Women Voters In Indian Democracy

7m ago
359.24 KB
5 Pages
Last View : 1m ago
Last Download : n/a
Upload by : Brenna Zink

NOTESWomen Voters inIndian DemocracyA Silent RevolutionMudit Kapoor, Shamika RaviAn analysis of the politicalparticipation of women bycomparing the turnout of womenvoters to men in all the stateelections from 1962 till 2012reveals a steady and sharp declinein the gender bias in voting overtime. This phenomenon is seenacross all the states, includingthe traditionally “backward”states of north India.Mudit Kapoor (Mudit [email protected]) andShamika Ravi (Shamika [email protected]) are atthe Indian School of Business, Hyderabad.Economic & Political WeeklyEPWmarch 22, 2014The persistence of gender inequality which is embodied in “missingwomen”, a concept developed bySen (1990, 1992), is a common phenomenon in low-income countries and issuspiciously high in countries such asIndia and China. Through a political context, our study aims to contribute to thebroader literature and understanding ofgender inequality.More specifically, in this article, westudy the gender bias in political participation by analysing women voter turnout in Indian democracy from 1962 till2012 and discover a silent revolution.The main finding of our analysis is asteady and a sharp decline in the genderbias in voting over time. In particular,we find that the sex ratio of voters whichis defined as the number of women votersto every 1,000 men voters, increased veryimpressively from 715 in the 1960s to883 in the 2000s. We also find that thisphenomenon of declining gender bias invoting is across all the states, includingthe traditionally backward “BIMARU”states of Bihar, Madhya Pradesh (MP),Rajasthan and Uttar Pradesh (UP). Ondecomposing the data further, we discover that this decline in gender bias ofvoters is solely driven by the dramaticincrease in women participation in theelections since the 1990s, while menparticipation has remained unchanged.The right to equality in voting is abasic human right in liberal democracy.Women enjoy this right to equality invoting, and by casting a vote they makea formal expression of their individualchoice of political parties, representativesor of broad policies. The fact that morewomen are voluntarily exercising theirconstitutional right of adult suffrageacross all states in India is testimony tovol xlix no 12the rise of self-empowerment of womento secure their fundamental right tofreedom of expression. This is an extraordinary achievement in the world’slargest democracy with 717 million votersof which 342 million voters are women.Existing literature has documented asignificant gender gap in various sectorsincluding health, labour market opportunities, education and political representation. Anderson and Ray (2010) providea decomposition of missing women by ageand cause of death, and their findingssuggest that excess female mortality isa universal phenomenon and shouldnot be attributed merely to parentalpreferences. Duflo (2012) provides anexcellent overview of the literatureby reviewing the relationship betweengender inequality and economic development. In contrast to existing literaturewhich finds stark persistence in genderinequality over time in various outcomes, we report a more positive phenomenon of sharply declining gender biasin political participation measured byvoter turnout. We document the growingpolitical empowerment of women votersin India. The reason why this is animportant development is because it isnot an outcome of any specific top-downpolicy intervention to raise voter turnout of women, but is driven over time bythe voluntary act of self-empowermentof women.This article also differs from the existing research on women and politics in afundamental way. While previous workshave focused on women as policymakers, our emphasis is on women as votersin a political system. This is a criticalarea of research as women voters comprise a significant share in any electionwithin a representative democracy.For our analysis, we use constituencylevel data from the Election Commission of India (ECI). We analyse all stateassembly elections held from 1962till 2012, for 16 large states. These 16states together represent more than93% of the total electors in India. This isa rich data set which has been usedby earlier papers that have exploredwomen’s political participation and crime63

NOTES(Iyer et al 2012), how voters respond tocriminal charges of candidates (Duttaand Gupta 2012) and relationship ofgrowth to election outcomes in India(Gupta and Panagariya 2012).The rest of the article is organised inthe following manner: we briefly describeECI data in Section 1. Within the datasection, we also describe the constructionof variables of interest that we analysefrom the data. The results are analysedin Section 2 where we highlight the consistent trends of reducing gender gap invoter turnout in all states of India overthe last five decades. We discuss andconclude our findings in Section 3.female poll percentagest (( Nsi 1 female votersitNs i 1 female electorsitmale poll percentagest Nsi 1 male votersit)* 100 .(3))* 100.(4) Nsi 1 male electorsitwhere s is the state, t is the year in whichthe election is held for the state assembly,i is the assembly constituency in states, and Ns is the total number of assemblyconstituencies in state s. In Table 1 wesummarise the years in which the assembly elections were held in each of thebig 16 states in India.Table 1: States and the Election YearsStates1 DataThe first general election in India washeld in 1951-52 and so far 17 elections tothe House of the People have been held.At present there are 543 members in theLok Sabha. The Election Commission isa permanent constitutional body thatwas established in 1950.The ECI collects and documents election data for each and every parliamentary and the state assembly constituency.For each constituency it reports data onthe total number of electors and voterswhich are segregated by gender, the nameand gender of each candidate contestingthe election, party affiliation of eachcontestant and if the candidate is notaffiliated to any party then the candidateis categorised as an independent, and thetotal number of votes secured by eachcandidate in the election. This data isavailable for every general election heldin the parliamentary and the state assembly constituency from 1951 till 2012.For our analysis we use data at theconstituency level for the state assemblyelections held for 16 large states from1962 till 2012. These 16 large states represent more than 93% of the total electorsin India. Next we describe the construction of the variables of interest using thedata at the constituency level.sex ratio of votersst (( Nsi 1 female votersit Nsi 1 male votersitsex ratio of electorsst 64)* 1000) Nsi 1 female electorsit Nsi 1 male electorsit.(1)* 1000 .(2)held in each of the decades from the1960s till the 2000s. We illustrate thiswith the help of an example. In UP in thedecade of 1960s, three state elections wereheld in 1962, 1967 and 1969 and in eachof these elections the sex ratio of the voters was 538, 675, and 629 respectively.So we compute the simple mean of thesethree sex ratios of the voters which is614 and define that as the average sexratio of the voters in the 1960s in UP. Foreach state we do a similar computationfor all the decades. Our results reveal avery interesting pattern – for every statefrom the 1960s to the 2000s there hasElection YearsAndhra PradeshAssamBihar2009 2004 1999 19942011 2006 2001 19962010 Oct Feb 20002005 2005Gujarat2007 2002 1998 1995Haryana2009 2005 2000 1996Himachal Pradesh 2007 2003 1998 1993Karnataka2008 2004 1999 1994Kerala2011 2006 2001 1996Madhya Pradesh 2008 2003 1998 1993Maharashtra2009 2004 1999 1995Odisha2009 2004 2000 1995Punjab2012 2007 2002 1997Rajasthan2008 2003 1998 1993Tamil Nadu2011 2006 2001 1996Uttar Pradesh2012 2007 2002 1996West Bengal2011 2006 2001 19961989 1985 1983 1978 1972 1967 19621991 1985 1983 1978 1972 1967 19621995 1990 1985 1980 1977 1972 1969 1967 19671971197719711962196719621967 19651962196219671967 1962196219671974 1969 1967 19621969 1967 1962Data Source: Election Commission of India.2 ResultsThe first set of results are summarised inTables 2a and 2b. Before we explain theresults it is important to describe howwe construct the average sex ratio ofvoters for every state for the electionsbeen a significant improvement in the sexratio of the voters; however, in Haryanaand Tamil Nadu (TN) the improvementis marginal compared to other states.Some of the largest gains happened inthe poorest states of India, the so-calledTable 2a: Average Sex Ratio of the VotersTable 2b: Average Sex Ratio of the ElectorsStates1960s 1970sStatesAndhra PradeshAssamBiharGujaratHaryanaHimachal PradeshKarnatakaKeralaMadhya PradeshMaharashtraOdishaPunjabRajasthanTamil NaduUttar PradeshWest Bengal886 910 917 924 971620 716 780 868 896567 573 625 701 783786 824 799 824 853807 806 806 815 810675 802 957 970 1,029797 843 856 886 913981 1,034 1,017 1,027 1,046558 677 677 730 804794 853 823 859 855572 639 679 802 866777 810 810 793 904603 745 735 760 861949 917 926 923 961614 667 660 671 761670 707 831 880 8861980s 1990s 2000s1960s1970s 1980s1990s 2000sAndhra Pradesh996 1,011 997 1,000 1,018Assam831 855 869897 932Bihar884 897 894878 867Gujarat941 970 973946 943Haryana881 888 875857 838Himachal Pradesh 947 943 1008983 971Karnataka954 957 970964 966Kerala1,025 1,020 1,025 1,041 1,075Madhya Pradesh 988 990 977937 906Maharashtra934 955 948923 913Odisha923 925 912886 937Punjab846 854 837872 906Rajasthan922 938 908894 910Tamil Nadu1016 987 976984 998Uttar Pradesh846 850 830821 827West Bengal796 794 875901 912Source: Authors’ calculations from ECI data.Source: Authors’ calculations from ECI data.march 22, 2014vol xlix no 12EPWEconomic & Political Weekly

NOTES“BIMARU” states. However, we also observethat there are significant variations in thesex ratio of the voters across the states.In sharp contrast, in Table 2b, whenwe look at the average sex ratio of theelectors we do not observe a similarpattern. As a matter of fact, there aresome states for which the sex ratio ofelectors has marginally worsened fromthe 1960s to the 2000s. This suggests thatthe increase in the sex ratio of the votersis not driven by more women registeringto vote relative to men, but is, instead,driven by the fact that more women areactually casting their vote in the elections, relative to men. This is an important observation in the data.We also plot the sex ratio of voters andelectors for every state over the electionyears. We see that over the 50 years, thetrend in the sex ratio of voters is positivefor all the states, except for Haryana andTN where the voter sex ratios are nearlyflat. This means that in almost all thestates of India, more and more womenare casting their votes in comparison tomen. In sharp contrast, the trend in theelectors’ sex ratio for most of the statesremains more or less flat with the exception of Haryana where it has declined.Given that the electors’ sex ratio is aclose approximation of the gender ratio inadult population of the state, the trendin electors’ sex ratio reflects the trend inthe adult population sex ratio of thestate. Figure 1a shows the voters’ sex ratioand electors’ sex ratio for the traditionally backward states in India. Figure 1bshows the voters’ sex ratio and electors’sex ratio for the southern states in Indiaand finally Figure 1c (p 66) shows thevoters’ sex ratio and electors’ sex ratiofor the remaining seven large states.Next we do a fixed effects regressionanalysis to study the trends in the sexratio of voters and electors. We also do asimilar fixed effects regression analysisfor the female and the male poll percentage. In particular we run the followingregression:Yst state dummies β1post70 β2post80 β3post90 β4post00 stwhere yst is the outcome variable ofinterest. The different outcomes that wewill study are the following: (1) sex ratio ofvoters, (2) sex ratio of electors, (3) femaleEconomic & Political WeeklyEPWmarch 22, 2014Figure 1a: Bihar, Madhya Pradesh, Rajasthan, Uttar PradeshBiharMadhya Pradesh 900900 800 700 600500 1960 1000800 700 50019701980 Sex ratio of votersSex ratio of electors19902000Fitted valuesFitted values2010 600 19601970198019902000Fitted valuesFitted values Sex ratio of votersSex ratio of electorsRajasthan2010Uttar Pradesh1000900900 800 700600 1960 800 700 19902000Fitted valuesFitted values2010 600 19902000Fitted valuesFitted values 50019701980 Sex ratio of votersSex ratio of electors 196019701980 Sex ratio of votersSex ratio of electors2010Source: Authors’ calculations from ECI data.Figure 1b: Southern StatesAndhra PradeshKarnataka105010001000950 950 900850 1960 900 850 19701980 Sex ratio of votersSex ratio of electors19902000Fitted valuesFitted values20108001960197019801990 Sex ratio of votersSex ratio of electorsKerala2000Fitted valuesFitted values1060105010001040 1000 1020950 Sex ratio of electors19902000Fitted valuesFitted values2010 8501980 Sex ratio of voters 900 19702010Tamil Nadu 1080980 19701980 Sex ratio of voters1990Sex ratio of electors2000Fitted valuesFitted values2010Source: Authors’ calculations from ECI data.poll percentage, and (4) male poll percentage. The subscript s is the state, t isthe year in which the election is held forthe state assembly, post70 is a dummyvariable which equals 1 if the state elections are held on or after the year 1970 andbefore the year 1980. Similarly, post80 isa dummy variable which equals 1 if thestate elections are held on or after the year1980 and before the year 1990, post90 isa dummy variable which equals 1 if thestate elections are held on or after the year1990 and before the year 2000, post00 isa dummy variable which equals 1 if thestate elections are held on or after theyear 2000. st is the error term.vol xlix no 12The results of our regression analysisare in Table 3 (p 66) from column (1) tocolumn (8). In column (1) and column(2) of the table we report the results forsex ratio of the voters with and withoutthe state fixed effects respectively. Fromcolumn (1) we note that in the 1960s theaverage sex ratio of the voters was 715and it increased significantly by 168points to 883 in the 2000s. This is thekey finding of this article.The results are very similar whenwe control for the state fixed effects. Inparticular, from column (2) we see thatthe sex ratio of the voters has increasedvery significantly over time and has65

NOTESFigure 1c: Remaining Large StatesAssam1000 900 800 700 1970950880900860 198019902000Fitted valuesFitted values Sex ratio of votersSex ratio of electors2010 800 1960900850 600750 1960 1970 19801990Sex ratio of electors8202000Fitted valuesFitted values 800 80060019902000Fitted valuesFitted values2010West Bengal1000900 800 700 196019701980 Sex ratio of votersSex ratio of electors19902000Fitted valuesFitted values2010improved for every successive decade.When we compare the 1960s to the 1970swe note that the sex ratio of the voterswent up by 45, from the 1970s to the1980s it improved by 28, it further wentup by 30 from the 1980s to the 1990s andit improved dramatically by 52 when wecompare the 1990s to the 2000s. Theseresults reflect the main point of our articlethat gender bias in voting has declinedsteadily and very significantly over time.In columns (3) and (4) we report resultsfor the sex ratio of electors with andwithout the state fixed effects, respectively. We find that the sex ratio of theelectors has remained more or less thesame throughout the decades. This resultalso directly implies that an improvement in the sex ratio of voters is drivenby more women actually casting theirvotes in elections, relative to men. Andthat this result is not driven by morewomen getting registered to vote whichis reflected in the unchanging sex ratioof electors in the states over time.We directly test this by looking at thevoter turnout of women and men over50 years. In particular, we study the 198019902000Fitted valuesFitted values Sex ratio of votersSex ratio of electors 2010 1970 900 850 19701980 Sex ratio of votersSex ratio of electors19702010 Punjab 700 1960 950900 Odisha 850840800100090066 Sex ratio of voters950600 Maharashtra750HaryanaGujarat1000 800 750 70019801990 Sex ratio of votersSex ratio of electors2000Fitted valuesFitted values2010female and male poll percentage overtime. Female poll percentage is definedas number of female voters who actuallycast their vote to every 100 registeredfemale electors. The results for womenvoter turnout or female poll percentage arereported in columns (5) and (6) with andwithout the state fixed effects, respectively. Our results show that over time thefemale poll percentage has increased verysignificantly and it has been increasingdecade after decade steadily. The risingtrend started in the 1980s and picked upmomentum in the 1990s and the 2000s.In sharp contrast to the rising femalepoll percentage, when we look at themale poll percentage or male voterturnout (results in columns (7) and (8)with and without the state fixed effects, 19601970198019902000Fitted values Sex ratio of votersSex ratio of electorsFitted values2010respectively) we do not observe a similar pattern. In fact the male poll percentage has remained stable through thedecades while increasing only marginally in the 2000s, however, and it is notsignificant at the conventional 5% levelof statistical significance. This resultconfirms that more women relative to menare turning out to vote in the elections.This voluntary act of self-empowermentis the key driving force in reducing thegender bias in voting amongst Indiansover the last five decades.Next, we are curious to see whetherthis positive national phenomenon ofrising female voter participation alsoholds in the states which have been traditionally considered backward in India.For this, we repeat the same regressionTable 3: Regression AnalysisSex Ratioof Voters(1)post70post80post90post00ConstantState fixed effectsAdj-R2ObservationsSex Ratioof Voters(2)Sex Ratioof Electors(3)53.6*45.1*** 17.1(30.1)(15.7)(16.9)100.3***73.1*** 30.5*(30.1)(14.7)(16.2)119.0*** 104.0*** 24.4(29.5)(16.5)(15.9)168.0*** 156.4*** 25.2(27.2)(16.4)(16.1)715.1*** 846.2*** 176Sex Ratioof Electors(4)Female 0**(7.43)996.4***(5.65)Y0.850176Female Poll Male PollMale PollPercentage Percentage 615176-1.11-1.27(1.66) (1.46)-0.18-0.89(2.01) (1.67)1.501.97(1.99) (2.05)2.812.45*(1.77)(1.45)65.8*** 71.1***(1.15)(1.44)NY0.0280.390176176Standard errors in parentheses, * p 0.10 ** p 0.05 *** p 0.01.post70 is a dummy variable which equals 1 if the state elections are held on or after the year 1970 and before the year 1980and 0 otherwise, post80 is a dummy variable which equals 1 if the state elections are held on or after the year 1980 andbefore the year 1990 and 0 otherwise, post90 is a dummy variable which equals 1 if the state elections are held on or afterthe year 1990 and before the year 2000 and 0 otherwise, post00 is a dummy variable which equals 1 if the state electionsare held on or after the year 2000 and 0 otherwise.march 22, 2014vol xlix no 12EPWEconomic & Political Weekly

NOTESTable 4: Regression Analysis for the “BIMARU” StatesSex Ratioof Voters(1)post70post80post90post00ConstantState fixed Sex Ratioof Voters(2)S

of women. This article also differs from the exist-ing research on women and politics in a fundamental way. While previous works have focused on women as policymak-ers, our emphasis is on women as voters in a political system. This is a critical area of research as women voters com-prise a signifi cant share in any election