Now I'm not advocating for these measures, I'm not suggesting these measures are great, but this is an example of how sociologists have attempted to try to quantify these sorts of processes and one of the take-home points that I want you to get from this is, this is not easy to do and we're still trying to figure out how to do this better and perhaps these are not, you know, getting back to some things I said yesterday, there're, perhaps there're some pretty strong limitations in using quantitative methods to try to study something like globalization, you know, there are some limitations to this, but at the same time, you know, we think that they can provide some important contributions. So a little bit about longitudinal methods, I'm not going to do the dance again like I did yesterday to try to sell you on longitudinal methods, but why longitudinal methods when you're studying these sorts of interrelationships, well development is it's a process, right, it's not static; these forms of globalization again are not static and environmental change really isn't static either and so ideally in order to study these interrelationships we want to have a repeated observations on the same cases for particular units of analyses and so yeah if you needed to be convinced as to why we might want to employ longitudinal methods well here you go, but there's also some particular methodological reasons that.
I think are really important in terms of allowing us to do more rigorous hypothesis testing since we're not using experimental methods, you know, we're analyzing secondary data we're using inferential statistics and so by using longitudinal data we're allow, we're able to better account for really omitted variable bias, this notion of heterogeneity bias, things that either we do, we might know that we're missing that we don't have measures for, or things that we don't know about yet, you know, this is this process of using different kinds of fixed effects in these models and so in simple terms when we're doing something like a two-way fixed effects model in a cross-national study where we have repeated observations on many nations, this means that in our fancy regression model,
We have a dummy variable for every country as well as a dummy variable for every time point, okay, to account for these things that are unique to each let's say country that don't vary through time as well as unobservable factors that are unique to each time point that don't vary across cases, okay. Now arguably that is going to explain a notable amount of variation in your dependent variable which is going to lead to perhaps more conservative estimates of the effects of your independent variables on your dependent variable. There's a lot of arguments within the methodological literature about all of this, but I think that this is an important thing to keep in mind from my point of view, you know, early on in my career that goes back about 15 years, I started out doing all cross-sectional stuff and then it got to a point where I could do more longitudinal research and from my point of view.
I thought well this allows me to be more conservative in my hypothesis testing, asking similar sorts of questions, but being able to do the research more rigorously and more conservatively cause the last thing I want to do is commit a type one error, yeah, falsely rejecting the null hypothesis that's the last thing I want to do and I think that using these methods helps me to lessen the likelihood of committing a type one error, again it's debatable I realize, but that's one of the, my selling points on why I think these methods are important. Okay, so in this literature though on environment development globalization at the comparative international level, these are some of the common dependent variables in this emerging area of literature and anthropogenic greenhouse gas emissions especially anthropogenic carbon emissions from the man, from the burning of fossil fuels and manufacturer of cement and there's also a difference between production-based emissions versus consumption-based emissions too as well, but a lot of us are analyzing CO2 data because they tend to be more reliable for comparisons between nations as well as comparisons through time,
We're also looking at other kinds of composite indicators like the ecological footprint, there's a tradition in the environmental social sciences to use ecological footprint for hypothesis testing, there's also a huge literature debating the methodology of the ecological footprint and I think that that literature on the methodology suggests that we should be really, really cautious when we're using ecological footprint for hypothesis testing because it's this big kitchen sink sustainability index that lumps a ton of stuff together and if you haven't ever looked at it before spend some time reading all of the fine print it's really impressive, but it lumps together a lot of stuff and I'm personally not all that comfortable anymore in using these very lumped together measures,
I'd rather use these more direct measures like CO2 or other air pollutants, industrial water pollution has been studied, the, and specifically industrial organic water pollution by, measure by biochemical oxygen demand, some folks have done some research on synthetic pesticide and fertilizer use and there's also been a tradition of work within sociology that Tom has contributed to greatly on deforestation and this one's tricky to do longitudinal research cross-nationally on deforestation because from a measurement perspective the ways in which we measure forest cover at this level of aggregation changes through time and also varies by nation, so it's really tricky and challenging and really difficult to do longitudinal research on deforestation, but we try, some of us have tried to do it. A little bit more about measurement though, actually about some of the CO2 data,
I usually spend a whole lecture with this slide up when I teach undergraduate courses in environmental studies to talk about a lot of things, but one of them though is the importance in how you can operationalize and measure an outcome in let's see three different ways it's commonly done to tell us very different things and so these are three figures of looking at national level anthropogenic emissions, total emissions annually per capita emissions and emissions per unit of GDP and this is just time series data, annual time series data for foreign nations - Brazil, China, India, and the U.S. I think the figures speak for themselves that the time trends look different for the three, for the four nations across these three measures.
There are important theoretical distinctions it, with, from a sociological perspective in terms of why you might want to use one or another of these as a dependent variable. A lot of inequality, international inequality scholars within sociology are very interested in per capita measures as a dependent variable, but from, really from a climate change mitigation perspective this is the most important one, if you think about it total emissions, or arguably cumulative emissions is even more important than this and so I'm going to show you some examples of different studies that use these different outcomes and I know that the reading that I'd suggested that you read and I'm sorry it's really long, if you tried to read the whole thing, but in that particular study we employ all three of these as dependent variables to try to do a very sort of objective systematic analysis of testing competing hypotheses between those theories that you heard about yesterday - ecological modernization theory and treadmill production theory. And then again, if you use a standard measure of development I think this is something that most of us are familiar with and these are adjusted for inflation, these GDP per capita data.
Some would argue that this is illustrative, this growing gap between the Global North and the Global South, but the scale of this also hides the fact that the GDP per capita of some of these nations are going up as well. I'm a little biased cause I'm a sociologist, so I like the per capita measures, but sort of peering in more closely on per capita CO2 and this is just looking at something that you're probably familiar with, but to further, to kind of underscore the point that if we look at average per capita CO2 let's say for the kind of the Global North versus the Global South through time,
Environment, Development, and Globalization - Part 01
Environment, Development, and Globalization - Part 02
Environment, Development, and Globalization - Part 03
Environment, Development, and Globalization - Part 04
0 Comments