Felicifia Update

I've been too busy recently scrambling to make grad school application deadlines to post much. I hope to have more time to write in the spring. It's certainly not for lack of things to say! The applications are for opportunities to study some form of quality of life maximization.

Meanwhile, we've registered felicifia.com, where we hope to roll out a completely revamped site on a platform more sophisticated than Blogger, possibly Scoop. Blogger has been great: free, fast, and functional, but I'm looking for new functionality.

Also, congradulations to Peter Singer on his cover article in the recent NY Times magazine: What Should a Billionaire Give – and What Should You?.

Happy holidays, or lack of holidays, whichever it is for you.

Utility Taxonomy

There is a strong consensus that members of certain other species have non-zero utility. However, I have not seen any attempts, however crude and approximate, to produce a "utility taxonomy", i.e. a classification of species by the relative amounts of utility their members are either capable of experiencing or are likely to experience during a normal state of being.

Having a utility taxonomy is essential to our efforts to understand how to maximize utility on this planet. We need to know how substantial the interests of livestock animals are as we attempt to optimize our agricultural practices. We need to know how much the interests of "wildlife" animals are worth as we consider actions that would disrupt their ecosystems. It might even be worth knowing whether we should smash and kill a mosquito whose presence we found mildly annoying.

In the absence of an existing utility taxonomy, let's make our own. I'll get things started, and we can refine it in the comment thread. Post anything you think may be helpful. As always, if you prefer to remain anonymous or otherwise don't want to use the comment thread, then email me at sethbaum [at] gmail.com.

One final thought before beginning: Humanity, ourselves included, really may not have a reasonable solution to this problem at present. Our work may prove completely off the mark. Thus, we should proceed with caution. However, we should proceed, as we've got to start somewhere with this. Personally, it feels strange trying to develop a system based on little more than crude intuition. So be it, I suppose.

---------

Some initial estimates:

I'll make the seemingly safe assumption that organisms that have no neurons also have no utility. Thus, everything that's not an animal gets a scaling factor of zero. Sponges, which are classified as animals, have no neurons and thus also get a zero. Humans get a one. Everything else gets something in between a zero and a one.

My guess is that other primates are pretty close to a one, maybe a 0.95. Beyond that, I just don't know.

---------

Some thoughts on the scaling system:

The post Total Human Earth Utility used a scale from zero to one, which was a normalized (scaled/divided by ten) version of the zero to ten scale used in the HPI list. In the post, zero was the minimum possible human utility level and one was the maximum. This scale makes what I believe to be a bad assumption that there is no negative utility. (For more on negative utility see the posts Utility Curves and Vegetarianism And Negative Utility.)

An option that may be convenient is to use a zero to one or zero to ten scale for each species and then scale those values based on how strong that species's utility is. In other words, when evaluating the well being of members of specific species, we can use the same zero to one/ten scale, and then when we need to compare values across species, we scale them accordingly.

This scaling may or may not be linear. If it was linear, then we could use the same scaling factor for all utility levels. If it was nonlinear, then the scaling factor would be a function of utility level. For example, (on a zero to ten scale):

Linear:
Scaling factor between humans and cows is 2,
i.e. (human utility) = 2 * (cow utility)
A cow 1 corresponds to a human 2.
A cow 2 corresponds to a human 4.
A cow 3 corresponds to a human 6.
etc.

Nonlinear (Quadratic):
Scaling factor between humans and cows is equal to the cow utility,
i.e. (human utility) = (cow utility) * (cow utility) = (cow utility)^2
A cow 1 corresponds to a human 1.
A cow 2 corresponds to a human 4.
A cow 3 corresponds to a human 9.
etc.

Nonlinear (General):
Human utility level is some function of cow utility level,
i.e. (human utility) = f(cow utility)

A sound first step would be estimating linear scaling factors.

In our estimates, we should make sure to note if we're estimating typical levels to potential levels. For example, on the HPI zero to ten scale, humans have the potential to experience a ten but typically experience around a six. (For more on this see the post Population Utility Distribution.)

Total Human Earth Utility

Eventually, the sun will die out and with it, our species. That is, unless we extend our existence by colonizing space (likely) or arranging a non-solar energy source (plausible), or unless we shorten our existence by something along the lines of climate change or nuclear winter (plausible). However, to put things into perspective, we present a simple, "back of the envelope" calculation of the total utility humanity will get from the sun on Earth.

Note: I am using the short scale for names of numbers. Thus, 10^9 is one billion and 10^18 is one quintillion. For more on this see Long and short scales.

The simplest form is:
  u = p * a * t
Here u is total utility, measured in QALYs. p is average population, measured in number of people. a is average individual utility level, measured in QALYs per person-year. t is total time Earth will be hospitable for humans, measured in years. Reasonable estimates are:
  p = 10^10 (10 billion) people, which is approximately where the world's human population is projected to level off at later this century (link)
  a = 0.6, which is the average human life satisfaction from HPI normalized (divided by ten) so it would be on a scale from zero to one.
  t = 10^9 (1 billion) years, which is approximately the lifespan of Earth's biosphere (Caldeira, K. and J.F. Kasting. The life span of the biosphere revisited. Nature 360: 721-723)

Using this model and these numbers, we get
  u = (10^10 people) * (0.6 QALYs per person-year) * (10^9 years)
  u = 6*10^18 QALYs

Thus, the total human Earth utility should be on the order of 6*10^18 or 6 quintillion quality of life-adjusted years.

How accurate do I think this value is? Barring any major disasters, it seems reasonable to expect that we'll bump that value for average QALYs per person-year (a) up from 0.6 to at least 0.8 or 0.9 as we improve the human condition across the planet. It is at least plausible that the given value for average population level (p) may actually prove accurate given our recent tendency to maintain population levels below carrying capacity in wealthier, more urban parts of the world. Finally, assuming the geoscience is correct (it probably is, but I'm no geoscientist), the total time (t) should be accurate. However, given the likelihood that our species does not end when the sun is no longer supportive (see below), the exact value for (t) is of little relevance to contemporary decision making.

Increasing (a) is a sound goal and an uncontroversial one by today's standards.

Increasing (p) is a trickier matter. These days, we're more likely to hear concern about overpopulation or overcrowding. Furthermore, it remains unclear what the Earth's long term human carrying capacity will be given today's rapidly changing situations for resources (generally worsening) and technology (generally improving). In addition, we may personally (i.e. selfishly) prefer maintaining a population well below carrying capacity, as is the case throughout today's wealthier, more urban population. We may eventually want to restructure our society to encourage more procreation as is already done in France (link) and perhaps elsewhere.

Increasing (t) is out of our hands. But remember, (t) is total time Earth will be hospitable for humans. Whether we make it that far or farther is very much in our hands. Given today's rapid rate of technological advancement and the billion or so years we have to work with, it's fair worry more about making it that far than about making it farther. If short term threats don't end us, we'll be well positioned to survive post-sun.

For more on ensuring the long term viability of humanity, see Reducing the risk of human extinction by Jason Matheny (key quote: "We review the challenges to studying human extinction risks and, by way of example, estimate the cost-effectiveness of preventing extinction-level asteroid impacts.") and the Lifeboat Foundation, which discusses a wide range of threats to humanity.

Proximity, Uncertainty, And Effort

A common criticism of utilitarianism is that it requires us to care just as much about individuals who we have nothing to do with as about those individuals closest to us. For example, US conservative columnist George Will on utilitarian philosopher Peter Singer:

But proximity, even familial attachment--these are moral irrelevancies in Singer's analysis of one's obligations to others. Should one spend a sum to ease the suffering of a family member or send the same sum to ease the sufferings of 10 Sudanese? Singer is consistent: In the Sudan the money will better serve the world's total amount of happiness."

"Life and Death at Princeton", Newsweek, 09/13/99, Vol. 134 Issue 11, p80

Utilitarianism requires us to weight all individuals' interests equally. It does not, however, require us to try to help all individuals equally. Indeed, in many common cases we actually should try to help those closest to us, because the closer they are, the better we understand how to increase their quality of life. Similarly, effort that could be spent figuring out how to increase the quality of life of more distant individuals could be spent actually increasing the quality of life of those closer to us.

For an extreme example, consider what shirt I should wear tomorrow. I should give it some thought so that I pick a shirt I feel like wearing which would be appropriate for tomorrow's weather, activities, etc. However, choosing what shirt I should wear would be a waste of your time, especially if you've got little to do with me.

I don't know for sure which shirt I should wear tomorrow in order to maximize quality of life on Earth. However, I can make a pretty good guess, and a much better one than you can. My estimations might look like:

in which case I'd choose shirt number two. However, unless you put a lot of extra effort into it, your estimations probably look like:

Thus, in order to maximize expected utility, I should choose the shirt and you should find something better to do with your time.

In a more serious example, consider where to donate surplus money. Money is interesting because it can go halfway around the planet to people you've never met before quite easily: just donate to Oxfam. However, if you haven't already, you should take at least a little effort to convince yourself that Oxfam really is a worthy recipient of your surplus money. Indeed, the whole Surplus Money post is about figuring out if we should donate to Oxfam or to some other organization. In other words, we're putting in effort to bring ourselves closer not just the to organizations but to the individuals that they would benefit.

Is this effort to analyze charities well spent? I believe so. First, we can easily share our results with the world both online and off, so others won't have to repeat our work. Furthermore, I personally enjoy the work and suspect it may help my career, which is worth something in its own right.

If you're interested in the work, join us! Otherwise, you can skip the effort and just take our advice or someone else's. Alternatively, perhaps you can make relatively effortless and effective donations such as via remittances. Would your remittances/etc beat Oxfam? That's probably your call, since we're unlikely to put the effort into analyzing the specific circumstances of the remittances/etc.

Utility Measures

Over the years, people have discussed if it's possible to measure utility or even to compare utility of different individuals on either a ordinal (ranked) or cardinal (absolute) basis. A common criticism of utilitarianism is that utility is incommensurable, i.e. that it can't be compared across different individuals, making it impossible to determine what actions maximize utility.

Personally, this view seems inconsistent with the underlying physiology of utility or with basic personal experience. To me, the question is not if we can measure utility but how accurately we can do so. Indeed, not only is it possible to measure utility, but we do it all the time. Here is an overview of the existing measures of utility I'm currently aware of. Please add others in the comments or email them to me for me to add.

I'll say up front that the one glaring absence is the complete lack of even crude measures of utility of non-humans, despite the general consensus that members of other species do have non-zero utility (i.e. they experience pleasure, pain, etc.).

Measures of individual utility:

Self-reported happiness as used in happiness economics. Here, the measure of an individual's utility is that individual's response to a question along the lines of "On a scale from one to ten, how happy are you?". While a five for you might be different from a five for me, it's probably at least similar, and in large studies, these differences get averaged out. For more, see the post Happiness Economics: Recommended Reading.

Disability-Adjusted Life Year (DALY) as used in public health. One DALY is one human life year in full health. A year in less than full health will will have a lower DALY value. It strictly focuses on health and not on quality of life more generally and thus is an imperfect but nonetheless useful statistic, such as for measuring the impact of a disease outbreak. For more, see Wikipedia

Quality-adjusted life years as used in medicine. Ultimately the QALY should replace the DALY as the more general unit of utility, but for now its use appears restricted to the medical community. For more, see Wikipedia

At the time of this post, measures of utility using brain imaging appear to be a nonexistent pipe dream.

Measures of collective utility:

Happy Planet Index (HPI), by the new economics foundation (London). It "shows the ecological efficiency with which human well-being is delivered around the world." (link) This one's a personal favorite. For more see the post Links, Links, Links, the HPI website, or Wikipedia.

Human Development Index (HDI), by the UN Development Program. It strives to measure well being from a bundle of external (i.e. non-neurophysiological) phenomena, including literacy and life expectancy. I consider it a middle ground between strictly financial measures such as GDP and pure quality of life measures such as HPI. See Wikipedia.

Gross National Happiness (GNH), a counterpart to GNP/GDP devised by the King of Bhutan to be focused more on happiness and on consumption. "the four pillars of GNH: economic self-reliance, a pristine environment, the preservation and promotion of Bhutan’s culture, and good governance in the form of a democracy." (link) For more see Wikipedia.

Genuine Progress Indicator, as in welfare and green economics. "The Genuine Progress Index is based on the fundamental understanding that social, economic and environmental realities are inextricably linked." (link) For more see Wikipedia.

Various financial measures including Gross National Product (GNP), Gross Domestic Product (GDP), and others, which measure "the total value of final goods and services produced in a year" (link) and are perhaps the most common measures of collective well being, especially among economists. While I generally agree with these measures' critics, the measures do provide at least some insight.

For criticism of the financial measures, I quote Warren Buffett: "If I wanted to, I could hire 10,000 people to do nothing but paint my picture every day for the rest of my life. And the GNP would go up. But the utility of the product would be zilch, and I would be keeping those 10,000 people from doing AIDS research, or teaching, or nursing. I don't do that though."
The quote is supposed to be from Warren Buffett Speaks by Janet Lowe, 1997, p. 165-166, ISBN 047116996X

Utility Cost Of Greenhouse Gas

In light of the recent Stern Review Report on the Economics of Climate Change (see also the Felicifia post Reviewing The Stern Review), let me propose to whatever audience this humble blog yet has:

What is the cost in utility of greenhouse gas?

In other words, how much utility must one derive from an action that results in greenhouse gas emission in order to offset the decrease in utility from the gas and therefore justify the action?

We should ultimately put this in practical terms, such as "Saving a gallon of gas is about as good as a human living an extra two months in good health." which would mean that if you can extend a human's life in good health by at least two months with by burning a gallon of gas, then you should do it, but if it's any less than that, then you shouldn't. (I'm just making up the two months number.) In addition to gallons of gas, I'd personally like to see a hamburger vs. veggieburger comparison: Sure, that hamburger might test better, but does it taste better enough? (Ignore animal welfare here for now; we'll come back to that soon enough.)

I know this site only has a small handful of readers thus far, but I'm very excited about the idea of online collaboration on the most pressing of topics and want to give it a spin to see what happens. Let's sort the problem out in the comment thread or, if you prefer to remain anonymous or otherwise don't want to use the comment thread, then email me at sethbaum [at] gmail.com.

Reviewing The Stern Review

Yestarday, the UK government released the Stern Review Report on the Economics of Climate Change by Sir Nicholas Stern. At a glance, it appears to be an excellent, encyclopedic reference on the topic. I'm tempted to call this the biggest event since Warren Buffett announced his massive donations to the Gates Foundation. To help us review the report further, I've skimmed through the Executive Summary (pdf) and pulled what I found to be the key quotes. The document is well-organized, so going to a quote's page will give you further detail on that topic. Also, here is a rebuttal of the Stern Review from the Telegraph and a rebuttal of the rebuttal from a RealClimate comment thread. (RealClimate may post on it shortly, in which case I'll change the link to that.)

"The Review first examines the evidence on the economic impacts of climate change itself, and explores the economics of stabilising greenhouse gases in the atmosphere. The second half of the Review considers the complex policy challenges involved in managing the transition to a low-carbon economy and in ensuring that societies can adapt to the consequences of climate change that can no longer be avoided." (p.1)

"Climate change presents a unique challenge for economics: it is the greatest and widest-ranging market failure ever seen." (p.1)

"...the evidence gathered by the Review leads to a simple conclusion: the benefits of strong, early action considerably outweigh the costs." (p.2)

The evidence shows that ignoring climate change will eventually damage economic growth. Our actions over the coming few decades could create risks of major disruption to economic and social activity, later in this century and in the next, on a scale similar to those associated with the great wars and the economic depression of the first half of the 20th century. And it will be difficult or impossible to reverse these changes. Tackling climate change is the pro-growth strategy for the longer term, and it can be done in a way that does not cap the aspirations for growth of rich or poor countries. The earlier effective action is taken, the less costly it will be.

At the same time, given that climate change is happening, measures to help people adapt to it are essential. And the less mitigation we do now, the greater the difficulty of continuing to adapt in future.

(p.2)

"Climate change threatens the basic elements of life for people around the world - access to water, food production, health, and use of land and the environment." (p.6)

"The impacts of climate change are not evenly distributed - the poorest countries and people will suffer earliest and most. And if and when the damages appear it will be too late to reverse the process." (p.7)

"Integrated assessment models provide a tool for estimating the total impact on the economy; our estimates suggest that this is likely to be higher than previously suggested." (p.8)

"Emissions have been, and continue to be, driven by economic growth; yet stabilisation of greenhouse-gas concentrations in the atmosphere is feasible and consistent with continued growth." (p.11)

"The Review estimates the annual costs of stabilisation at 500-550ppm CO2e to be around 1% of GDP by 2050 - a level that is significant but manageable." (p.12)

"Delay in taking action on climate change would make it necessary to accept both more climate change and, eventually, higher mitigation costs." (p.15)

"Policy to reduce emissions should be based on three essential elements: carbon pricing, technology policy, and removal of barriers to behavioural change." (p.18)

"Adaptation policy is crucial for dealing with the unavoidable impacts of climate change, but it has been under-emphasised in many countries." (p.21)

"An effective response to climate change will depend on creating the conditions for international collective action." (p.22)

"Creating a broadly similar carbon price signal around the world, and using carbon finance to accelerate action in developing countries, are urgent priorities for international co-operation." (p.23)

"Curbing deforestation is a highly cost-effective way of reducing greenhouse gas emissions." (p.25)