalized through social structures, culture in their own experiences and behaviours and from having Challenging rigid and interactions. Though men usually have more agency better and healthier relationships with women and girls. gender norms and than the women in their lives, men’s decisions and be- So challenging rigid gender norms and power dy- haviours are also profoundly shaped by rigid social and namics in households and communities and involving power dynamics cultural expectations related to masculinity. men and boys in making these changes are important. in households and Masculinity is the pattern of social behaviours or Engaging men in preventing gender-based violence, communities and practices associated with ideals about how men should supporting women´s economic empowerment, pursuing behave.1 Some characteristics of masculinity relate to change for reproductive health and acting as fathers or involving men and dominance, toughness and risk-taking, recently referred caregivers are examples of how men can challenge their boys in making these to as toxic masculinity or the man box, in that traditional notions of masculinity and of their own selves. changes are important

Notes 1. Ricardo and MenEngage 2014. 2. Heilman and others 2019. adulthood, as exempli!ed in the di&erences in had an imbalanced sex ratio at birth, today 21 labour force participation and the representa- countries have a skewed ratio. #e preference tion of women in decisionmaking positions for a son can lead to sex-selective abortions in business and in politics (see figure 4.4). and to a large number of “missing” women, For unpaid care work, women bear a bigger particularly in some South Asian countries.40 burden, providing more than three times as Discrimination continues through how much as men.39 And older women’s challenges households share resources. Girls and women accumulate through the life course: #ey are sometimes eat last and least in the household.41 less likely than men to have access to pensions, #e gender politics of food—nurtured by as- even though they are expected to live three sumptions, norms and practices about women years longer. Along the way, social norms and needing fewer calories—can push women into path dependence—how outcomes today a&ect perpetual malnutrition and protein de!ciency. outcomes tomorrow—interact to form a highly Education opportunities, including access complex system of structural gender gaps. and quality, are a&ected by both household and community social norms. Gender di&er- Birth, early childhood and school age ences manifest !rst in girls’ families over education as a human right and later over respect In some cultures traditional social norms can for women’s agency to decide to study and to a&ect girls even before they are born, since choose her preferred !eld. Social norms can some countries deeply prefer bearing sons de!ne the level of education a girl can attain over daughters. While in the 1990s only some or her choice of study. #e restriction, control countries had the technology available to de- and monitoring of a girl’s or woman’s behav- termine a baby’s gender and only 6 countries iour and decisionmaking about her education

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 159 or job, or her access to financial resources presents disproportionate risks to women’s and or their distribution, constitute economic girls’ health, re%ecting both biological di&er- violence against her (see spotlight 4.1 at the ences and social norms (see box 4.3). And early end of the chapter). And even when girls are marriage limits girls’ choices. educated as well as boys, other e&ects of ine- The adolescent birth rate among women quality—driven especially by gendered social ages 15–19 is 104.7 per 1,000 in Sub-Saharan norms—reduce the likelihood that women Africa and 63.2 in Latin America and the will later attain positions of power and partici- Caribbean. When a teenage girl becomes preg- pate in decisionmaking. nant, her health is endangered, her education Worldwide, one in eight age-eligible girls and job prospects can abruptly end and her vul- does not attend primary or secondary school. nerability to poverty and exclusion multiplies.49 Only 62 of 145 countries have gender parity Adolescent pregnancy, o'en a result of a girl’s in primary and secondary education.42 Despite lack of opportunities and freedom, can re%ect the progress in enrolment ratios for some a failure among those around her to protect her countries, large di&erences persist in learning rights. outcomes and education quality. Contraception is important in maintain- Even among children attending school, ing good sexual and reproductive health.50 determinants of occupational choices appear Contraceptive use is higher among unmarried very early. Girls are less likely to study subjects and sexually active adolescents, but so is the such as science, technology, engineering and unmet need for family planning, especially in mathematics, while boys are a minority of those Asia and the Paci!c and Sub-Saharan Africa studying health and education.43 (!gure 4.12). #ere is still a stigma in many countries around unmarried women needing Adolescence and early adulthood family planning services. And in some countries regulations prevent access to these servic- Adolescence is when girls’ and boys’ futures es. Moreover, many women cannot a&ord to start to diverge; while boys’ worlds expand, pay for health care. 44 Social norms and girls’ worlds contract. Every year 12 mil- Social norms and traditional behaviour gen- lion girls are victims of forced marriage.45 Girls erally pose a threat to women’s reproductive traditional behaviour forced to get married as a child are victims of health. Women are more vulnerable to a loss generally pose a threat a human rights violation and are condemned of agency to have a satisfying and safe sex life, to women’s health to live a life with heavily restricted choices and the capability to reproduce and the freedom to low human development. decide if, when and how o'en to do so.51 When Child marriage not only alienates girls men use their power to decide on women’s from their families and social networks but behalf, that limits women’s access to resources also increases their risk of becoming victims 46 of domestic violence. It exacerbates overall FIGURE 4.12 gender inequality in education and employment by greatly reducing a girl’s chances of Contraceptive use is higher among unmarried completing formal schooling and developing and sexually active adolescent girls, but so is the unmet need for family planning, 2002–2014 skills for employment outside the home.47 It also leads to early and multiple pregnancies, Currently married/in a union increasing health risks for both the married Unmarried and sexually active girls and their children, since the risks of 51 newborn death and infant mortality and 41 morbidity are higher in children born to women under age 20.48 20 23 The health effects of early marriage are among the many health risks that are higher for women and girls than for men and boys. One Contraceptive prevalence, Unmet need for of the most globally widespread cross-cutting any method (percent) family planning (percent) forms of horizontal inequality, early marriage Source: UNFPA 2016.

160 | HUMAN DEVELOPMENT REPORT 2019 and dictates women’s behaviour. More broadly, income regions have a narrower gap in unpaid if women are seen as objects rather than agents care work. #e regions with the widest gaps in households and communities, this form of are the Arab States, South Asia, Sub-Saharan horizontal inequality can lead to violence and Africa and Latin America and the Caribbean— harassment (see spotlight 4.1 at the end of the the same regions that have the widest gaps chapter), a&ecting women’s mental health.52 for women´s labour force participation (!g- ure 4.13). #e struggle to reconcile care work Adulthood and older age responsibilities with paid work can lead women to occupational downgrading, where they Globally, women do more unpaid work than choose employment below their skill level and men do.53 However, the global gender income accept poorer working conditions.57 gap is 44 percent (see Statistical table 4). Some constraints faced by women are invis- Gender di&erences in paid and unpaid work ible when gaps are seen in isolation. Statistics and the gradients in empowerment combine typically record achievements (the func- Gender differences in multiple elements that restrict women’s choices. tionings) but not the full set of choices (the paid and unpaid work #e gaps illustrate the multidimensional e&ects capabilities). #is partial view tends to hide and the gradients of gender inequality on occupation choices, the multidimensional biases in choices wom- in empowerment income and women’s !nancial independence en face. Take, for instance, a quali!ed woman and resilience to external shocks. who has children and must decide between combine multiple A key constraint on women’s decisionmaking taking a job and staying home. Workplace elements that restrict 58 is their disadvantages in the amount of unpaid inequalities (including pay gaps and the risk women’s choices work they do, bearing disproportionate re- of harassment), social norms (pressure to ful!l sponsibility for housework, caring for family the role of mother) and imbalances at home (a members and performing voluntary communi- greater load of domestic unpaid work), among ty work.54 On average, women spend about 2.5 other factors, may deter her from participating times as much time on unpaid care and domes- in paid work. #e woman’s choice may bring tic work as men do.55 #is a&ects women’s la- feelings of guilt or regret. A large proportion bour force participation, lowers economywide of female homemakers feel that by staying productivity and limits their opportunities to home they are giving up a career or economic spend time in other ways.56 #is sort of gender independence. A large proportion of mothers inequality is linked to levels of income: Higher employed in paid occupations face the stress of

FIGURE 4.13

The gap in unpaid care work persists in developing economies

Proportion of time spent on unpaid domestic chores and care work (percent) Female Male

21.6 21.0 19.4 19.2 16.8 15.2

5.8 5.5 5.9 5.2 5.6 3.0

Arab States East Asia and the Europe and Latin America South Asia Sub-Saharan Paciﬁc Central Asia and the Africa Caribbean

Note: Aggregation rule has been relaxed; estimates not published in dashboard. Source: Human Development Report Ofﬁce.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 161 feeling that their choice implies su&ering for stability61 or to legal discrimination and gender their children (!gure 4.14). norms.62 Women face restricted resources in Moreover, home-based inequalities exacer- areas besides !nance, with climate change, in bate market-based gender inequality through particular, exacerbating existing inequalities in the motherhood pay gap—a term that can refer women’s livelihoods and reducing their resil- to the di&erence in pay between mothers and ience (box 4.5). childless women, or to that between mothers As noted, girls and women of reproductive and fathers, rather than between all working age (15–49 years old) are more likely than men and women. #e motherhood pay gap is boys and men of the same age to live in poor usually bigger in developing countries, and in households (!gure 4.16). #is challenges the all countries it increases with the number of “headship de!nitions” approach to household children a woman has. The combination of composition for examining poverty pro!les, in low earnings and dependants makes women which households with a male earner, a non- over-represented among poor people during income earner spouse and children are more their reproductive age: Women are 22 percent likely to have poor women. Children and other more likely than men to live in a poor house- dependants can be an important vulnerability hold between the ages of 25 and 34.59 factor for women in their reproductive health. Women’s ﬁnancial According to the World Bank’s 2017 Global For both genders, pooling resources and having independence can Findex, of the 1.7 billion unbanked adults more adults working for pay in a household can in the world, 56 percent are women, while in protect them from falling into poverty, as can be dependent on developing countries women are 9 percentage education, especially for women.63 socioeconomic factors points more likely to be unbanked than men.60 For most people lifetime working conditions such as profession, #e Arab States and Sub-Saharan Africa have have a great impact on economic conditions the lowest percentage of women with an ac- and autonomy in older age. For women— earnings and income count at a !nancial institution or with a mobile over-represented among older people—earlier stability or to legal money-service provider, but the percentage gender gaps in health, wages, productivity, discrimination and is below 80 percent in all developing country labour participation, formal versus informal regions (!gure 4.15). Women’s !nancial inde- work, remunerated versus nonremunerated gender norms pendence can be dependent on socioeconomic work, continuity in the labour market and the factors such as profession, earnings and income ability to own property and save are likely to

FIGURE 4.14

A large proportion of employed women believe that choosing work implies suffering for their children, while a large proportion of female homemakers feel that by staying home they are giving up a career or economic independence, 2010–2014

Employed women who agree that Homemakers who agree that a job is children suffer if women have a job the best way for a woman to be independent

Arab States Arab States

East Asia and the Paciﬁc East Asia and the Paciﬁc

Europe and Central Asia Europe and Central Asia

Latin America and the Latin America and the Caribbean Caribbean

South Asia South Asia

Sub-Saharan Africa Sub-Saharan Africa

0 20 40 60 80 0 20 40 60 80

Source: Human Development Report Ofﬁce calculations based on data from wave 6 (2010–2014) of the World Values Survey.

162 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.15

The percentage of women with an account at a ﬁnancial institution or with a mobile money-service provider is below 80 percent in all developing country regions in 2018

Women with access to an account with ﬁnancial institution or a money service provider, ages 15 and older, simple average (percent) 100

0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Arab States East Asia and Europe and Latin America and South Asia Sub-Saharan Africa Developed the Paciﬁc Central Asia the Caribbean Source: Human Development Report Ofﬁce based on data from the Global Findex database.

BOX 4.5

Climate change and gender inequality

Women tend to be responsible for procuring and pro- reduction and the mitigation of climate change effects viding food in households and are the primary workers and environmental degradation. engaged in subsistence agriculture. They make up an Greater female participation in natural resource average of 43 percent of the agricultural workforce in management, productive agricultural activities and nat- Greater female 1 developing countries. ural disaster responses can enhance the effectiveness participation in natural Even so, they experience inequitable access to and sustainability of policies and projects. Closing the land and agricultural inputs,2 which can affect their pro- gender gap in agricultural productivity would increase resource management, ductivity in the sector, generating a gap in comparison crop production by 7–19 percent in Ethiopia, Malawi, productive agricultural with men’s productivity. In Ethiopia, Malawi, Rwanda, Rwanda, Tanzania and Uganda.4 Tanzania and Uganda the gender gap in agricultural Climate change can affect women’s income, edu- activities and natural productivity ranges from 11 percent to 28 percent.3 The cation, access to resources, access to technologies and disaster responses difference is due to access to credit, ownership of land, access to information.5 It is entangled with economic use of fertilizers and seeds, and availability of labour. and social consequences for women. Women in devel- can enhance the As in many other dimensions, gendered norms and oping countries are highly vulnerable when they depend effectiveness and traditions at the household level are behind the ineq- heavily on local natural resources for their livelihood. sustainability of uitable allocations of production factors, thus limiting Yet women are powerful agents of change. As key play- women’s agency, decisionmaking power and partici- ers in core productive sectors, they are well placed to policies and projects pation in the labour market. Furthermore, the gender identify and adopt appropriate strategies to address agricultural gap hinders poverty reduction, inequality climate change at the household and community levels.

Notes 1. FAO 2011. 2. UN Women, UNDP and UNEP 2018. 3. UN Women, UNDP and UNEP 2018. 4. UN Women, UNDP and UNEP 2018. 5. Brody, Demetriades and Esplen 2008.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 163 FIGURE 4.16 expectations may lead to a perceived clash—a con%ict, for example, of women’s rights with Girls and women of reproductive age are more traditional values—or reveal subconscious likely to live in poor households than boys and men biases. Still, even norms can be shi'ed towards Poverty rate (percent) gender equality. 25 #e shi' can be supported with a proactive stance, generating new regulations and policy 20 interventions that mainstream gender equality 15 and women’s empowerment. This has been Female happening but has not been enough to create 10 long-term changes in stereotypes and tradition- Male al gender roles. Entrenched inequalities persist 5 due to discriminatory social norms and harmful 0 behaviours and practices that undermine im- 0–14 15–24 25–34 35–39 40–49 50–54 55–59 60+ plementation. Well intentioned interventions Age group (years) might fail or might have unintended conse-

Source: Munoz Boudet and others 2018. quences if policymakers do not consider deeply rooted norms and practices. For instance, a$rmative action or positive discrimination become later gender gaps in well-being.64 #e has sometimes overlooked or underplayed the gap widens when pension systems are based e&ects of social norms on overall outcomes.67 on contributory schemes, and even more when E&orts to promote women’s representation they take the form of individual accounts.65 In in positions of leadership have yet to succeed, most developed countries women have equal and major prejudices persist about women’s access to pensions. But in most developing ability to participate politically and function in countries with data, there is a women’s pension high o$ce. Representation quotas for women gap (see Dashboard 2 in the statistical annex). sometimes do not deliver the envisaged trans- formation and risk promoting tokenism by introducing women’s presence while power Empowering girls and remains entrenched in traditional hierarchies women towards gender and privileges based on other identities such as equality: A template to reduce class, race and ethnicity. horizontal inequalities Varied alternatives should be priorities in light of multiple and complementary identities Expanding opportunities for women and girls; rather than competing, con%icting ones—the promoting their economic, social and political multiple identities of an individual as a wom- participation; and improving their access to an, a mother, a worker and a citizen should be social protection, employment and natural re- mutually supportive, not counterposed. So, sources make economies more productive. Such choices that enhance multiple freedoms are to investments reduce poverty and inequality and be prioritized over choices based on a singular make societies more peaceful and resilient.66 All identity that diminish other freedoms. Any that is well known. Social norms are shi'ing approach addressing gender inequality should The backlash against towards changed gendered roles in society. But consider the multidimensional character of changing gender while some conventional gender norms evolve gender and be sensitive to local social norms. in the private and public domains, their e&ects Norm-aware interventions for women focus on roles in households, are also facing a backlash from the conventional supporting them by providing solutions that workplaces and power relations between men and women in work around existing social norm constraints. politics affects entire today’s social hierarchies. Options to reduce gender inequalities—and societies inﬂuenced by #e backlash against changing gender roles many other horizontal ones—need to consider in households, workplaces and politics a&ects how to directly target changes in unequal pow- shifting power relations entire societies in%uenced by shi'ing power er relationships among individuals within a relations. #e resistance to changes in gender community or to challenge deeply rooted roles.

164 | HUMAN DEVELOPMENT REPORT 2019 #is may include a combination of e&orts in change—targeting both women and men is Analysis that goes education, raising awareness by providing new crucial. #e importance of adequately engaging beyond averages information and changing incentives. men and boys in overcoming gender inequality An additional and important considera- or addressing their own gender-related vulnera- requires more and tion to in%uence change in social norms and bilities is acknowledged, but actions have a long better data to keep traditional gender roles is for options to be way to go. pushing for gender inclusive of both women and men, which may Finally, analysis that goes beyond averages re- equality and to make hold also for other horizontal inequalities. quires more and better data to keep pushing for When choosing among alternatives—whether gender equality and to make other horizontal other horizontal norm-aware or those pursuing social norm inequalities visible (box 4.6). inequalities visible

BOX 4.6

Better data are needed on gender inequalities

Gender data face challenges of quantity and quality. The disaggregating by sex and age, using outdated measure- first refers to not having enough data to depict women’s ments of time use and collecting data only on households current situation. For instance, among the Sustainable instead of individuals. Changes in these measurements Development Goals over 70 percent of data for 58 indi- can affect indicators such as the Multidimensional cators linked to gender equality and women’s empow- Poverty Index, calculated for households rather than indi- erment is missing.1 The second refers to current data viduals, so that complementary research may be needed that might not accurately reflect reality and that might to clarify the relationship between gender and poverty.2 underestimate women’s roles and contributions. More information is needed to get a better picture Some organizations perceive collecting and produc- of gender biases specific to a region, country or coming gender data as expensive in time and cost. Some munity, as with information on the impact of media and data collection methods are outdated and biased against social networks in reinforcing traditional norms and women because they follow gender social norms, such stereotypes.3 as interviewing only the male head of household, not

Notes 1. Human Development Report Ofﬁce calculations based on data from UN Women (2017). 2. UNDP 2016. 3. Broockman and Kalla 2016; Paluck and others 2010.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 165 SPOTLIGHT 4.1 Women’s unequal access to physical security—and thus to social and political empowerment

Violence against women is one of the cru- Decomposing these factors reveals inequality in Violence against ellest forms of women’s disempowerment. the experience of violence, an insight that can women is one Magnifying inequality, it happens throughout help in designing more focused interventions. the lifecycle, in di&erent spaces—households, For instance, although violence can occur at all of the cruellest institutions, public spaces, politics and on- education levels, greater education attainment forms of women’s line—in all societies, among all socioeconomic can protect women from partner violence. disempowerment groups and at all levels of education. And it Educated women have better access to informa- can be perpetuated re%ects the same social norms that legitimize tion and resources that help them identify an harassment and discrimination. abusive relationship and end it.3 Women’s eco- through social norms More than a third of women—and more nomic empowerment through participation in than two-thirds in some countries—have ex- the workforce had mixed associations with the perienced physical or sexual violence in%icted risk of intimate partner violence,4 challenging the by an intimate partner or sexual violence in- notion that economic empowerment protects %icted by a nonpartner (!gure S4.1.1).1 Some women from gender-based violence. #is !nding 20 percent of women have experienced sexual highlights the heavy in%uence of social norms in violence as children. Nearly a quarter of girls women’s perceptions of their status in society in ages 15–19 worldwide report having been vic- some cultures. In developing countries women tims of violence a'er turning 15.2 And violence make up a large proportion of the informal sector is typically underestimated because of stigma, workforce with low-paying jobs, a structure that denial, mistrust of authority and other barriers might perpetuate the myth of male superiority.5 to women reporting an incident. Violence against women can be perpetuated Intimate partner violence has been recurrent- through social norms. For example, female ly associated with such factors as age, wealth, genital mutilation and cutting remain wide- marital status, number of children, education spread. An estimated 200 million women and attainment and economic empowerment. girls living today have undergone female genital mutilation, even though most men and women oppose the practice in many countries where FIGURE S4.1.1 it is performed.6 Violence against women and About a third of women ages 15 and older have girls is sustained by individual behaviours and experienced physical or sexual violence inﬂicted beliefs as well as by social norms from the com- by an intimate partner, 2010 munities and networks that can slow change.

Percent Violent actions, attitudes and behaviours are triggered by unequal power relations dictat- Arab States ing gender roles at the household level. Some 35 examples are beliefs that a man has a right to 25 physically discipline a woman for an incorrect Sub- East Asia behaviour, divorce is shameful or sex is a man’s Saharan 15 and the Africa Paciﬁc right in marriage. 5 When women assert autonomy or aspire to exert power at any level—from the household to the national government—they o'en face Europe and South Asia Central Asia a backlash that can include violence (psychological, emotional, physical, sexual or economic), whether as discrimination, harassment, Latin America and the Caribbean assault or femicide. More than 85 percent Source: WHO 2013. of female members of European parliaments

166 | HUMAN DEVELOPMENT REPORT 2019 have experienced psychological violence, and or unwanted comments or purposely touching 47 percent have received threats of death, or brushing up against someone. Women are Women are rape, beating or kidnapping (!gure S4.1.2).7 harassed mostly in public spaces, their workplac- Moreover, the only country in the world that es, their residences or their schools.10 harassed mostly has legally made political violence a separately #rough social media and other online plat- in public spaces, 8 de!ned crime is Bolivia. Elsewhere, lacking forms and applications, women are vulnerable their workplaces, laws, regulations and sanctions, women are le' to harassment and bullying in a new space—the unprotected from this type of violence. In 2016 digital public space. Ensuring that this space their residences the #Not#eCost campaign was launched to is safe and empowering for women and girls is or their schools raise awareness and stop violence against wom- a new challenge. Some 73 percent of women en in politics. #e name alludes to how women online have been exposed to some type of cyber- are told that harassment, threats, psychological violence, and women are 27 times more likely abuse and other forms of violence are “the cost” than men to be the victims of cyberviolence.11 of participating in politics.9 Traditional gender Besides the impact of violence against women norms play a role in such political violence. and girls in other spaces, cyberviolence impedes Globally, there are some e&orts to !ght the their digital inclusion and keeps them from backlash. Political violence and sexual harass- enjoying digital dividends. Even though tech- ment and assault received attention in 2017 nology can connect and empower, it can also when American actress Alyssa Milano called for reinforce traditional gender roles and normalize women to come forward with their experienc- stereotypes re%ecting a culture of misogyny and es. Some 1.7 million tweets using the hashtag marginalization. Security and harassment are #MeToo responded, and 85 countries had at among the top !ve barriers to women’s mobile least 1,000 #MeToo tweets. #e movement gave phone ownership and use.12 Online harassment, visibility to the issue and propelled initiatives sexist attitudes and misogynistic remarks can to conduct more research on sexual harassment undermine women’s sense of legitimacy, compe- and assault, especially in the United States. Some tence and safety, making them mistrust technol- 81 percent of women and 43 percent of men in ogy and even opt out of its use. Besides hindering the United States reported experiencing some technological inclusion, violence against women form of sexual harassment or assault in their life- and girls in this space has a cumulative emotional time. #e most common forms of sexual harass- and physical cost on them. ment are whistling, honking, saying disrespectful For each demographically “missing” woman, many more fail to get an education, job or political responsibility they would have obtained FIGURE S4.1.2 if they were men.13 Gender is a global factor in Female members of European parliaments unequal human autonomy, physical security experience high rates of acts of political violence and social, economic and political empower- against women, 2018 ment. Women’s human development depends on socioeconomic enabling factors such as Reported incident Experienced this form of violence the ability to pursue a profession, to attain Percent income stability and to achieve earnings comparable to men’s. Women’s empowerment in health, education, earning opportunities, and political rights and participation can change 39 23 social decisionmaking and development (!g- ure S4.1.3). Women’s human development also 21 requires positive gender norms and an absence 25 22 20 8 of gender discrimination, with laws preventing 8 5 7 Online Threats Psychological Sexual Physical unequal treatment, harassment and violence violence of physical harassment harassment violence against women. Education, reproductive rights violence and political participation are key assets in all these areas, while the right to human security is Source: IPU 2019. fundamental.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 167 FIGURE S4.1.3

Traditional social norms encourage different forms of violence against women

Mental health

H e a Women’s l th empowerment in Psychological/ Education health, education, emotional earning opportunities, and political rights and participation Forms of Physical can change social Labour Economic violence against Sexual integrity decisionmaking women and development

Financial Physical inclusion

Security

Source: Human Development Report Ofﬁce based on UN General Assembly (2006).

8 Government of Bolivia 2012. 9 NDI 2019. Notes 10 Kearl 2018. 11 Broadband Commission for Digital Development Working 1 WHO 2013. Group on Broadband and Gender 2015; Messenger 2017. 2 UNICEF 2014a. 12 GSMA Connected Women 2015. 3 Flake 2005; Waites 1993. 13 Duflo 2012. 4 Sardinha and Catalán 2018. 14 Caprioli 2005. 5 Uthman, Lawoko and Moradi 2011. 15 Ouedraogo and Ouedraogo 2019. 6 UNICEF 2018a. 16 Stone (2015) as cited in O’Reilly, Ó Súilleabháin and Paffenholz 7 IPU 2019. (2015).

168 | HUMAN DEVELOPMENT REPORT 2019