Research

WORKING PAPERS


Toxic Truth: Lead Exposure and Fertility Choices

(with Karen Clay and Edson Severnini) (JMP)

Although environmental insults have been shown to affect short- and long-run economic outcomes, how they evolve over time and may have recurrent effects through different channels has been overlooked. In this study we examine the impact of lead on fertility in U.S. counties in the 1980s, when airborne lead concentration decreases considerably due to the phase down of lead in gasoline, and more recently, when legacy deposition in soil is still an issue. Evidence from the epidemiology literature suggests that exposure to lead impairs the reproductive system of both males and females, leading to lower fecundity (i.e. the physiological ability to have children). It is unclear, however, whether this would cause a decrease in fertility (i.e. the actual production of offspring). Households could take actions to avoid lead exposure and/or to remediate undesirable consequences of some exposure. In this paper, we leverage the implementation of Clean Air Act regulations and the 1944 Interstate Highway System Plan within a fixed-effect instrumental variable approach, and find two main results. First, exposure to airborne lead causes a reduction on the number of births and birth rates, indicating that avoidance and/or compensatory behavior might not fully offset the pathologic effects of lead. Our preferred estimate indicates that, had airborne lead remained in the 1978 level, approximately 200 babies would have not been born annually. Second, such a reduction in fertility rates seems smaller for high-educated households (those with mothers with high school or higher education), suggesting that defensive investments may play a role in attenuating the potentially harmful effects of lead. In the end, back-of-the-envelope calculations based on our estimates indicate that the benefits of the phasing down of lead in gasoline in the 1980s exceeded the costs of the policy.


The Effect of Weather on Mortality in Russia: What if People Adapt?

(Revision requested by Journal of Environmental Economics and Management)

The response to climate change is likely to involve a host of short- and long-run adaptation strategies. Previous research has focused on specific adaptation responses: e.g. change in electricity use, time allocation change. Little is known about overall adaptation response when all potential adaptation techniques are taken into account. In this paper I construct predictions for the climate change impact on mortality in Russia while accounting for a general adaptation response. Namely, using regional monthly mortality and daily temperature data, I estimate a flexible non-parametric relation between weather and mortality. I find that people are better adapted to temperatures that they have been exposed to more often: damages from high temperatures are smaller for regions where the average summer temperature is higher and damages from low temperatures are lower in regions where winters are usually more severe. On the basis of these estimates I propose a way to account for general adaptation response to climate change: If some currently cold region were to regularly experience high temperatures in the future, its response would be similar to the response of warm regions that currently experience such high temperatures on a regular basis. Because the analysis involves observations from a single country it is possible that adaptation would be similar across regions. I illustrate this approach constructing predictions for the impact of climate change on mortality in Russia using business-as-usual temperature predictions from climate change models. No-adaptation specification predicts a one percent increase in mortality by 2070-2099. However, when a broad adaptation response is taken into account my model predicts a decrease in mortality by 0.7 percent by 2070-2099. 


Adaptation to Climate Change through Migration

In this paper I study migration as an adaptation mechanism to climate change. I estimate a discrete location choice model, in which households choose residence locations on the basis of potential earnings, moving costs, climate amenities, and population density. I treat population density as endogenous using geological structure as an instrument. This model allows me to estimate counterfactual migratory responses and welfare changes resulting from non-marginal changes in temperature, such as these predicted by most climate models. I also account for general equilibrium effects on population densities arising from individual migration decisions. I find that the costs of climate change are likely to be quite large. In the absence of migration, American households would require their incomes to increase by 20-30 percent on average to attain their present day level of utility. The distribution of those costs is uneven across geographical locations. Some areas in the South would require more than 50 percent increases in terms of current incomes, while some northern locations actually see benefits around 20 percent. Allowing for migratory responses decrease those extremes considerably because of the resulting shifts in population densities.  For the hardest hit areas, migration would reduce the costs by more than 10 percent (4-5 percentage points). Areas benefiting the most from climate change without migration would see their benefits reduced due to migratory inflows from other locations. 


When Are Natural Resources Curses or Blessings? Evidence from the United States 1936-2015 (with Karen Clay)

This paper uses new state-level panel datasets spanning 1936-2015 and the three most valuable natural resources during the period – oil & natural gas, coal, and agriculture – to investigate when natural resources are a curse or a blessing as measured by a range of economic outcomes including growth in per capita income, growth in population, and growth in employment.  The long time period and the use of three resources facilitates comparisons over time and across resources with declining and increasing employment and across non-renewable and renewable resources. To examine the relationships between resources and outcomes, we use a shift-share approach that is closely related to Allcott and Keniston (2015) and Bartik (1991). National resource-specific employment is interacted with cross sectional variation in resource specific reserves in 1935 and is related to economic outcomes. The paper has two main findings. First, if we require increases and decreases in resource-specific employment to have symmetric effects, oil & gas and coal appear to be blessings across a range of outcomes. Increases in employment in these sectors are positively and significantly related to growth in per capita income, population, and employment. Agriculture is neither a blessing nor a curse. That is, the effect on outcomes is rarely statistically significant. Second, if we allow the coefficients to differ across increases and decreases in resource employment, we find evidence of both blessings and curses. The differential effects of busts are large and statistically significant for oil & natural gas during 1936-1974 and for coal during 1975-2015. 

 

International Politics and Oil Trade: Evidence from Russian Firms

(with Kevin K.Tsui)

It is widely argued that oil exporters could use their natural resources as a weapon to punish adversaries and reward allies. Yet empirical analysis of these claims has been elusive due to lack of data. Using a novel dataset on Russian companies’ oil exports over 1999–2011, we show that a decline in relations between Russia and another country, measured by divergence in their United Nations General Assembly (UNGA) voting patterns, considerably reduces the value of Russian oil exports to that country. The effect is more pronounced for state-owned companies. A deterioration in political relations and associated decrease in oil exports are costly for Russian companies. They experience a decline in profitability following a breakdown in political relations between Russia and those companies’ main export destination countries. Finally, we show that a deterioration in political relations with Russia is costly for the countries importing oil from Russia as their total oil imports decline, suggesting that, at least in the short run, it is costly for these countries to find close substitute for Russian oil. Notably, such adverse effect of political relations on oil importers is pretty recent phenomenon observed over 2000-2011, which coincide with the rise of Vladimir Putin to power, such patterns are absent in the earlier 1992-1999 period.


WORK IN PROGRESS


Long Term Effect of Polluton on Children’s Health: Evidence from Russia


Temperature and Health Outcomes: Evidence from Moscow Ambulance Daily Calls Data


Supply-Side Constrains of Adaptation to Climate Change: Air Conditioners Purchases during 2010 Heat Wave in Russia


© Margarita Portnykh 2016