Good Example Of Key Results Article Review

International Differences in Emissions Intensity and Emissions Content of Global Trade: A Review of the Article

Introduction (Weaknesses and Strengths)
It is important to understand the emissions content of trade, for reasons that should be obvious in light of the current worldwide battle against global warming. In this study we compare the national emissions content of each country against the global average. This study is subject to the experimental weakness known as aggregation bias, in which the behavior of a group is mistaken for the behavior of each individual within that group. By diversifying the number of economic sectors, we can reduce the effect of this source of aggregation bias. The study found that the differences in emissions between each country had little effect on their strength as a trade competitor in the global market.

Issues and Methodology

The study found that the international differences in emissions intensity does not significantly affect international patterns in trade. Emissions intensity was measured as industrial CO2 emissions (measured in kilotons) divided by real GDP. The empirical results of the study were in agreement with many previous studies that answered related questions about the emissions content of trade.
The study looks at year 2000 data on production techniques and bilateral trade data for 39 countries and 41 industrial sectors. First, it combines OECD input-output data with International Energy Agency (IEA) emissions data, sorted by industrial sector and country. Then, taking the origin of international goods into account, it calculates sector-specific and country-specific emissions intensity coefficients. The sectors of industry in the IEA emissions data included electricity and heat production, which accounted for 35 percent of world CO2 emissions in 2000. It also included unallocated auto producers, other energy industries, manufacturing and construction, transport, and other sectors. The validity of the emissions allocation methods was checked against official allocations obtained from surveys by national governments.
Just 39 countries account for 76 percent of the world's total emissions, according to the IEA. The ratio of CO2 emissions from electricity and heat production to CO2 emissions from industrial production varies from country to country. It is a variable that depends on the characteristics of electricity generation plants, capital shares in production, emissions from transportation, and the size of each country's manufacturing sector. For example, France derives 79 percent of its energy from nuclear energy, so its CO2 emissions from electricity generation is only 10 percent of the national total. On the opposite end of the spectrum, India's production of CO2 emissions from electricity generation accounts for 81 percent of the national total. The study found that industrial activities account for between 34 and 82 percent of each country's total emissions.
There is a non-linear relationship between a country's level of development and its emissions intensity. For example, countries with less capital and resources will likely produce fewer CO2 emissions per unit of output. But a capital and labor-rich country such as the United States will have both higher productivity and less tolerance for pollution. Emissions intensity tends to decline with rising TFP, or total factor productivity. The TFP is a measure of an economy's long term technological change. The higher a country's TFP, the lower its requirements of capital and labor per unit of output. Previous studies have shown there to be a high correlation between TFP and energy conversion efficiency. This may be achieved by using more advanced technology that requires fewer resources of all kinds.
The data indicates that there is an overall negative but weak correlation between capital intensity and CO2 emissions intensity. One of the strongest determinants of CO2 emissions intensity was actually found to be a country's most prevalent technology choice, specifically coal usage for the production of electric power. The fossil fuel emissions for each country that came from coal ranged from 1 percent for Argentina to 82 percent for China. The choice in power generation technologies is affected by a country's endowment in alternative methods such as hydro-power and nuclear techniques, and is also affected by factors such as national income and technical expertise (a form of human capital).
The emissions efficiencies were modeled by country and by industrial sector. Emissions intensity was quantified by the amount of CO2 emissions required to produce one US dollar's worth of output. Efficiency coefficients were calculated using a regression formula that incorporated country-specific efficiency differences, normalized to the United States. This way, negative values of country-specific coefficient estimates indicate countries that produce fewer emissions per unit of output than the United States. The calculation of these coefficients showed that less than a quarter of the countries in the data set use significantly cleaner production techniques than the United States.
The analysis of the data focuses on measuring the importance of the composition and technique in determining the emissions content of trade. The two effects have different policy implications for each country. The composition effect refers to a change in the share of “dirty” (emissions-intensive) goods that comprise the overall GDP. This change can come about because of a price change that favors their production, such as a change due to improved manufacturing efficiency. The technique effect is in reference to a change in emissions intensity. This effect could be caused, for example, by a higher pollution tax being levied. The technique effect generally refers to the idea that as the GDP per capita of a country rises, so does the demand for more environmental regulation, thereby reducing the pollution output of industrial production. The fact that coal usage and country-level technical proficiency are good indicators of the emissions intensity of a country suggests that the technique effect may be important in determining the emissions content of trade as well.
The study compared the predictions of two empirical models for measuring the emissions content of trade. The models differ in their theoretical assumptions about whether the price of emissions rights is the same across countries. The models that were used were the HOV (Heckscher-Ohlin-Vanek) model and the TZ (Trefler and Zhu) model. The HOV model assumes that the composition effect alone accounts for the observed international patterns of emission content of trade. This model is most appropriate for when country-specific differences in production techniques are minimal. It further assumes that markets are equally competitive, and that there are no barriers to trade. The TZ model is most appropriate when production techniques vary across countries. In the TZ model, the factor content of imports and exports are measured using the producer countries' techniques.
Standard test procedures were used to evaluate the predictive performance of the HOV and TZ models. First, a sign test to measure the number of coincidences between the measured emissions content of trade and the predicted emissions content of trade gave an indication of which was the more accurate model. A 100 percent coincidence would show that the measurements were perfectly consistent with the predictions. Then, a slope test was used to measure the emissions content of trade on a linear regression of the predicted emissions content without a y-intercept. Finally, a variance ratio was computed for each factor by dividing the variance of measured values by the variance of predicted values. A ratio of 1:1 suggests perfect correspondence.
The HOV model's performance was mostly inconsistent with observation. Its sign fit was 56 percent, which is reasonable, but its slope coefficient was 0.048, which is very low and indicates a low level of correspondence between predicted and measured outcomes. The slope coefficient is determined by taking the measured emissions content of trade on the y-axis, and the predicted emissions content of trade on the x-axis, and drawing a line through the data points for each country. For example, if the model predicted that the United States had an emissions output of 200,000 kt, and the actual measured output was 500,000 kt, a data point would be recorded at (200000, 500000). The variance ratio for the HOV model was 0.070, also low.
The TZ model fared much better at accurately predicting the emissions content of trade. Its sign fit was 87 percent, which indicates a high rate of agreement. The slope of its regression line is 10 times higher than that of the HOV model (0.447), also a good indication of its high predictive power. Its trade variance ratio is also high, at 0.451.
The TZ model is able to accurately predict which countries are the leading exporters of emissions. China and the United States were found to be the leading exporters of emissions, while France, Germany, Italy, Japan, and the United Kingdom were found to be importers of emissions. Several developing countries in Eastern Europe, including the Czech Republic, Poland, and the Slovak Republic were also found to be net exporters of emissions. These results seem to confirm the hypothesis that the net exports of developing countries are more emissions-intensive than those of developed countries.
The HOV model, which allows only for the composition effect and holds the technique effect constant, was found to demonstrate only a statistically insignificant relationship to measured outcomes. On the other hand, the TZ model, which allows the technique effect to vary, was found to demonstrate a highly statistically significant correlation between predicted and measured values. The difference between the two models is due to differences in how they view the importance of production techniques. This evidence suggests that emissions intensities differ across countries because of variations in production techniques.
One of the weaknesses of this study was that it was subject to a type of error known as aggregation bias. Aggregation bias is an effect that happens when the incorrect assumption is made that what is true of the group is true of the individual. For example, it may be that the machinery sector of industry includes a emissions-intensive “heavy” sector and a less emissions-intensive “light” sector. If the companies that produce lower emissions are located within more developed countries, then the methods that were used for this analysis would detect cross-country differences in technique in the machinery sector, when in fact the differences would be attributable to the composition effect. Aggregation bias, therefore, can lead to an incorrect interpretation of results.
However, one of the strengths of the study was that the data sets were constructed so as to minimize the level of aggregation. By using 41 industrial sectors in the data set instead of a lower number, the sectors were less aggregated. This allowed for a more accurate comparison of emissions data across a large number of developed and developing countries.

The overall conclusions of this study were that emissions intensities differ across countries because of of differences in production techniques. The results negate the hypothesis that international differences in emissions intensity play a substantial role in determining patterns of trade.

Conclusion (Future Avenues of Research)

This study employed tools of empirical international trade to show that current emissions costs differ because of international differences in production techniques. It has shown that there is a negative relationship between a country's emissions intensity and its level of economic development. The paper addresses questions of which countries' production and trade are most emissions-intensive and why. The results of this study are of particular interest in light of global efforts to reduce climate change, and will have an influence on international policy. Future studies may examine how the imbalance in national CO2 emissions intensity between rich and poor nations contributes to global warming trends.

Works Cited

Douglas, S. and Nishioka, S. “International differences in emissions intensity and emissions content of global trade.” Journal of Development Economics 99. (2012): 415-427. Web.