Global Infant Mortality: Correcting for Undercounting

Fiscal decentralization and China's regional infant mortality

“…The empirical model is Jin and Sun (2011) equation one but modified by the addition of several macroeconomic variables plus their assumption of a normal error term is relaxed to account for the one-sided nature of IMR underreporting as suggested by Anthopolos and Becker (2010).…”

“…positive truncations of a normal distribution with unknown mean (m) and variance (var-u). Anthopolos and Becker (2010) assume that "u" has a half-normal distribution which restricts "m" to be equal to zero and nested within the more general truncated normal. As theory provides no guidance as to what the underlying density function of IMR underreporting might be, we examine both half-normal and a more generalized truncated normal distribution.…”

“…Indeed, the one-child policy has led to millions of females aborted according to the most recent (2005) mini census (Ridley, 2012) and over-reporting of twins that do not actually exist (Huang, Lei, & Zhao, 2014). Infant mortality in middle income countries is most likely to be underreported in general (Anthopolos & Becker, 2010), and China as a middle income country with a strict population control policy in a society biased toward male births is therefore quite likely to have high IMR underreporting. Underreporting suggests that the true IMR is consistently above the reported rate of IMR variation across time and provinces in China including very recent censuses (e.g., for the 2010 census, see Du & Zhang, 2014;Zhao & Yang, 2013).…”

“…Bauer, 1990;Fare & Grosskopf, 2005), it has only recently been considered in explaining infant mortality (Anthopolos & Becker, 2010) and never within China. The SFM with a one-sided error is applied with the assumption that the reported IMR is much lower than the true IMR resulting in a one-sided "frontier" that, if underreporting did not happen, would represent a true IMR frontier.…”

Air Pollution and Infant Mortality: Evidence from the Expansion of Natural Gas Infrastructure

“…There are usually wide differences in the infant mortality rates reported by the national statistical offices of individual countries and those released by international agencies such as the WHO. Note that the under-reporting is mostly one-sided, i.e., there are far more unreported infant deaths than unreported live births that survive the first year in life (Anthopolos and Becker, 2010). The statistics compiled by these international agencies usually adjust for under-reporting by combining information from various demographic surveys, census data, the official vital statistics registries, and by using modeling assumptions in an attempt to reduce the impact of under-reporting (Gruber et al, 2012;Lopez et al, 2001).…”