The Gender Impact of Crime on Time Use and Children Development
in the Colombian Rural Areas
Manuel FernándezAna María Ibáñez
Ximena Peña
June 4th 2010
Objectives
• Examine the coping strategies households adopt when confronting a violent shock. – Heterogeneous responses to violent shocks by
gender– Whether this heterogeneity leads to changes
in bargaining power within the household.
• Evaluate the impact of violent shocks:– On child development– Differentiated impact of shocks on boys and
girls
Results
• Violence is not randomly targeted, but purposively directed to:– Relatively better-off groups – Certain municipalities.
• Coping strategies of households:– Collective strategies: reliance on collective
action, cooperation with the government authorities, and seeking support from formal organizations or other community members
– Individual strategies to mitigate the impact of the shocks is less frequently used
Results
• Need of households to generate additional income seems to be pushing males to participate in formal labor markets.
• Women partially substitute men in agricultural activities in their land plot and as seasonal agricultural workers, which implies reductions in their leisure time.
• Families are unable to protect their children from the impact of violent shocks– Children’s anthropometric measures are affected,
especially for boys– No evidence of a negative effect on cognitive
development.
Colombia: 50 years of internal conflict
Attacks on Armed Forces and Infrastructure 1998-2009
500
100
01
500
200
02
500
300
0T
otal
Atta
cks
2000 2005 2010Year
Source: Vice-Presindent's Observatory of Human Rights
Homicide Rates per 100,000 Inhabitants 1946-2009
02
04
06
08
0
1940 1950 1960 1970 1980 1990 2000 2010Year
Source: National Police Department
Number and Victims of Massacres 1993-2009
Persons Wounded and Killed by Landmines 1990-2009
05
001
000
150
0T
otal
vic
tims
of m
assa
cres
05
01
001
502
002
50N
umb
er o
f ma
ssac
res
1995 2000 2005 2010Year
Massacres Total victims of massacres
Source: Vice-Presindent's Observatory of Human Rights
02
004
006
008
001
000
1990 1995 2000 2005 2010Year
Wounded Killed
Source: Vice-Presindent's Observatory of Human Rights
Internal conflict
• 56% of municipalities have had at least one massacre between 1993 and 2009
• The Vice-President’s Observatory of Human Rights reports a total of 8,900 victims of massacres in the period described. 4,778 (53%) in the past ten years.
• 22 massacres in 2009
Victims of massacres
0
>4 - 22
23 - 52
53 - 109
110 - 209
210 - 569
0 270 540135 Kilometers
Map 1. Victims of Massacres(1993-2009)
Source: Vice-President's Observatoryof Human Rights
• 95.6% of municipalities have had people forcefully displaced between 1997 and 2009
• The Presidential Agency for Social Accion (Acción Social) reports that 3.3 Million people have been forcefully displaced (7.9% of the Colombian population).
• The Internal Displacement Monitoring Centre (IDMC) reports that Colombia has the second highest magnitude worldwide after Sudan
Number of persons forcefully displaced
0
>0 - 5,135
5,136 - 10,658
10,659 - 18,464
18,465 - 33,151
33,152 - 67,189
0 270 540135 Kilometers
Map 2. Forced Displacement(1997-2009)
Source: Acción Social
• The Colombian Longitudinal Survey of Wealth, Income, Labor and Land (CLS-WILL)
– 10,000 households: 6,000 urban and 4,000 rural.– Rural sample covers (mostly) small agricultural
producers of stratum one and is representative of four micro-regions. The sample covers 1,000 households for each micro-region.
– The first wave was collected between April and July 2010
– The preliminary results contained in this paper were calculated using observations from the 2,501 rural households available in mid-May
The Data
Households Affected by Violent ShocksShocks by household
% of households affected
Idiosyncratic shock 33 1.32%Covariate shock 281 11.24%General Shocks 312 12.48%
Incidence of shocks: 11.24% covariate and 1.32% idiosyncratic
Types of Shocks
Shock Description
Idiosyncratic shockDestruction or theft of households goods
Idiosyncratic shock Direct victimizationCovariate shock Attacks from armed groupsCovariate shock Gang related violenceCovariate shock Cattle theftCovariate shock ExtortionsCovariate shock Massacres
Distribution of Time Use by Gender and Violent Shocks
No Violent shocks
Violent shocks
No Violent shocks
Violent shocks
Variable Sample Mean Sample Mean Sample Mean Sample Mean
0.07 0.11 *** 0.23 0.25
(0.14) (0.16) (0.27) (0.28)
0.04 0.04 0.26 0.23
(0.12) (0.13) (0.29) (0.27)
0.40 0.38 0.43 0.44
(0.18) (0.17) (0.18) (0.17)
0.50 0.47 * 0.08 0.07(0.21) (0.20) (0.13) (0.12)
Number of Observations
2,004 251 1,890 236
Difference Between Means (t-
test)*
Standar errors in parentheses(*)*** p<0.01, ** p<0.05, * p<0.1
% of time spend in domestic chores and taking care of children and other members of the house
Female Male
% of time working inside the household farm
% of time working in other households farms
% of leisure time and other activities
Difference Between Means (t-
test)*
Participation in Labor Markets
No Violent shocks
Violent shocks
No Violent shocks
Violent shocks
Variable Sample Mean Sample Mean Sample Mean Sample Mean
0.14 0.18 * 0.36 0.43 *(0.34) (0.38) (0.48) (0.49)
0.09 0.15 ** 0.28 0.26(0.28) (0.36) (0.45) (0.43)
0.09 0.15 ** 0.49 0.48(0.29) (0.36) (0.50) (0.50)
Number of Observations
1,898 234 1,794 223
Standar errors in parentheses(*)*** p<0.01, ** p<0.05, * p<0.1
Difference Between Means*
Difference Between Means*
Worked for a salary in the past 12 months
Tried to find a job in the past 12 months
Worked as a seasonal worker in the past 12 months
Female Male
Nutritional Status of Children and Peabody Picture Vocabulary Test
Results by Gender and Violent Shocks
No Violent shocks
Violent shocks
No Violent shocks
Violent shocks
Sample Mean Sample Mean Sample Mean Sample Mean
-0.24 -0.52 * -0.40 -1.02 **(1.70) (1.03) (1.47) (1.76)
-0.78 -0.93 -1.03 -1.72 **(1.99) (1.14) (1.66) (2.16)
93.68 94.86 94.45 95.57(16.85) (15.87) (18.06) (16.99)
Number of Observations
Nutritional Status
307 58 321 48
Number of Observations Peabody
Picture Vocabulary Test Results
488 76 495 74
Standar errors in parentheses(*)*** p<0.01, ** p<0.05, * p<0.1
Difference Between Means (t-
test)*
Peabody Picture Vocabulary Test
Results
Female Male
Z-Score weight for age
Z-Score height for age
Difference Between Means (t-
test)*
Descriptive Statistics Control Variables
No Violent shocks
Violent shocks
Variable Sample MeanSample Mean
47.24 43.88 *** ***
(13.69) (13.83)
0.19 0.24 **
(0.39) (0.43)
0.76 0.71 *
(0.43) (0.46)5.18 6.03 *** ***(3.52) (3.70)
-0.02 0.13 ** ***
(0.98) (0.15)1.05 0.81 ***(1.59) (1.35)0.44 0.43(0.69) (0.69)1.49 1.35 *(1.39) (1.28)0.31 0.30(0.59) (0.57)
Number of Observations 2,216 285
Standar errors in parentheses(*)*** p<0.01, ** p<0.05, * p<0.1
Higher level of education in household
Number of members from extended family
Number of children under 5 years of age
Number of children between 5 and 18
Number of members 65 or older
Wealth Index
Equality of Distribution Functions (Kolmogorov
Smirnov)*
Difference Between Means (t-
test)*
Househols with both head and spouse
Age household head
Female household head
Probability of Receiving a Violent Shock in the Household in the Past 12 Months (marginal effects after probit)
(1) (2) (3)
-0.005* -0.004* -0.004*(0.003) (0.002) (0.003)
0.000 0.000 0.000(0.000) (0.000) (0.000)
0.024 0.007 0.007(0.021) (0.018) (0.018)
-0.021 -0.011 -0.012(0.019) (0.017) (0.018)
0.003 0.002 0.002(0.002) (0.002) (0.002)
0.064** 0.045* 0.046*(0.026) (0.024) (0.024)
0.073*** 0.051** 0.052**(0.027) (0.024) (0.025)
0.099*** 0.079*** 0.080***(0.028) (0.027) (0.027)
0.096*** 0.077*** 0.079***(0.029) (0.028) (0.028)
-0.005(0.007)
0.001(0.011)
0.001(0.007)
-0.008(0.012)
Observations 2501 2501 2501Log-Likelihood -859.2 -766 -765.0Pseudo-R2 0.0279 0.133 0.134Municipal Fix Effects No Yes YesRobust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
Number of members 65 or older
Wealth Quantile 3
Wealth Quantile 4
Wealth Quantile 5
Number of members from extended family
Number of children under 5 years of age
Number of children between 5 and 18
Wealth Quantile 2
Age household head
Age household head squared
Female household head
Househols with both head and spouse
Higher level of education in household
Strategies to cope with Violent Shocks by Gender of the Households Head
Male household head Female Household headVariable Mean Mean
18.70% 23.17%(0.39) (0.42)
13.48% 10.98%(0.34) (0.31)
20.87% 13.41%(0.41) (0.34)
0.43% 2.44%(0.07) (0.16)
4.78% 3.66%(0.21) (0.19)
12.61% 12.20%(0.33) (0.33)
53.48% 56.10%(0.50) (0.50)
Number of Observations 230 82Standar errors in parentheses(*)*** p<0.01, ** p<0.05, * p<0.1
Other
Nothing was donde
Households increased cooperation with authorities and/or got together with other households to defend themselves
Seek help from national and international institutions
Seek help from other members of the family or the community
Some member of the family raised the hours worked or went looking for a job
Spend their savings or sold/rent the assets of the household
Effect of Violent Shocks on Time Use
Household controls include: Age of the household head, female household head dummy, households with both head and spouse dummy, higher level of education in the household, number of members from extended family, number of children under 5 years of age, number of children between 5 and 18 years of age and number of members 65 or older
Variables
% of time working inside the household
farm
% of time working in other households farms
% of leisure time and other
activities
% of time spend in domestic chores and taking care of
children and other members of the house
-0.324*** -0.702*** -0.037*** 0.530***(0.013) (0.021) (0.005) (0.007)
-0.027 -0.081** 0.027** 0.005(0.028) (0.037) (0.012) (0.017)
0.104*** 0.112* -0.029** -0.012(0.032) (0.064) (0.015) (0.022)
Constant 0.190*** 0.414*** 0.305*** -0.019(0.044) (0.070) (0.020) (0.025)
Observations 4381 4381 4381 4381Log-Likelihood -2110 -2082 1483 -477.6Total number of violent shocks 487 487 487 487Municipal Fixed Effects Yes Yes Yes YesHousehold controls Yes Yes Yes YesEstmated total effect on women 0.077 0.031 -0.002 -0.007t-stat 3.238 0.520 -0.201 -0.534p-val 0 0.603 0.841 0.593
*** p<0.01, ** p<0.05, * p<0.1
Sex (1 if Female)
Violent shock (1 if recieved a violent shock)
Sex x Violent schock
Robust standard errors in parentheses (cluster by household)
Effect of Violent Shocks on Time Use
-0.100
-0.050
0.000
0.050
0.100
0.150
Effect of Violent Shock Difference in Difference Estimation Total Effect on Women
Effects of Violent Shocks on Labor Outcomes
Household controls include: Age of the household head, female household head dummy, households with both head and spouse dummy, higher level of education in the household, number of members from extended family, number of children under 5 years of age, number of children between 5 and 18 years of age and number of members 65 or older
Tried to find a job in the past 12 months (1 if
he/she tried to find a job)
Worked for a salary in the past 12 months (1 if he/she
worked for a salary)
Worked as a seasonal worker in the past 12 months (1 if he/she worked as a seasonal worker)
-0.231*** -0.267*** -0.450***(0.012) (0.013) (0.013)
-0.061** 0.056* -0.042(0.031) (0.033) (0.034)
0.085** -0.024 0.078*(0.035) (0.040) (0.040)
Constant 0.576*** 0.465*** 0.497***(0.040) (0.044) (0.01)
Observations 4149 4149 4149R-squared 0.118 0.160 0.262Total number of violent shocks 457 457 457Municipal Fixed Effects Yes Yes YesHousehold controls Yes Yes YesEstmated total effect on women 0.024 0.032 0.037t-stat 0.968 1.174 1.464p-val 0.333 0.241 0.143
*** p<0.01, ** p<0.05, * p<0.1
Sex x Violent schock
Sex (1 if Female)
Violent shock (1 if recieved a violent shock)
Robust standard errors in parentheses (cluster by household)
Effects of Violent Shocks on Labor Outcomes
Tried to Find a Job in The Past 12 Months
Worked for a Salary in the Past 12 Months
Worked as a Seasonal Worker in the Past 12 Months
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
Effect of Violent Shock Difference in Difference Estimation Total Effect on Women
Effects of Violent Shocks on Child's Nutritional Status as Measured by Height for Age z-score by Age Group
Household controls include: Age of the child in months, female household head dummy, boths parents live with the child dummy, higher level of education in the household, number of members from extended family, number of children under 5 years of age, number of children between 5 and 18 years of age and number of members 65 or older
age<=20 20<age<40 age>40 All Sample
0.104 0.118 0.049 0.107(0.217) (0.111) (0.083) (0.077)
-0.592* -0.306 -0.235 -0.403**(0.325) (0.223) (0.192) (0.175)
0.727* 0.153 0.128 0.351*(0.419) (0.247) (0.238) (0.194)
-0.228 0.217 0.177 0.059(0.288) (0.152) (0.112) (0.116)
0.012 0.164 0.243* 0.186(0.264) (0.155) (0.125) (0.117)
0.130 0.378*** 0.201 0.231**(0.217) (0.141) (0.135) (0.097)
-0.744** 0.274* 0.078 -0.082(0.366) (0.150) (0.130) (0.140)
Constant 0.528 -0.332 0.158 0.137(0.426) (0.366) (0.373) (0.181)
Observations 231 251 252 734R-squared 0.167 0.138 0.140 0.068Total number of violent shocks 39 33 34 106Regional Fixed Effects Yes Yes Yes YesEstmated total effect on girls 0.135 -0.154 -0.107 -0.0514t-stat 0.519 -1.264 -0.780 -0.486p-val 0.604 0.208 0.436 0.627
*** p<0.01, ** p<0.05, * p<0.1
Sex (1 if Female)
Violent shock (1 if recieved a violent shock)
Sex x Violent schock
Robust standard errors in parentheses (Cluster by household)
Wealth Quantile 2
Wealth Quantile 3
Wealth Quantile 4
Wealth Quantile 5
Effects of Violent Shocks on Child's Nutritional Status as Measured by Height for Age z-score
by Age Group
Age<=20 Months 20>Age>=40 Months Age>40 Months All Sample-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Effect of Violent Shock Difference in Difference Estimation Total Effect on Women
Effects of Violent Shocks on Child's Nutritional Status as Measured by Weight for Age z-score by Age Group
Household controls include: Age of the child in months, female household head dummy, boths parents live with the child dummy, higher level of education in the household, number of members from extended family, number of children under 5 years of age, number of children between 5 and 18 years of age and number of members 65 or older
age<=20 20<age<40 age>40 All Sample
0.231 0.007 0.040 0.075(0.219) (0.109) (0.103) (0.078)
-0.522* -0.484* -0.302 -0.466***(0.300) (0.280) (0.198) (0.179)
0.405 0.313 0.094 0.309(0.397) (0.314) (0.245) (0.193)
-0.125 0.174 0.475*** 0.189(0.293) (0.160) (0.165) (0.129)
0.184 0.139 0.277** 0.234**(0.253) (0.163) (0.113) (0.109)
0.064 0.314** 0.301** 0.233**(0.236) (0.140) (0.133) (0.103)
-0.420 0.082 0.147 -0.053(0.357) (0.147) (0.137) (0.134)
Constant 0.246 -0.130 -0.303 0.161(0.446) (0.350) (0.517) (0.196)
Observations 231 251 252 734R-squared 0.113 0.120 0.091 0.059Total number of violent shocks 39 33 34 106Regional Fixed Effects Yes Yes Yes YesEstmated total effect on girls -0.117 -0.171 -0.208 -0.158t-stat -0.489 -1.186 -1.470 -1.557p-val 0.626 0.237 0.143 0.120
*** p<0.01, ** p<0.05, * p<0.1Robust standard errors in parentheses (Cluster by household)
Violent shock (1 if recieved a violent shock)
Sex (1 if Female)
Wealth Quantile 2
Wealth Quantile 3
Wealth Quantile 4
Wealth Quantile 5
Sex x Violent schock
Effects of Violent Shocks on Child's Nutritional Status as Measured by Weight
for Age z-score by Age Group
Age<=20 Months 20>Age>=40 Months Age>40 Months All Sample-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Effect of Violent Shock Difference in Difference Estimation Total Effect on Women
Effects of Violent Shocks on Child's Peabody Picture Vocabulary Test Results
Household controls include: Age of the child in months, female household head dummy, boths parents live with the child dummy, higher level of education in the household, number of members from extended family, number of children under 5 years of age, number of children between 5 and 18 years of age and number of members 65 or older
age<=5 5<age<=7 age>7 All Sample
-0.690 0.597 -2.755** -1.213(1.349) (2.184) (1.226) (0.912)
-1.611 -4.380 0.352 -1.169(1.783) (3.055) (2.377) (1.742)
0.645 6.786 4.238 2.358(2.663) (4.675) (2.925) (2.212)
Wealth Quantile 2 2.127 3.621 1.886 2.259*(1.644) (3.159) (1.918) (1.362)
Wealth Quantile 3 0.786 9.206*** 6.857*** 5.471***(1.844) (3.044) (1.808) (1.385)
Wealth Quantile 4 3.246 4.665 4.595** 3.751***(2.018) (3.519) (1.916) (1.401)
Wealth Quantile 5 1.603 8.194** 5.072** 4.720***(1.85) (3.47) (2.15) (1.62)
Constant 118.164*** 51.284*** 101.382*** 83.251***(6.253) (12.344) (6.927) (3.572)
Observations 339 281 513 1133R-squared 0.290 0.235 0.202 0.236Total number of violent shocks 46 50 54 150Regional Fixed Effects Yes Yes Yes YesEstmated total effect on women -0.967 2.406 4.589 1.189t-stat -0.479 0.643 2.515 0.734p-val 0.633 0.521 0.0122 0.463
*** p<0.01, ** p<0.05, * p<0.1
Sex (1 if Female)
Violent shock (1 if recieved a violent shock)
Sex x Violent schock
Robust standard errors in parentheses (Cluster by household)
Effects of Violent Shocks on Child's Peabody Picture Vocabulary Test Results
Age<=5 Years 5>Age>=7 Years Age>7 Years All Sample-6
-4
-2
0
2
4
6
8
Effect of Violent Shock Difference in Difference Estimation Total Effect on Women