Alma,

First off, good motivational video, even if it does use overtly populist appeals at times which ignore the tough questions. Generally speaking, even those tools have value in making people aware of the full costs of their actions.

Second, sorry to take so long to reply. I began writing a response and then realized that it is a bit like Alice falling down the rabbit hole. Unfortunately, I have not studied environmental economics at all so I cannot wave my hand at some general conclusions from that field. However, concepts from the fields of development econ, growth econ, and general macro theory can be used to help understand the human framework in which a solution has to be formulated.

The hardest part of the issue is its complexity. There are so many factors that go into it; talking about it qualitatively inevitably faces significant diminishing returns. To put together my thoughts, I have been putting together a macro model with the salient structural elements so that I can be rigorous about my understanding of the problem. Not only is this a perfect application of a good deal of the theory I've been working with lately, but a proper DSGE growth model with well formulated resource constraints (all growth models that I am aware of have abstracted away from such constraints) could potentially be a paper topic. (edit: there is some literature out there)

I'm going to start by discussing the permanent income hypothesis qualitatively (although I will use a few equations), and I will report back when I feel more confident with the ways that the various institutional/social/physical frictions interact.

Understanding the way that people form their intertemporal preferences is a huge part of the problem. Taking a standard instantaneous utility function, and then assuming that societies seek to maximize their total future-discounted expected utility, the first order maximization condition gives what is called the consumption euler equation:

*U'(c_t)=beta*E_t[(1+r)*U'(c_t+1)]*where U'(c) is marginal utility, beta is the discount factor (0<beta<1), E_t is the expectations operator conditioned on time t information, and r is the expected real interest rate: 1 + r = (1 + nominal interest rate)/(1 + expected inflation).

The consumption euler equation expresses the hypothesis of rational expectations, people make their decisions about the future using all available current period information. If rational expectations holds, then to the extent that people are informed and we have enough information consumption should be smoothed over time. That is, consumption should approximate a stecady state. This idea (credited to Milton Friedman) is called the permanent income hypothesis.

What the video you posted in fact

**does** is to help society converge our consumption stream to a new (more correct, fully informed) sustainable steady state. The problem is that information is often costly to obtain (and there may be more social norm and institutional factors at play too retarding the spread of information). The way I see it is, people do hold close to rational expectations (approximately, artificial neural networks are a way to simulate approximate rational expectations). With the proper information, consumption should trend around the steady state growth path as postulated by the permanent income hypothesis.

Unfortunately, consumption in itself is an abstract phenomenon, as is utility. However, one way to try to see if this holds true is in the dual variables -- the price variables (the interest rate). Suppose that at some time t we are at the steady state. Then, the euler equation reduces to

*U'(c)=beta*r*U'(c)* meaning that

*1 + r=1/beta*values of beta between .96 and .99 (~4.1% and ~1.2% respectively) are often used in calibrations of models and tends to fit data well (note, not in short run situations of hyperbolic preferences -- a topic which potential relevance, but I'll ignore it for now). If the interest rate is higher than

*1/beta*, what does this mean? Then,

*U'(c_t)>E_t[U'(c_t+1)]* which implies (due to diminishing returns to consumption, ie concavity) that

*c_t<E_t c_t+1*. people expect to consume more in the future. If the interest rate is lower, then just the opposite holds and people expect to consume less.

Look at the following data on interest on US bonds (blue) and interest on inflation indexed bonds (red). The spread is a measure of expected inflation, and so in theory (to the extent that lending to the federal gov't is risk free) the red line measures the expected real interest rate:

First notice that at the beginning of the data series, this measure suggest that the real interest rate was at about 4%. However, with the onset of the recent slowdown, the real interest rate has fallen. Arguably, people are realizing that they can't afford their current mode of consumption and are expecting to cut back.

Now, I'm not arguing that this necessarily directly has to do with living on a finite planet. It does, however, suggest that we have tools to indirectly observe people's preferences and expectations. Interest rates are a potential insight into understanding how people will behave with respect to their future consumption.

Note that this doesn't imply that permanently low interest rates will CREATE a better future. In fact, if one used policy to put downward pressure on interest rates you would in fact encourage consumption (the opportunity cost of consumption today is the interest rate). BUT, in the transition from a over consuming steady state growth path to sustainable level of consumption, you would expect to see a fall in interest rates as people updated their expectations to information about the future and then a rise again in interest rates as we approach the new steady state growth path.

Ok. I need to get some work done and I've probably bored the shit out of everyone. There is plenty more to be said, and I'm going to create a more formal model of how a complex macro economy with imperfect product markets would react over time to running up against its resource constraint.