Academia.eduAcademia.edu

Figure 5 – uploaded by Abdulaziz Reshid

See full PDFdownloadDownload figure
Note: The estimates are obtained from a linear probability regression of unintended indicator on family characteristics (listed in the horizontal axis). The coefficients reported in the x-axis represent the effect of family background on the probability of unintended birth. The omitted categories for ethnicity, regional states and religion are Oromo ethnic group, Christian religion and Oromia regional state, respectively. The 95% confidence interval is represented by the horizontal line corresponding to every individual estimate. Figure A2: A comparison of families with intended and unintended children

Figure 5 Note: The estimates are obtained from a linear probability regression of unintended indicator on family characteristics (listed in the horizontal axis). The coefficients reported in the x-axis represent the effect of family background on the probability of unintended birth. The omitted categories for ethnicity, regional states and religion are Oromo ethnic group, Christian religion and Oromia regional state, respectively. The 95% confidence interval is represented by the horizontal line corresponding to every individual estimate. Figure A2: A comparison of families with intended and unintended children

Related Figures (15)

Discrimination and Integration Studies
Discrimination and Integration Studies
Table 1: Test of Random Assignment of Child Gender Notes: a) The dependent variable is the girl dummy regressed on pre-determined characteristics of the mother and children. b) The null hypothesis is that the pre-determined characteristics do not jointly predict child gender. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Table 1: Test of Random Assignment of Child Gender Notes: a) The dependent variable is the girl dummy regressed on pre-determined characteristics of the mother and children. b) The null hypothesis is that the pre-determined characteristics do not jointly predict child gender. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Notes: The information on number of vaccinations has 584 missing observations, which reduced the baseline sample size in columns 1-3 from 15,843 to 15,259. The sample size for columns 4 to 6 is low because we exclude children who are younger than 12 months. The controls in columns 3 and 6 include mother characteristics (age, age at birth, age at marriage, years of education, religion, ethnicity, residence, ideal family size, number of prenatal visits, number of tetanus shots and non-home delivery), child characteristics (age in months, birth order, number of brothers and number of sisters) and survey year fixed effect. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 2: The gender gap in vaccination among intended and unintended children.
Notes: The information on number of vaccinations has 584 missing observations, which reduced the baseline sample size in columns 1-3 from 15,843 to 15,259. The sample size for columns 4 to 6 is low because we exclude children who are younger than 12 months. The controls in columns 3 and 6 include mother characteristics (age, age at birth, age at marriage, years of education, religion, ethnicity, residence, ideal family size, number of prenatal visits, number of tetanus shots and non-home delivery), child characteristics (age in months, birth order, number of brothers and number of sisters) and survey year fixed effect. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 2: The gender gap in vaccination among intended and unintended children.
Note: The above models are estimates based on a sample for 2000 and 2005. Data on number of months of breastfeeding duration is not available for 2011. The controls are described in footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 3: The gender gap in breastfeeding months among intended and unintendec children.
Note: The above models are estimates based on a sample for 2000 and 2005. Data on number of months of breastfeeding duration is not available for 2011. The controls are described in footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 3: The gender gap in breastfeeding months among intended and unintendec children.
Note: The data on anthropometry is obtained by directly measuring the height of the child who is at home during the survey. Children below the age of five but not present during the interview are missing in the sample, leaving us with 12,073 observations. The controls are described in the footnotes of Table 2. We also added square root of child age in months to improve the precision of our estimate. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 4: The gender gap in anthropometric outcomes among intended and unintended children
Note: The data on anthropometry is obtained by directly measuring the height of the child who is at home during the survey. Children below the age of five but not present during the interview are missing in the sample, leaving us with 12,073 observations. The controls are described in the footnotes of Table 2. We also added square root of child age in months to improve the precision of our estimate. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. Table 4: The gender gap in anthropometric outcomes among intended and unintended children
behavior. [his indicates that families in Ethiopia do not exhibit a son-biased fertility stopping
behavior. [his indicates that families in Ethiopia do not exhibit a son-biased fertility stopping
Notes: Columns 1-3 estimate the model after excluding children above 15 months. Columns 4-6 restrict the sample to children between 12 to 15 months. The controls are described in footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Notes: Columns 1-3 estimate the model after excluding children above 15 months. Columns 4-6 restrict the sample to children between 12 to 15 months. The controls are described in footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Notes: a) A total of 2,320 respondents have a non-number response to the survey question on ideal family size, of which approximately 1,826 respondents have a non-numeric response of “up to God” or “up to Allah.” Women with a response on ideal family size above 13 children are coded as 13. b) The information on vaccination has 584 missing observations. c) A child is considered fully vaccinated if he/she received all the nine recommended vaccinations by 12 months or older. d) The data on breastfeeding duration is not available for 2011. Hence, the analysis on
Notes: a) A total of 2,320 respondents have a non-number response to the survey question on ideal family size, of which approximately 1,826 respondents have a non-numeric response of “up to God” or “up to Allah.” Women with a response on ideal family size above 13 children are coded as 13. b) The information on vaccination has 584 missing observations. c) A child is considered fully vaccinated if he/she received all the nine recommended vaccinations by 12 months or older. d) The data on breastfeeding duration is not available for 2011. Hence, the analysis on
Note: The above models are estimates based on a sample for 2000 and 2005. Data on number of months of breastfeeding duration is not available for 2011. The controls are described in the footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. breastfeeding duration is solely based on the information from 2000 and 2005. e) Children below the age of five but not present during the interview are missing in the sample, leaving us with 12,073 observations.
Note: The above models are estimates based on a sample for 2000 and 2005. Data on number of months of breastfeeding duration is not available for 2011. The controls are described in the footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively. breastfeeding duration is solely based on the information from 2000 and 2005. e) Children below the age of five but not present during the interview are missing in the sample, leaving us with 12,073 observations.
Note: Observations with non-numeric responses, approximately 2320, are excluded from the sample. The controls are described in the footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Note: Observations with non-numeric responses, approximately 2320, are excluded from the sample. The controls are described in the footnotes of Table 2. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Table A4: Stated preferences of wives and husbands on ideal family size
Table A4: Stated preferences of wives and husbands on ideal family size
Table A5: Gender gap in anthropometric outcomes among intended and unintended children, 0-15 months Note: The controls are described in the footnotes of Table 4. We also added the square root of child age in months to improve the precision of our estimate. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Table A5: Gender gap in anthropometric outcomes among intended and unintended children, 0-15 months Note: The controls are described in the footnotes of Table 4. We also added the square root of child age in months to improve the precision of our estimate. The standard errors are reported in parentheses. The signs ***, *, and * represent a 1, 5 and 10 percent significance level, respectively.
Figure II. Countries where excess female child mortality (ages 1-4) was found in the 2000s
Figure II. Countries where excess female child mortality (ages 1-4) was found in the 2000s
Note: The estimates are obtained from a linear probability regression of unintended indicator on family background (reported on the horizontal axis) and their interactions with the girl dummy. The coefficients reported in the x-axis are estimates from the interaction between girl indicator and family background. The omitted categories for ethnicity, regional states and religion are Oromo ethnic group, Christian religion and Oromia regional state, respectively. The 95% confidence interval is represented by the horizontal line corresponding to every individual estimate. The estimates for Tigre ethnicity, Addis Ababa (rural) and Tigray regions are statistically significant at 5% level, while other estimates are not statistically significant. Figure A3: Gender difference on the effect of family background on unintended birth Figure A3: Gender difference on the effect of family background on unintended birth
Note: The estimates are obtained from a linear probability regression of unintended indicator on family background (reported on the horizontal axis) and their interactions with the girl dummy. The coefficients reported in the x-axis are estimates from the interaction between girl indicator and family background. The omitted categories for ethnicity, regional states and religion are Oromo ethnic group, Christian religion and Oromia regional state, respectively. The 95% confidence interval is represented by the horizontal line corresponding to every individual estimate. The estimates for Tigre ethnicity, Addis Ababa (rural) and Tigray regions are statistically significant at 5% level, while other estimates are not statistically significant. Figure A3: Gender difference on the effect of family background on unintended birth Figure A3: Gender difference on the effect of family background on unintended birth
Related topics:
EconomicsMedicineUnintended birth
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved