My paper with Bijetri Bose and Shamma Alam, "Impacts of the COVID-19 lockdown on healthcare inaccessibility and unaffordability in Uganda," is now available as Ahead of Print at American Journal of Tropical Medicine & Hygiene.
I have a new working paper on how fertility behavior in Sub-Saharan Africa differ from Latin America, East Asia, and South Asia. The paper examines completed fertility using all DHS and MICS data from the four regions. Abstract is below and the Online Appendix is here.
Sub-Saharan Africa's fertility decline has progressed much slower than elsewhere. However, there is still substantial disagreement about why, partly because four leading potential causes—cultural norms, expected offspring mortality, land access, and school quality—are challenging to measure. I use large-scale woman-level data to infer what role each explanation plays in fertility differences between Sub-Saharan Africa and East Asia, South Asia, and Latin America, based on estimations of fertility outcomes by region, cohort, area of residence, and grade level. I show that the differences in fertility between Sub-Saharan Africa and the other regions first increase and then decrease with years of education. For women without education, fertility rates in Sub-Saharan Africa are comparable to those in Latin America. Similarly, for women with secondary education or higher, fertility rates in Sub-Saharan Africa align with those in South and East Asia. There are substantial and statistically significant differences for women with some primary education for all three comparison regions. The differences are more pronounced for children ever born than for surviving children. Overall, the results suggest that offspring mortality and the lower quality of primary schooling are the dominant reasons why fertility decline in Sub-Saharan Africa lags behind other regions.
Bijetri Bose, Shamma Alam, and I have a new paper, "Impacts of the COVID-19 Lockdown on Healthcare Inaccessibility and Unaffordability in Uganda," forthcoming in the American Journal of Tropical Medicine and Hygiene. The abstract is:
Several studies have reported adverse consequences of the COVID-19 lockdowns on the utilization of healthcare services across Africa. However, little is known about the channels through which lockdowns impacted healthcare utilization. This study focuses on unaffordability as a reason for not utilizing healthcare services. We estimate the causal impacts of the COVID-19 lockdown on healthcare inaccessibility and affordability in Uganda relative to the nonlockdown periods of the pandemic. We use nationally representative longitudinal household data and a household fixed-effects model to identify the impact of the lockdown on whether households could not access medical treatment and whether the reason for not getting care was the lack of money. We find that the lockdown in Uganda was associated with an 8.4% higher likelihood of respondents being unable to access healthcare when treatment was needed relative to the nonlockdown periods. This implies a 122% increase in the share of respondents unable to access healthcare. As lockdown restrictions eased, the likelihood of being unable to access medical treatment decreased. The main reason for the increase in inaccessibility was the lack of money, with a 71% increase in the likelihood of respondents being unable to afford treatment. We find little evidence that the effects of the lockdown differed by wealth status or area of residence. Our results indicate the need for policymakers to consider immediate social support for households as a strategy for balancing the disruptions caused by lockdowns.
Shamma Alam, Ishraq Ahmed, and I have a new working paper out examining the effects on food insecurity from Uganda's two Covid lockdowns. Abstract is below:
We examine the short- and medium-run impacts of two of the strictest Covid-19 lockdowns in the developing world, employing longitudinal data from Uganda. Household fixedeffects estimations show significant, immediate increases in food insecurity after the first lockdown and a continued negative impact three months after its lifting. The second lockdown’s medium-term impact was even worse, likely because of a compounding effect of a concurrent drought. The rising food insecurity was partly the result of the lockdownrelated reductions in the availability of paid work. Agricultural households were more likely to continue working and consequently saw smaller increases in food insecurity. Furthermore, the likelihood of engaging in agricultural work increased after the first lockdown, suggesting a switch to agriculture as a coping mechanism. The other coping mechanisms that households typically rely on for idiosyncratic shocks failed in the face of a worldwide shock, contributing to the sizeable increase in food insecurity.
My book chapter, "Economic Shocks and Children’s Education," for the Handbook of Labor, Human Resources and Population Economics has been published online. The chapter is joint with Shamma Alam and Claire Simpson.
My paper, "Birth Spacing and Fertility in the Presence of Son Preference and Sex-Selective Abortions: India's Experience Over Four Decades," is now available ahead of print on Demography's website: https://doi.org/10.1215/00703370-9580703. Demography is now an open-access journal, so the PDF is free to download for everybody. The paper should be out in the first 2022 issue.
My paper, "Birth Spacing and Fertility in the Presence of Son Preference and Sex-Selective Abortions: India's Experience Over Four Decades," is now in line for publication in Demography. The likely publication date is December 2021. Until then, you can find the final version here, together with the online appendix, and the GitHub repository with the paper and code (note, you will have to get the data yourself from DHS).
Well, it took long enough, but my paper on sex selection, birth spacing, and fertility in India has finally been accepted for publication in Demography. I will update the paper link and add the GitHub repository once the final version has been submitted. The upside of the long wait is that Demography is now open access.
My paper on birth spacing in India has been conditionally accepted at Demography. The new version based on Editor and referee comments is now available.
The new abstract is below:
Over the past four decades, the Hindu women in India most likely to use sex-selective abortions—well-educated women with no sons—had the most substantial lengthening of birth intervals and the most biased sex ratios. As a result, we now see cases that reverse the traditional spacing pattern, with some women with no sons having longer birth intervals than those with sons. Those least likely to use sex-selective abortions—less-educated women in rural areas—still follow the traditional pattern of short spacing when they have girls, with only limited evidence of sex selection. Because of the rapid lengthening in spacing, the standard fertility rates substantially overestimated how fast cohort fertility fell. Despite a convergence, cohort fertility is still 10%–20% higher than the fertility rate and above replacement level for all but the best-educated urban women. Infant mortality has declined substantially over time for all groups, with the fastest decline among the less educated. Short birth spacing is still associated with higher mortality, although considerably less so for the best-educated women. There is no evidence that repeated sex-selective abortions are associated with higher infant mortality for the child eventually born. Finally, it does not appear that the use of sex selection is declining.
Some shameless self-promotion: I will be the Howard J. Bosanko Professor of International Economics and Finance for 2020-23.
I plan to organize a series of Bosanko Lectures that will touch on different aspects of the challenges and opportunities that Sub-Saharan Africa faces – population, human capital, and business climate.
If any of you have ideas for speakers, preferably in academia or think tanks, I would love to hear them.
The announcement is at https://www.seattleu.edu/business/news-events/ (scroll down a bit).
I have just finished revising my paper on birth spacing and sex selection in India. The paper is here and the abstract below.
Using four rounds of India's National Family and Health Surveys and a competing-risk hazard model, I show that Hindu women's average birth intervals increased over the last four decades for all education groups. The most significant increases are among the women most likely to use sex selection. Despite the rise in average intervals, the likelihood of very short spacing did not change substantially. Hence, the increases come predominantly from the longer birth intervals getting even longer. As a result of the longer spacing, fertility rates significantly overestimated how fast cohort fertility fell. Although cohort fertility and the fertility rate have started to converge, the cohort fertility is still substantially higher than the fertility rate. Furthermore, cohort fertility is still at or above replacement level for all but the best-educated urban women. Finally, infant mortality risk has declined substantially over time for all groups, but fastest for the lower education groups, who are now close to the level of women with the most education. Short birth spacing is still associated with higher mortality, although the difference is small for the best-educated women. There is no evidence that the increasing use of sex selection is associated with higher infant mortality risk.
I am looking for a research assistant (RA) for a project that examines how the determinants of urban fertility vary across countries and to help organize a conference around the same topic. The RA will assist with data management, literature searches, and running estimations. The ideal candidate is interested in quantitative analyses, has a working knowledge of RStudio, and can commit to 5-10 hours a week for the Fall quarter. The pay is $16 an hour, and the hours are flexible. To apply, please email me a short statement of interest, a resume, an unofficial transcript, and an example of your R code, if you have one, by close of business Monday 21 October 2019.
Please contact me if you have any questions about the project or the position.
PS To apply for this position, you have to be a student at Seattle University (graduate or undergraduate).
One would expect that online labor markets are competitive. After all, there are many employers and many employees, and most of the labor markets have a high degree of flexibility. Recent papers suggest, however, that employers have substantial market power, even in markets such as Mechanical Turk. This is based on a low labor supply elasticity facing employers (if an employer changes the offered wage, the response in terms of how many people will want to work is relatively small). In a new paper, "What Labor Supply Elasticities do Employers Face? Evidence from Field Experiments," Nail Hassairi and I, show, using experimental data from Mechanical Turk, that this result is likely due to the way prior experiments and research was done. The paper is available here and the abstract is below.
We provide experimental evidence on the labor supply elasticity faced by employers, which is an essential measure of employer market power. We offered two different types of jobs, each with large randomized variations in pay, and observed the amount of work performed. We find no evidence of the strong employer market power suggested by prior research, with our elasticities close to unity. Furthermore, elasticities based on the total amount of work are significantly larger than if we use worker-level data as prior studies have done. Finally, elasticities differ by job type, suggesting that worker characteristics play a crucial role.
As a recent convert from Stata to R, one of my main problem is that the quality of the documentation in R is nowhere near what I was used to from Stata. That makes sense. Stata is a relatively expensive piece of software, and they sweat the details to ensure that the user is happy. R is more hit and miss.
A problem I ran into this week was trying to do a double-censored normal regression in R. This is where the outcome of interest can take one of three forms: left-censored, right-censored, or no censoring (where you observe the actual value). If you are an economist, think a Tobit model with both upper and lower censoring points. To make the problem even more fun, the censoring points varied across observations. Both cnreg and intreg in Stata do this, but I was having a hard time figuring out how to do this is in R.
As it turns out, the answer is interval regression from the survival package, which can also fit Tobit models. It took me a while to get past the name, and all the examples were of actual intervals, while my data are all points. To make it worse, coding it was not very intuitive and the documentation of limited help. Hence, this post is mainly to remind my future self how to do it.
To make it easy to see what is going on, I am going to use simulated data. The outcome is censored from below at zero and from above at 50 for the first 500 observations and 40 for the last 500 observations. You do not need the two censoring variables (censoring_low and censoring_high), but I find it easier to read.
library(tidyverse)
library(survival)
set.seed(2)
# Get 1,000 observation
b0 <- 17
b1 <- 0.5
id <- 1:1000
x1 <- runif(1000, min = -100, 100)
sigma <- 4
eps <- rnorm(1000, mean = 0, sigma)
df <- bind_cols(data_frame(id), data_frame(x1), data_frame(eps))
# Set up the data
df <- df %>%
mutate(
y = b0 + b1 * x1 + eps
) %>%
mutate(
# Convert all negative to zero
y_cen = if_else(y < 0, 0, y)
) %>%
mutate(
# Convert first 500 obs > 50 to 50
censoring_high = id < 500 & y > 50,
y_cen = replace(y_cen, censoring_high, 50),
# Convert last 500 obs > 40 to 40
censoring_low = id >= 500 & y > 40,
y_cen = replace(y_cen, censoring_low, 40)
)The trick is that you need to create a survival object with two variables indicating the left and right side of the "interval." Ironically, it was the documentation for Stata's intreg that helped me see how to do this. For those observations that are points and uncensored, you have the same value in both variables. For left-censored observations, you need minus infinity in the left variable and the censoring point in the right variable. For right-censored observations, you need the censoring point in the left variable and infinity in the right. In principle, you should be able to use NA instead of infinity, but I could not get it to work.
# Define left and rigth variables
df <- df %>%
mutate(
left = case_when(
y_cen <= 0 ~ -Inf,
y_cen > 0 ~ y_cen
),
right = case_when(
!censoring_low & !censoring_high ~ y_cen,
censoring_low | censoring_high ~ Inf
)
)Once you have the data set up, you use the survival object as the outcome and "interval2" as type together with a Gaussian distribution. Confusingly, you do not need a censoring indicator at all, despite the documentation's claim that this is unusual and is equivalent to no censoring. The model generates the censoring status automatically.
# OLS model
ols <- lm(y_cen ~ x1, data = df)
summary(ols)
# Survival model
res <- survreg(Surv(left, right, type = "interval2") ~ x1,
data = df, dist = "gaussian"
)
# Show results
summary(res)The summary does not provide much information on the amount of censoring, and I need that for my result table, so here is one way of getting that information. It is a little rough, but it works.
# Show descriptive stats on censoring (easier with tidyverse)
# Updated on 2018-11-28 to address updated packages
# Thank you to Kirk Wythers for alerting me to this
y_out <- as.tibble(res$y[ , 1:3])
# 'time1' in include only y values for observations used
obs_used <- length(y_out$time1)
# 2: left censored, 1: point, 0: right censored
censoring <- y_out %>%
group_by(status) %>%
summarise(
count = n()
)
left <- censoring %>%
filter(status == 2) %>%
select(count)
right <- censoring %>%
filter(status == 0) %>%
select(count)
# Example for including in LaTeX file
cat(
"Of the", obs_used, "observations,", as.numeric(left[1,1]), "are left censored and", as.numeric(right[1,1]), "are right censored."
)My latest paper, on how birth spacing changed in India with the introduction of sex selection, is now available. I am presenting a poster on this paper this coming Friday at the Population Association of America's annual meeting in Denver.
Title:
Abstract:
Strong son preference is typically associated with shorter birth spacing in the absence of sons, but access to sex selection has the potential to reverse this pattern because each abortion extends spacing by six to twelve months. I introduce a statistical method that simultaneously accounts for how sex selection increases the spacing between births and the likelihood of a son. Using four rounds of India’s National Family and Health Surveys, I show that, except for first births, the spacing between births increased substantially over the last four decades, with the most substantial increases among women most likely to use sex selection. Specifically, well-educated women with no boys now exhibit significantly longer spacing and more male-biased sex ratios than similar women with boys. Women with no education still follow the standard pattern of short spacing when they have girls and little evidence of sex selection, with medium-educated women showing mixed results. Finally, sex ratios are more likely to decline within spells at lower parities, where there is less pressure to ensure a son, and more likely to increase or remain consistently high for higher-order spells, where the pressure to provide a son is high.