Paul Collier: The sharp increase in the world price of staple foods is an inconvenience for consumers in the rich world, but for consumers in the poorest countries, especially in Africa, it is a catastrophe. Despite the predominance of peasant agriculture, most African countries are net food importers and food accounts for over half of the budget of low-income households. This is the result of decades of agricultural stagnation combined with growing populations. Although many of the net purchasers are rural, evidently the problem is at its most intense in the urban slums. These slums are political powder kegs and so rising food prices have already triggered riots. Indeed, they sow the seeds of an ugly and destructive populist politics.

Why have food prices rocketed? Paradoxically, this squeeze on the poorest has come about as a result of the success of globalization in reducing world poverty. As China develops, helped by its massive exports to our markets, millions of Chinese households have started to eat better. Better means not just more food but more meat, the new luxury. But to produce a kilo of meat takes six kilos of grain. Livestock reared for meat to be consumed in Asia are now eating the grain that would previously have been eaten by the African poor. So what is the remedy?

The best solution to a problem is often not closely related to its cause (a proposition that that might be recognized in the climate change debate). China’s long march to prosperity is something to celebrate. The remedy to high food prices is to increase food supply, something that is entirely feasible. The most realistic way to raise global supply is to replicate the Brazilian model of large, technologically sophisticated agro-companies supplying for the world market. To give one remarkable example, the time between harvesting one crop and planting the next, in effect the downtime for land, has been reduced an astounding thirty minutes. There are still many areas of the world that have good land which could be used far more productively if it was properly managed by large companies. For example, almost 90% of Mozambique’s land, an enormous area, is idle.

Unfortunately, large-scale commercial agriculture is unromantic. We laud the production style of the peasant: environmentally sustainable and human in scale. In respect of manufacturing and services we grew out of this fantasy years ago, but in agriculture it continues to contaminate our policies. In Europe and Japan huge public resources have been devoted to propping up small farms. The best that can be said for these policies is that we can afford them. In Africa, which cannot afford them, development agencies have oriented their entire efforts on agricultural development to peasant style production. As a result, Africa has less large-scale commercial agriculture than it had fifty years ago. Unfortunately, peasant farming is generally not well-suited to innovation and investment: the result has been that African agriculture has fallen further and further behind the advancing productivity frontier of the globalized commercial model. Indeed, during the present phase of high prices the FAO is worried that African peasants are likely to reduce their production because they cannot finance the increased cost of fertilizer inputs. While there are partial solutions to this problem through subsidies and credit schemes, large scale commercial agriculture simply does not face this problem: if output prices rise by more than input prices, production will be expanded because credit lines are well-established.

Our longstanding agricultural romanticism has been compounded by our new-found environmental romanticism. In the United States fears of climate change have been manipulated by shrewd interests to produce grotesquely inefficient subsidies for bio-fuel. Around a third of American grain production has rapidly been diverted into energy production. This switch demonstrates both the superb responsiveness of the market to price signals, and the shameful power of subsidy-hunting lobby groups. Given the depth of anti-Americanism in Europe it is, of course, fashionable to criticize the American folly with bio-fuels. But Europe has its equivalent follies.

First, the European Commission is now imitating the American bio-fuels policy. At present the programme is small enough to be unimportant, but we need to pull it back before it does real damage. We have surely learnt enough about European agriculture to realize how important it is to kill this incipient scam before we are engulfed by it. But the true European equivalent of America’s folly with bio-fuels is the ban on GM. Europe’s distinctive and deep-seated fears of science have been manipulated by the agricultural lobby into yet another form of protectionism. The ban on both the production and import of genetically modified crops has obviously retarded productivity growth in European agriculture: again, the best that can be said of it is that we are rich enough to afford such folly. But Europe is a major agricultural producer, so the cumulative consequence of this reduction in the growth of productivity has most surely rebounded onto world food markets. Further, and most cruelly, as an unintended side-effect the ban has terrified African governments into themselves banning genetic modification in case by growing modified crops they would permanently be shut out of selling to European markets. Africa definitely cannot afford this self-denial. It needs all the help it can possibly get from genetic modification. Not only is Africa currently being hit by rising food prices, over the longer term it will face climatic deterioration in the context of a rapidly growing population.

While the policies needed for the long term have been befuddled by romanticism, the short term global response has been pure beggar-thy-neighbour. It is easier for urban slum dwellers to riot than for farmers: riots need streets, not fields. And so, in the internal tussles between the interests of poor consumers and poor producers, the interests of consumers have prevailed. Governments in grain-exporting countries have swung prices in favour of their consumers and against their farmers by banning exports. These responses further politicize and fragment an already confused global food market. They increase the risks of investing in commercial-scale food production and drive up prices further in the food-importing countries. Unfortunately, trade in agriculture has been the main economic activity to have resisted being subject to global rules. We need stronger and fairer globalization, not less of it.

here

Eliezer a week ago:

At the frontier of scientific chaos and scientific confusion, you find problems of thinking that are not taught in academic courses, and that have not been reduced to calculation. … It will seem that you must do philosophical thinking in order to sort out the confusion. But … it is usually not a professional philosopher who wins all the marbles – because it takes intimate involvement with the scientific domain in order to do the philosophical thinking. … There is … [a] place for professional philosophers in the world. Some problems are so chaotic that there is no established place for them at all in the halls of science. But those “professional philosophers” would be very, very wise to learn every scrap of relevant-seeming science that they can possibly get their hands on.

Once upon a time all academia was “philosophy”, all using the same method of informal argument. One by one groups focusing on particular topics developed specialized methods, and split off to form new disciplines. Philosophy is now the “miscellaneous” discipline, the only one left engaging many big hard questions. Philosophers mostly use this method:

Compare intuitions about selected cases with general principles expressed in words. Discuss wording ambiguities and find extreme case-principle conflicts. Suggest new wordings for better matches.

Being drawn to big hard questions, I love that philosophers engage those questions. My main lament is philosophers’ reluctance to calculate; they mostly use their standard method, even when more exact formal models are available. Two examples:

Born rule in many worlds — physicists mostly punt to philosophers, who use flimsy excuses to declare meaningless the use of specific quantum models to calculate the number of worlds that see particular experimental results. This leaves them free to settle the question by proposing abstract principles that imply the Born rule. (At least a few do this semi-formally.) Two recent workshops here and here, my stuff here.

Rationality of disagreement – Economists studied this since Aumann ’76, but mostly as a theory foil, not to critique human disagreement. Recently philosophers have written dozens of papers on when it is rational to disagree, basically ignoring the Aumann-started literature. Some say disagreement is so obviously rational that if models say otherwise, so much the worse for models. Others give flimsy reasons for dismissing model relevance, but mostly I think they can’t be bothered to follow the calculations. My overview here.

Of course the real problem is that academia discourages interdisciplinary work. Researchers using one method give too little consideration to people or work using other methods, making it hard to mix or switch methods. I might well commit similar sins if similarly empowered.

My other laments about philosophers follow from heavy reliance on their main method: they seem too enamored of words over more formal notation, and they seem to trust their intuitions way too much.

Besides Eliezer’s comment, this post was also sparked by an enjoyable day I spent last Monday talking to Princeton philosophers: Prof. Adam Elga for lunch, then guest lecturing for his graduate seminar, attended by Prof. Thomas Kelly and blogger Richard Chappell, and then dinner with Richard, Joshua Harris, and other students. Previously, I’ve spend weeks with Oxford philosophers, including Nick Bostrom, Nicholas Shackel, and Toby Ord, and years with philosophically well-read GMU colleagues (Tyler Cowen is even well published).

(Thanks JJ.)

Here’s Richard Thaler and Cass Sunstein in today’s LA Times:

Those who design supermarkets and school cafeterias are engaged in what we call “choice architecture”: the organization of the context in which people make decisions. Choice architects are everywhere. If you design the ballot that voters use to choose candidates, you are a choice architect. If you are a doctor and must describe the alternative treatments available to a patient, you are a choice architect. If you design the form that new employees fill out to enroll in the company healthcare plan, you are a choice architect. If you are a parent, describing possible educational options to your son or daughter, you are a choice architect. If you are a salesperson, you are a choice architect (but you already knew that).

There are many parallels between choice architecture and more traditional forms of architecture. A crucial parallel is that there is no such thing as a “neutral” design. Cognitive psychology and behavioral economics have shown that small and apparently insignificant contextual details can have a major effect on people’s behavior.

They are right about the importance of context and framing. And I very much like the idea of “choice architecture,” in its place. I agree that there is generally no “neutral” design. But this opens them up to an obvious line of argument. What kind of architecture are we aiming at? A mundane utilitarian edifice or a challenging vertiginous space? Homey comfort or antiseptic austerity?

Sunstein and Thaler may wish to design the presentation of choices to bias decisions in favor of, say, happiness. But other choice architects may be more interested in biasing our choices toward virtue or toward participation in great collective projects. Obviously everyone is a “choice architect” to some degree in his or her daily intercourse with others. And some people, like marketers and salespeople, try to shape choices for a living. The thing is, we often rightly resent their attempts to manipulate us, but at least we are more or less in control of our exposure to such people. But when choice architecture is implemented politically, we cannot opt out of these attempts at manipulation, attempts which may or may not be benign. That’s a big problem because political choice architecture may do a great deal to shape us, even if, in its “libertarian paternalist” incarnation, it makes a show of leaving the ultimate choice open to individuals.

For example, I would object if President John McCain implemented a policy of opt-out national service because such a policy would communicate all-too-clearly that individuals need some kind of special justification or rationale not to serve the state. The default rule itself contains meaningful content. If allowed to stand, such a policy could shape norms and individual preferences in a direction antagonistic to the value of autonomy. Soon enough we might find ourselves asking, “Why should you be able to opt out at all?” The paternalistic nudge may “leave the choice open” but accepting the legitimacy of the certain nudges may imperil liberty.

Back to Thaler and Sunstein:

Let’s return to the cafeteria line. If, all things considered, you think the arrangement of food ought to nudge kids toward what’s best for them, then we welcome you to our new movement: libertarian paternalism. We are keenly aware that both those words are weighted down by stereotypes from popular culture and politics. Why combine two often reviled and seemingly contradictory concepts? The reason is that if the terms are properly understood, both concepts reflect common sense. They are far more attractive together than alone — and taken together, they point the way to a whole new approach to the role of government.

The libertarian aspect of the approach lies in the straightforward insistence that, in general, people should be free to do what they like. They should be permitted to opt out of arrangements they dislike, and even make a mess of their lives if they want to. The paternalistic aspect acknowledges that it is legitimate for choice architects to try to influence people’s behavior in order to make their lives longer, healthier and better.

OK. But there’s sure a lot of disagreement about “better” isn’t there? I sense that the fact of pluralism isn’t their chief concern. And the “libertarian” part of this I suspect really is a ruse. If Sunstein and Thaler were our wise choice archietcts, would it be legal or illegal for employers to not offer employment contracts without opt-out savings/investment accounts? Forcing some people to frame choices they offer to others in a way that will bias those people’s choices can’t be libertarian in any meaningful sense.

Then there is the problem of the meaning of default rules. Thaler and Sunstein suggest “If we want to increase the supply of transplant organs in the United States, we could presume that people want to donate, rather than treating nondonation as the default.” But isn’t this the sort of presumption that itself contains a great deal of normative and symbolic content? Does it not say, “Your body presumptively belongs to the commonwealth, and you must take special action to use it as you and your family wish?” Wouldn’t the very existence of such a default rule bias subsequent political deliberation against alternative policies, like legalizing markets in organs and tissue?

Individual choices made again and again create habits. Coordinated patterns of individual actions create norms. Choice architecture not only nudges us to do what we already want to do, but over time shapes what we want and shapes the social context and meaning of choice. By modifying the local frame of choice, the architect systematically affects the global frame of future choices. Suppose manipulating the context of micro-level individual choices eventually shifts political preferences. Do we think it is okay for the state to aim at producing a population with different political preferences, so that they will vote for the things that we, the choice architects, know will make them better off? (My critique of Social Security is that this is terribly illiberal and is exactly what happened.) Obviously this is completely pernicious and unacceptable. Which may be one reason why a chaotic ad hoc gallimaufry of completing choice frames, which add up to nothing in particular and tilts at no one set of values may be precisely what leaves us best off in the end.

—

Dissertation topic: If there is no “neutral” choice architecture, does that mean liberal neutrality is impossible? Short answer: No. It means that neutral neutrality is impossible.

April 4, 2004

The Secret Source of Google’s Power

by skrenta at 2:11 PM

Much is being written about Gmail, Google’s new free webmail system. There’s something deeper to learn about Google from this product than the initial reaction to the product features, however. Ignore for a moment the observations about Google leapfrogging their competitors with more user value and a new feature or two. Or Google diversifying away from search into other applications; they’ve been doing that for a while. Or the privacy red herring. No, the story is about seemingly incremental features that are actually massively expensive for others to match, and the platform that Google is building which makes it cheaper and easier for them to develop and run web-scale applications than anyone else.

I’ve written before about Google’s snippet service, which required that they store the entire web in RAM. All so they could generate a slightly better page excerpt than other search engines.

Google has taken the last 10 years of systems software research out of university labs, and built their own proprietary, production quality system. What is this platform that Google is building? It’s a distributed computing platform that can manage web-scale datasets on 100,000 node server clusters. It includes a petabyte, distributed, fault tolerant filesystem, distributed RPC code, probably network shared memory and process migration. And a datacenter management system which lets a handful of ops engineers effectively run 100,000 servers. Any of these projects could be the sole focus of a startup.

Speculation: Gmail’s Architecture and Economics

Let’s make some guesses about how one might build a Gmail.

Hotmail has 60 million users. Gmail’s design should be comparable, and should scale to 100 million users. It will only have to support a couple of million in the first year though.

The most obvious challenge is the storage. You can’t lose people’s email, and you don’t want to ever be down, so data has to be replicated. RAID is no good; when a disk fails, a human needs to replace the bad disk, or there is risk of data loss if more disks fail. One imagines the old ENIAC technician running up and down the isles of Google’s data center with a shopping cart full of spare disk drives instead of vacuum tubes. RAID also requires more expensive hardware — at least the hot swap drive trays. And RAID doesn’t handle high availability at the server level anyway.

No. Google has 100,000 servers. [nytimes] If a server/disk dies, they leave it dead in the rack, to be reclaimed/replaced later. Hardware failures need to be instantly routed around by software.

Google has built their own distributed, fault-tolerant, petabyte filesystem, the Google Filesystem. This is ideal for the job. Say GFS replicates user email in three places; if a disk or a server dies, GFS can automatically make a new copy from one of the remaining two. Compress the email for a 3:1 storage win, then store user’s email in three locations, and their raw storage need is approximately equivalent to the user’s mail size.

The Gmail servers wouldn’t be top-heavy with lots of disk. They need the CPU for indexing and page view serving anyway. No fancy RAID card or hot-swap trays, just 1-2 disks per 1U server.

It’s straightforward to spreadsheet out the economics of the service, taking into account average storage per user, cost of the servers, and monetization per user per year. Google apparently puts the operational cost of storage at $2 per gigabyte. My napkin math comes up with numbers in the same ballpark. I would assume the yearly monetized value of a webmail user to be in the $1-10 range.

Cheap Hardware

Here’s an anecdote to illustrate how far Google’s cultural approach to hardware cost is different from the norm, and what it means as a component of their competitive advantage.

In a previous job I specified 40 moderately-priced servers to run a new internet search site we were developing. The ops team overrode me; they wanted 6 more expensive servers, since they said it would be easier to manage 6 machines than 40.

What this does is raise the cost of a CPU second. We had engineers that could imagine algorithms that would give marginally better search results, but if the algorithm was 10 times slower than the current code, ops would have to add 10X the number of machines to the datacenter. If you’ve already got $20 million invested in a modest collection of Suns, going 10X to run some fancier code is not an option.

Google has 100,000 servers.

Any sane ops person would rather go with a fancy $5000 server than a bare $500 motherboard plus disks sitting exposed on a tray. But that’s a 10X difference to the cost of a CPU cycle. And this frees up the algorithm designers to invent better stuff.

Without cheap CPU cycles, the coders won’t even consider algorithms that the Google guys are deploying. They’re just too expensive to run.

Google doesn’t deploy bare motherboards on exposed trays anymore; they’re on at least the fourth iteration of their cheap hardware platform. Google now has an institutional competence building and maintaining servers that cost a lot less than the servers everyone else is using. And they do it with fewer people.

Think of the little internal factory they must have to deploy servers, and the level of automation needed to run that many boxes. Either network boot or a production line to pre-install disk images. Servers that self-configure on boot to determine their network config and load the latest rev of the software they’ll be running. Normal datacenter ops practices don’t scale to what Google has.

What are all those OS Researchers doing at Google?

Rob Pike has gone to Google. Yes, that Rob Pike — the OS researcher, the member of the original Unix team from Bell Labs. This guy isn’t just some labs hood ornament; he writes code, lots of it. Big chunks of whole new operating systems like Plan 9.

Look at the depth of the research background of the Google employees in OS, networking, and distributed systems. Compiler Optimization. Thread migration. Distributed shared memory.

I’m a sucker for cool OS research. Browsing papers from Google employees about distributed systems, thread migration, network shared memory, GFS, makes me feel like a kid in Tomorrowland wondering when we’re going to Mars. Wouldn’t it be great, as an engineer, to have production versions of all this great research.

Google engineers do!

Competitive Advantage

Google is a company that has built a single very large, custom computer. It’s running their own cluster operating system. They make their big computer even bigger and faster each month, while lowering the cost of CPU cycles. It’s looking more like a general purpose platform than a cluster optimized for a single application.

While competitors are targeting the individual applications Google has deployed, Google is building a massive, general purpose computing platform for web-scale programming.

This computer is running the world’s top search engine, a social networking service, a shopping price comparison engine, a new email service, and a local search/yellow pages engine. What will they do next with the world’s biggest computer and most advanced operating system?

• Integrated: Consolidates insights from economics, political science, psychology, sociology, and the other social sciences into one parsimonious and systematic framework.
• Theory: Theoretical in the sense of a scientific theory — making testable predictions about yet-unobserved phenomena. Employs formal (including mathematical) analyses and models.
• Societal: Not just “social;” intended to account for society-as-unit (society, per Jervis, at the “system” level) as well as for lower level “social” interactions between individuals and small groups. Explains the interactions and interfaces between the two levels.
• Dynamics: Accounts for change in societies and interaction effects between societies in a rigorous and systematic way. Not an equilibrium theory, but subsumes equilibrium theory in the sense that it accounts for the appearance, maintenance, and disappearance of equilibria within the more general framework that describes change.

The beginning…

Researchers use economic games to investigate how people cooperate in real-life. Now a team led by Benedikt Herrmann, at the University of Nottingham, have identified striking differences in the way university students from different countries play one such game known as The Public Goods Game. Compared with students from developed Western nations, students from less democratic countries like Saudi Arabia, Oman and Belarus tended to punish not only free-loaders, but also cooperative players, with the result that cooperation in their groups plummeted.

…

Anti-social punishment occurred more in those countries, including Belarus and Saudi Arabia, shown by surveys to have less faith in the rule of law and less belief in civic cooperation. In a commentary on the findings, published in the same journal, Herbert Gintis of the Sante Fe Institute, said the results challenge the way people have tended to view capitalist democracies. “The success of democratic market societies may depend critically upon moral virtues as well as material interests, so the depiction of civil society as the sphere of ‘naked self-interest’ is radically incorrect,” he wrote.

From BPS Research Digest, via Marginal Revolution. Full citation for the study in question:

Herrmann, B., Thoni, C., Gachter, S. (2008). Antisocial Punishment Across Societies. Science, 319(5868), 1362-1367. DOI: 10.1126/science.1153808