The Ecological Benefits of Smart Growth: Where is the Science?

0:00I am an economist by training, a strong free market economist, so I certainly share a lot in common with the von Mises Institute and the folks here who do such a good job. But I migrated several years ago over into the School of Forestry and Wildlife Sciences here at Auburn. This has allowed me to, it's a very sort of refreshing sort of thing to do in the middle of a career, is to take what you've always thought of as a model that applies pretty broadly into a completely new area where you haven't thought much about before. And it turns out there's a lot of applications for free market economics and public choice theory in the whole world of natural resources. of Sources.

0:45And one of the problems with the natural resource community is that there tends to be a mentality that focuses on command and control as a way of getting people to do what they know in their heart of hearts is the correct thing to do with the environment. And it's that sort of point in time that forms the point of departure for my discussion here. One of the probably biggest command and control-oriented programs or sets of principles that exist in the natural resources community, policy community, goes under a broad umbrella called smart growth, and that smart growth is essentially a set of principles by which the advocates of these policies want us to live our lives, and they're perfectly prepared to use the and the Apparatus of the Government, whether that comes in the form of tax incentives or various subsidies or outright regulation that tells you what you can and cannot do with your property and how you must live.

1:53The point is that it is a very, very widespread sort of movement. And one of the particular principles of this movement is that they like people to live in big cities. They publicly advocate high density human habitation. And the theory there is that if you cram everybody into the cities, then this leaves all the rest of the area pristine for all the critters. And so, at least in principle, the thought process is that there are ecological benefits that result from the pattern in which we conduct ourselves, the human populations.

2:46And here is wording from this principle within Smart Growth that tells you what this link is between the spatial organization of human populations and the sort of ecological consequences they expect. You build the buildings vertically rather than horizontally and you get what somebody refers to as efficient use of land and resources. and Resources, and I sort of wonder what they mean by efficient, but here's what it literally says, it reduces the footprint of new construction, preserving green space to absorb and filter rainwater, reduce flooding and storm water drainage needs, and so forth. And I've emphasized that because in effect what the argument here is, is that somewhere, somehow, by doing all and all this stuff, the world's a better place, all right?

3:43That just jumps out at me and I think that's got to be testable, okay? So I'm going to test that sucker. Just to give you a little bit of reference about how powerful this smart growth set of philosophies and advocates has become, I did a Google search on myself this morning, 149,000 references, Google search on Ludwig von Mises, Google search on smart growth, 12 million references. Google search on free market economics, 12.3 million. So the smart growth advocates, there's a lot of attention being paid to them. So what I'm interested in pursuing then is the question, what's wrong with a picture?

4:35The Smart Growth Advocates are perfectly happy to use the public sector to guide, as it were, how we live our lives. And their willingness to do that, it seems to me, is founded on two very, very strong assumptions. The first is that in fact there is this link between how we live our lives in a spatial context and ecological consequences. And then secondly, if there is some link, let's assume that number one is correct, If there is some link that command and control policies or subsidies or the sort of standard instruments of government intervention in the economy are, you know, sort of the efficient way of doing that, I'm not going to deal with number two because I don't need to deal with number two if I can show that number one has no basis to it.

5:25So that's what I want to explore. I want to explore that link between how we organize ourselves spatially and some indicator of ecological wellness as it were. So let's explore this link. In forestry, I can tell you exactly what the argument is because my colleagues in the School of Forestry and Wildlife Sciences make this argument. If you plant trees very densely in what we call managed forests, so they've got all that spacing worked out to what the optimal spacing for planting is and you fertilize them and you grow them thick and you grow them fast and you cut them, that in effect what that permits you to do is with plantation forestry you cut the same amount of timber volume on a smaller amount of land which allows sort of a lot of other land to be freed up for natural purposes.

6:22And if you're not doing that, then you're cutting the same volume of timber off of a much larger area of land. So there's that issue of density. Grow them thick, grow them in a small place, and you can get what you need. And that somehow there's these ecological benefits compared to letting the trees grow naturally and cut out of that natural area. And here's, it seems to me, the suggestion that's being made there is that in the aggregate, and I've got aggregate emphasized here, is that in the aggregate, this sort of managed population of trees has ecological benefits overall as compared to just growing them naturally and cutting them naturally. So now, let's look at the smart growth analogy. The smart growth analogy to what I see in forestry is almost identical.

7:19So here's a passage from the smartgrowth.org website. As we build, we replace our natural landscape, forests, wetlands, grasslands with streets, parking lots, rooftops and other impervious surfaces. The effect of this conversion is that stormwater, runoff which prior to development is filtered and captured by natural landscape, is trapped above impervious surfaces and runs off into streams, lakes and estuaries picking up pollutants along the way. Runoff can be reduced through clustering of development. There's the analogy to intensively managed forests, right? and thereby leaving larger open spaces and buffers although compact development generates higher runoff and pollutant loads within the development total runoff and pollutant loads are offset by reductions in surrounding undeveloped areas that is a the perfect analogy and so implicitly here the arguments being made that when you look at the world overall there are better patterns of Development and Worst Patterns of Development.

8:24And the better pattern of development is this one, where you put all those people, cram them into big cities, and you leave the countryside untrammeled. I can tell you who a big proponent of this is, probably the most eminent biologist in the world today, Ed Wilson at Harvard, came down here a couple of years ago, and he told A couple of things that seems to me are fairly self-evident, and you'll see that they're borne out by the data that I'll talk through here in a minute. First of all, the ecological success of any species creates opportunities for some species and imposes costs on others.

9:15In the words of economics, a successful species crowds out. Species that are competitive with that species. Okay, so surely in terms of things that we're very familiar with, population density, the more people there are, yeah, it's going to create problems for other species. Okay, the more black bears there are, the more deer. How many of you have azaleas that are suffering at the moment under the onslaught from a bunch of hungry deer? The point is a pretty simple one. Yeah, we know that human population matters and economics gives us a convenient way of thinking about that through the crowding out effect. And population density is probably the most widely used measure of human presence in a particular location.

10:02and you'll see that indeed population density consistently matters in models we're looking at some measure of ecological disturbance or distress okay now here it seems to me is the probably the most critical line in my presentation is this next one population density by itself may not be a consistent metric of human presence of the impact of human presence okay and by By way of illustrating this, let me refer to a researcher at Michigan State named Jack Lu, he is an ecologist, and he makes the following argument, that if you look at two equivalent populations in two equivalent areas, so you got the same population densities, then would you expect those two identical population densities to draw the same resources from the natural resource base?

10:59The answer is not necessarily, and his argument is that in one case, if that population is organized into more households, smaller number per household, but more households, there's a larger draw on the natural resource base than if that population is organized into fewer numbers of households of a larger average size. okay because you end up with more houses in this one here more appliances more of everything which which implies a bigger draw on the resource base okay so he's the first one that I know of and his his Jack Lew's paper just came out in Nature about two years ago a little over two years ago he's the first one that I know of who has suggested that Not all population densities are created equal, okay?

11:58And indeed, I think there's a very simple way of thinking about the impact of population. So if you consider two states are identical in all respects except the distribution of their respective populations, would you expect the ecological footprint of man to be the same in both cases? Our knee-jerk reaction is to say, probably not. So, pick this example. You've got two states. They've both got equivalent populations of one million each. In one state, the urban population is 75% and in another state, the urban population is only 10%. Would you expect, they've got the same populations, same areas, same population densities. One is just more concentrated than the other. Would you expect the ecological consequences to be the same?

12:47And as I say, you need your reactions to say, well, of course not. Well, let's take that one step further then. Consider two states with identical urban populations. But in all other respects, they're identical. Would you expect this ecological footprint associated with urbanization to be identical in the two states? So here we've got those same two states, a million population, which means that for the state as a whole, they got the same population density, right? And in this case, the urban population is distributed entirely in one city of a million, and in that state, it's distributed in 50 cities, each with a population of 20,000.

13:34You expect the ecological effects to be the same of that human population? And my guess is most of us think about that and they say, well, there's got to be differences. And yet, you ought to just put the brake on your natural tendency to say, yeah, go get them. We've got to do something to save the environment. Those million people have got to eat. And it doesn't matter whether they're located in a city or in one city or located in 50 cities. They've still got to eat. They've got to get drinking water. They're probably all moving around. They're probably all wearing clothes, maybe. And so they're drawing on a resource base. They're both in principle drawing on at least some resource base that they share no matter where they're living.

14:29So your knee-jerk reaction gets tempered pretty quickly by the realization that they're drawing Coming from a resource base that probably is pretty similar no matter how the human population is configured.

14:44So here's what I set out to do, is I want to investigate whether the ugly facts actually support what the smart growth crowd believes is a very elegant theory, which is that we ought to cram all those people into a big city and that there are ecological benefits to doing so. And I'm going to do this in the following manner. I do not want to get into a shouting match with a bunch of ecologists about specific ecological effects of cities. So in other words, I'm willing to concede, yeah, big cities may have runoff associated with all that impervious surface. OK, I don't have a problem with that. To me, the issue is, are there broad aggregate indicators of Environmental Well Being that can tell us whether any little bitty effects, specific effects that may be occurring, end up somehow really mattering in terms of the environment.

15:42And so I'm going to use a very aggregate indicator which is the presence or the intensity of ecologically imperiled species because they'll tell you whether or not there's something And what I'm going to do then is I've got data, this will be more apparent as we go on, but I'm going to essentially run multivariate regression analysis of a following general relationship, the species fragility in an area, as dependent on a number of ecological factors and a number of human induced factors. Among the ecological factors that I might potentially look at, if I'm looking at the number of endangered species, then you surely want to control for the total number of species.

16:39You almost certainly want to control for the number of endemic species. And those of you who do not know what endemics are, endemics are species that are unique, only found in a particular area and nowhere else. It turns out there's a very strong relationship between the number or percent of endemic species in a country or in a state or in a location and the number that are ecologically imperiled because the definition of endemic means that if they're found here and nowhere else it means they've carved out a little niche for themselves which means there probably aren't a lot of them, which automatically sort of clues you in maybe they're sort of ecologically fragile. And then you can look at a lot of indicators about those ecological niches, like annual rainfall, miles of coastline, average height above sea level.

17:34There are a lot of geophysical indicators that might clue you in to the extent of environmental diversity within a location. And then in terms of human sorts of variables, population density would be at the top of your list. We crowd out other species, no doubt. The urban population, the degree of urbanization. Here's the one I'll be particularly interested in, which is the distribution of that urban population. And then things like per capita income, education, maybe the more highly educated a population you have, The more interest there is in being environmentally friendly. I don't know.

18:23Here's some issues that I had to deal with. First issue is, what am I going to use as a measure of species fragility? Second is, what am I going to use as a measurement of the distribution of urbanization? and the third is, what level am I going to be permitted to conduct the analysis at? This will become more apparent in a little while. The ecologists all want you to conduct an analysis at a very, very micro scale, a very small scale, because they're convinced if you pour concrete here and you're covering up some little flower, that by definition, that means the world's a worse place. And they don't like to think in terms of larger scales. And so the scale issue, it turns out, is a very, very big issue in ecology.

19:14The guy I did the first part of this analysis with, Roger Brown, when he defended his dissertation, this was a chapter of it. And all of my colleagues just absolutely came unglued. We had a shouting match in his defense. And the dean of forestry was just shocked, because nothing like this had ever happened before. But it all boiled down to a scale issue. And so we've had to be pretty sensitive to this. So the metric for species fragility. I can tell you, because I've done the work, that if you look at a conventional measure of ecological fragility of species like listings under the Endangered Species Act, that those listings are not completely unbiased.

20:02They are subject to political forces. There have been several papers that demonstrate this. And so we rely on a listing put out by a company called NatureServe. NatureServe is a derivative organization of the Nature Conservancy. They split away from the Nature Conservancy several years ago to concentrate on purely monitoring and providing statistical information about endangered species. and they have their own metrics for endangerment and are, I think, are widely regarded as these essentially unbiased source of information about what they call species that are at risk of extinction.

20:51The distribution of urbanization. Economics has a lot to offer actually in this discussion because we've got a very convenient measure of concentration in economics. It's called the Gini coefficient and the Gini coefficient ranges on a scale from 0 to 1. This is formally how it's calculated. I'm not going to march through the calculation. I'll show you a nice diagram that will depict it in a minute. But the point is that on a scale from 0 to 1, with 1 being highly concentrated and 0 being very diffuse, there's a very nice measure which tells you whether the population is spread out evenly over an area, uniformly, or whether the population is highly concentrated into particular locations.

21:39So now I hope it becomes apparent what I'm going to do. I'm going to look at the population and I'm going to apply that measure to it and I can tell for a state, all across a state or all across any unit I want to look at, whether the population of humans is concentrated or whether it's diffused. And I can tell whether that has an impact on species fragility. So that's where I'm headed. Here's what the Gini coefficient looks like, diagrammatically. This diagonal line here, in effect, represents purely uniform distribution of humans across whatever area they inhabit. So 20% of the people would be spread over 20% of the area, 40% of the people would be spread over 40% of the area, and so forth.

22:30So this would represent purely uniform distribution of people across the area you're talking about. Deviations from that, concentration would follow this curved line. So what this would really mean here is if this line was really spread all the way over here, then you'd end up with 95% of the people in 20% of the area. And so the population would be very skewed and very concentrated. And in the limit, if you went all the way along the edge here, and you had 100% of your people in, you know, 0.01% of the land area, then you'd have an extremely concentrated population. So this Gini coefficient is a sort of a convenient way that economists have to measure how concentrated the population, human population is in any geographic area you choose to consider.

23:33And the question is how are we going to measure human presence as it were? Well, we can measure population directly from the census, We can measure the number of households from the census, which sort of gets more towards incorporating Jack Lew's concept of the humankind's impact on the environment. We can look at roads. We can look at the distribution of nighttime lights. A lot of mapping satellite imagery has now become available, which would allow you, in effect, to take a given area, look at all the pixels, Add up all the pixels that are in a certain light category and divide by the total number of pixels in the area and you get a measure of nighttime light concentration.

24:18And impervious surface area, we're actually going to look at the first four of those. So I wasn't content just to look at the number or the distribution of population. I wanted to deliberately look at alternative measures to see whether I got the same results. Here's the scale, part of the scale issue is a data issue. Ideally the ecologists want us to keep looking more and more and more and more closely, but the problem is a data problem. At the state level in the U.S. I can get data on all of the variables I need. But the particular variable that you have to have in a model is the number of endemic species to an area. And that data only exists at the state level.

25:06It's slowly becoming available at the county level, but we don't have a lot of them yet. At the sub-county level, the data problems become overwhelming. It's possible, I would guess, within five years to do a county-level analysis. You can also do this on a transnational level, which means you can look at countries as your

26:02or at-risk species in each state against all these explanatory variables why do we exclude Hawaii because islands are just different everything about an island is different the in particular the critters can't escape if some exotic being comes in from the outside that the species aren't prepared to deal with Ecologically speaking, they can't escape. And so the effects are magnified. This is very, very well known in the ecological literature. And so we have omitted Hawaii from our analysis of the states. I'll show more of this to you presently. So what do we find? Across the 49 states, here's our means.

26:50I particularly want to point out a couple of things. The Gini coefficients. Genie coefficients are very, very consistent for three of the four measures of human presence. The mean genie coefficient is at about 0.8 or so. With roads, it's different just because roads don't necessarily always follow the human population. Sometimes they just go off because the roads are all decided by politicians' pork barreling. But the other thing I want to point out about these genie coefficients is the minimum. The minima Gini coefficient on those three indicators is up close to between 0.55 and 0.6. So even the minimum degree of concentration is fairly concentrated.

27:45And here's our regression model. We're going to, our dependent variables is the percent of endangered species in a state. We're going to control for the percent of endemic species in that state, the area of the state, the population. Once you've fixed area, then when you enter population, in effect, you've entered population density. Here's our measure of the concentration of that population in the state. And then we have a set of regional dummies, dummy variables. And then I will replace population density, I will replace systematically with the other measures of human presence. So I'll look at households, I'll look at roads, and I'll look at nighttime lights.

28:37And here is a very, very busy figure. So I'll explain it fairly simply. Firstly, first of all, the percent of endemic species is by far, in a way, the biggest explanatory variable with respect to the percent of ecologically fragile species in the state. Very consistently, you will get an extremely high, large coefficient and extremely significant statistical explanatory power. Look at these four variables. These are the indicators of population density or human presence. They all vary systematically, are highly significant indicators. No question about it.

29:22Humans crowd out other species or they stress other species. That's a very consistent finding with the literature. I'm not disputing that at all. Here, it seems to me, is what the issue is under dispute. The next four measures are measures of the spatial arrangement of those humans. And the smart growth folks claim that the spatial organization matters. Right? They say the more you put those folks into the big cities, the better off we ought to be ecologically. I can't find it. I can't find it and I don't care how you define the human population. I cannot get a statistically significant coefficient on a variable looking at the concentration of the human population.

30:15We note some regional effects that are pretty significant and I don't have yet a ready explanation for the regional effects. Let me turn now to our international analysis, which I'm still in the process of marching through with my doctoral student, Ron Pandit. We've got data from about 124 countries around the world where we can get data on endemic species, total species, threatened and endangered species, and all the other variables that we are looking at. What would you expect? Would you expect human populations to be more spread out or less spread out in other countries than the United States?

31:02What do you think would happen to the Gini coefficient around the world? I was expecting it to go down. I was expecting there to be some countries where all the people were still out in the bush living in the Stone Age. Okay, so they're all spread out. Well, it doesn't happen. That's why I focused on this a little while ago. Look at the minimum Gini coefficient for population for all the countries around the world. It's almost as high on average for all the countries of the world as it was just for the United States. So it turns out that concentration of human populations is very high all around the world. in the world. And when I first started this, I had people very specifically say your analysis of the United States is no good because the United States is just not the same as it is everywhere else. Well, sorry for those ugly facts rearing their head again. It turns out the United States is a lot like the rest of the world. The rest of the world, human populations

32:01Cluster Together We have actually added one additional variable into the next set of models we ran internationally. We knew for each country the percentage of protected areas of the land base that had been set aside in protected areas for wildlife. And so we were able to enter that as an additional explanatory variable. It turns out that it never matters. Set that land aside and the poachers still get on there and kill the critters. So the percent of protected area in a country just doesn't seem to influence the fraction of endangered species in a country.

32:51But once again for all around the world when we look at the percent of endemic species, Exceptionally Highly Significant as an explanatory variable for the percent of endangered species.

33:08Population density matters, just the way you expect it to. More people, more endangerment. This is the Gini coefficient of population. Look at this. Gini coefficient matters. I'll come back to that. This is per capita GDP. This is like, this is a variable that we enter in to control for what's known as the environmental Kuznets curve. Anybody know what the environmental Kuznets curve is? Environmental Kuznets curve in a nutshell is a statistical relationship between the economic well-being of people in a country or a state and indicators of environmental stress.

33:53So the logic runs as follows. When you're absolutely dirt poor and you're struggling to survive, you eat those critters because you're just struggling to survive. 25, as you develop economically, and you can feed yourself and you've got clothing and you've got houses, you start to appreciate the finer things in life, like saving all those critters because you like looking at them rather than eating them. So there ends up being this relationship between income or economic well-being. As you get more and more, you're willing to trash your environment to gain a little bit economically, over some range. And then as you continue to develop economically, your trashing of the environment diminishes, where you take better care of the environment. So there's what's known as an inverted U-shaped relationship between economic well-being and species fragility.

34:43And we have got a number of different formulations where we look at this relationship and we don't find much evidence of it. Now, back to this one. This is my variable of interest. Look what happens when we add a dummy variable for whether or not the country was an island. That's this set of results right here. Island countries matter. They got lots of fragile species. And it matters so much that it completely invalidates any notion that the distribution of people matters. The spatial distribution of people matters. And for that matter, it drives out of significance even population density.

35:35Okay, so you better, this is clearly a variable that matters in a model, and you better not just exclude it from your model because you think it doesn't matter and that island countries are somehow, you know, not worth including in your model. And here's why. Look at the differences between island countries and non-island countries. Population density in island countries is six times higher than population density in non-island countries. The percent of endemic species that live there and nowhere else is triple. The percent of ecologically fragile species is also more than triple, almost quadruple. quadruple. So these characteristics of island countries are just radically different than characteristics of countries that aren't island countries. So you've got to have them in there and when you put them in there, you just, you don't find any evidence that the spatial and the International Organization of Human Matters.

36:45So where do we go from here with our analysis? Well, first of all, we could address the scale issue. So we could come down to finer levels of analysis and say, okay, we're pretty sure that across the United States as a whole, How you organize people spatially doesn't seem to have aggregate environmental effects, but in smaller locations the spatial organization of people may indeed matter. So we'd like to be able to bring that analysis down to a finer scale and see whether the conclusions are generalizable from a large scale down into a finer scale. We still remain for us to be able to get the data on that. That's coming, but it's not available yet.

37:35Here's what I'd really like to do. I'd really like to not work with established political jurisdictions. Countries are artificial constructs, right? Somebody just drew a line and said, that's my country. Take it if you can. But animals don't know that and species don't know that and I think really the best test of what we're looking at would be to take a big map of the world and throw darts all around the world as random points and take a radius around each of those points and identify endangered species and then identify the spatial distribution of the humans and see whether in a purely random analysis, purely random, you could link the spatial distribution and artificial distribution of humans with some measure of species fragility.

38:35Maybe, maybe this can be done. I can do this in the U.S. because NatureServe can actually get me the data. If I give them a latitude and a longitude point, they can draw a radius around it and give me the information on endangered species. But I can't do this for an international analysis. Lastly, I'd like to disentangle my dependent variable. Remember, my dependent variable is just the number of imperiled species in a country, or the number of imperiled species in a state. Well, the problem with that is that not all imperiled species are created equal. Okay, and let me show you what I mean by that. It turns out that the species that are tracked, they track them by major taxa.

39:25So they've got the amphibians, they've got birds, they've got reptiles, they've got plants, they've got fishes, they've got invertebrates. But look at this, look at the number, look at the sample size. Vascular plants in our sample dominate, there's almost 6,000 of them. Look at 139, 214, 65. The average country in terms of its flora and fauna, in terms of its fauna, the numbers are much lower than for the vascular plants. So I think one argument that you can make is to say, LeBan, your results are dominated by the vascular plants. And people, the impact of people and how humans are configured spatially may not have an impact because, well, the plants just don't seem to be very sensitive to that, but if you looked at the mammals, for example, or you looked at the amphibians, that there may be taxa-specific effects that indeed you could identify if you were just focused on those rather than on aggregating all of these different taxa of species together in an analysis. And so this is surely the next step that we will

40:39that we'll actually be able to take in Ron Pandit's dissertation work. We'll split that out and see how sensitive the individual tax results are or whether they're consistent with the aggregate results that we have. But the bottom line is that, at least at the moment, in an analysis that I'm perfectly happy to admit is a relatively aggregated level, We don't find any scientific basis for the very strong statements that are made in the smart growth literature to the effect that there are these ecological benefits to cramming people into very densely populated cities. Surely there are localized effects to doing that, but in the aggregate, when we look at the composition of species fragility, We don't find any evidence that the spatial organization of humans matters.

41:37So I've been pretty blunt in my criticism of the foundation of the smart growth prescriptions for human populations. And incidentally, Auburn, the city of Auburn, is a smart growth city, if you did not know that. So, it seems to me there's a lack of science that underpins what has become a very, very powerful movement. Thank you very much.