Sunlight and Happiness

Take an African animal and put it in Norway , or Seattle. Do you think it would be happy in that climate? But that's us. Growing up in Pennsylvania, I was familiar with the long summer days and long winter nights of temperate regions - but the week between Christmas and New Year, standing in Regent's Park in London during my first visit to the U.K., and looking at the tired glow of the sun through the clouds barely shoulder-high on the horizon at 11 in the morning - I started to realize why humans are so eager to to reduce their latitudes.

In the U.S., much of which is at sunnier latitudes, estimates of sufferers of clinical seasonal affect disorder or just winter blues run 10-15%. There are also claims that alcoholism correlates positively with latitude.

So I was not surprised when I looked at a sunshine map of the U.S., and saw a stripe of cloudiness along the Appalachians that matched (somewhat) with a portion of the unhappy stripe in a recent happiness survey:





The unhappy stretch extends from the west side of the Appalachians in Tennessee and Kentucky all the way through Arkansas and Oklahoma, which aren't as gloomy. I often point out to whiners in the Pacific Northwest that they aren't any cloudier than the poor folks in West Virginia and central/western PA.

Oddly enough, I noticed earlier that the unhappy stripe correlated even better with voting Republican for President in the 2008 U.S. elections.

Institutions and Values Matter

This evening I was playing around with cost of living, quality of life, latitude and average temperature numbers. Basically I was investigating the idea that countries at very high or low latitudes had a worse quality-per-cost ratio than those in temperate regions; that is to say, sure, maybe Norway has a higher quality of life, but it costs a lot more to live there than Spain, and do you get that much more? There is a lot of forced investment in infrastructure because of the marginal environment that they don't have to worry quite as much about in Spain.

I found only extremely weak connections, so I won't bother posting the data and graphs. But this isn't the first time I've looked for such connections. In fact whenever I've looked for connections between some economic or social indicator on one hand, and a non-human aspect of a country's real estate on the other (latitude, resources, climate) I either find no relationship, or a relationship that cannot be separated from the confounding facts of history, like the inheritance of certain values and institutions, particularly from Enlightenment Europe as it was colonizing the West. In fact, the first question on examining such relationships is whether we should try to separate the trend from history.

More and more I find myself siding with the development experts who state that it's the institutions of a country that matter more than anything else - not its resources or climate, temporarily wealthy petro-theocracies notwithstanding. One assumes that the values of the people have to support such institutions. As the U.S. is learning, you can't just drop a democracy onto a culture that has no history of open discourse and personal responsibility, even before the Hussein regime, and in fact I would argue that Indo-European cultures in general have at their root a value of parliamentary decision-making and openness that is rare elsewhere; why would the world's oldest parliament, the Althing, have appeared in Norse Iceland in the guts of the brutal Middle Ages?

Another way of emphasizing the importance of institutions, and underlying values, is in traditional economics terms; that it's the labor and not the land that makes life better and generates wealth. This shouldn't be surprising; it tracks the development of technology, which continually increases the potential productivity of human beings and their power to shape their environment. The last school of economics that discounted the role of labor entirely was the physiocrats in the eighteenth century, and no one has been able to make that mistake since. This can also explain the (near) disappearance of slavery, from a purely cynical economic standpoint. Three thousand years ago, there wasn't a whole lot more you could contribute as a scribe or farmer than you could as a slave. By the nineteenth century, the institution had shrunk enough in importance that moral concerns could override whatever clout the related industries retained, first in England, and later in the U.S. In 2009, from an economic standpoint, the idea of forcibly feeding and housing a person so they can pick plants instead of voluntarily build a better microchip seems patently absurd. In 2109 it will seem moreso.

The question remains then of how to measure the goodness of institutions without the tautology of just saying that whatever raises per capita income and gross national happiness must be good. Measuring values would be trickier still. Encouraging values that support good institutions and therefore the elimination of misery is the most important and difficult question of all.

Proposal: Adopt a Universal Phonetic Alphabet Based on Roman Characters

See my original post at halfbakery.com, which includes follow-up comments.

The benefits of literate people around the world being able to communicate, regardless of spoken language, are obvious. When building a writing system, there are two possible approaches.

1) Use symbols based on meaning. In such systems there are necessarily a lot of these (in the thousands. Chinese uses this strategy.)

2) Use phonetic values (an alphabet or syllabary). English (and many other non-Roman-alphabet languages) use this system; the number of symbols is often substantially less than 100.

I propose an alphabet, rather than a system of ideograms, and specifically a phonetic version of the Roman alphabet, because

a) well over half of humans live in countries where the Roman alphabet has official or co-official status (3.8 billion)

b) alphabets are easier to learn (if you are a first-language Chinese-speaker and need 3,000 characters to read a newspaper, how hard can it be to learn 30 more?)

The main benefit will be facilitation of second-language learning, rather than universal communication (any monographic alphabet-user learning Japanese or Chinese can attest to this). Because of course different languages have different sounds, an expanded Roman alphabet could be used (mimicking the International Phonetic Alphabet?) That would be more fair, so everyone has to learn a somewhat new system, even people who already read Roman characters.

Someone posted a proposal here that all languages be written in ideograms. Beyond the difficulty of teaching ideograms to alphabet-readers, it's very difficult to adopt these symbols between languages, given differences in word order and grammar. The best-known example, Japanese, uses a klugey system of Chinese characters with home-grown syllabary characters scotch-taping them together within Japanese grammar; and the ideograms do drift from their original meanings, defeating the purpose anyway of adopting such a difficult system.

The problem of implementation is first and foremost a political one, of convincing the Chinese and Arabic-speaking governments of educating their citizenry to be at least bi-scriptural. But Turkey has done so, without which change it's doubtful whether there would even be an argument today over whether they could join the EU. [Added later: it turns out that there were serious proposals in Meiji Japan to Romanize Japanese; they would have beat Ataturk to the punch. There are actually books from this era written in Romanji-Japanese. A study of why it didn't catch on would be informative to this proposal.]

A Proposal: Compile Constitutions in Programming Language So They're Consistent

This was originally posted at halfbakery.com, where you can see the follow-up comments.

Most democracies have largely secular, rational post-enlightenment systems of government whose power flows neither from gun barrels nor arguments from authority to continue operating. However, because of the advance of technology, the laws these governments pass (and the way they can operate) will continue running into situations that their founders couldn't possibly have anticipated.

Currently many of these problems are solved by court rulings, which establish precedents. These precedents accumulate until there are layers upon layers, not all of them consistent with each other. Laws passed by legislative bodies also can take the form of an inconsistent patchwork that fail to take into account what went before.

By writing a constitution in a logical programming language that generates new laws and automatically checks for internal conflicts, these inefficiencies and inconsistencies can be avoided. Governments would become much truer to the ideal of being made of laws, and not of men.

Inflection Points in History: 1965 and 1990

It's tempting to try to find a point in time when an old zeitgeist fled and a new one took over. Anyone who does this in print must recognize that they're generalizing. After all, even in periods of real tumult, the zeitgeist is really just a constellation of attention-grabbing characteristics that mostly move independently of one each other. Art history example: the Renaissance is widely considered to have become the Mannerist period by the time Michelangelo began work on the the Last Judgment in 1537. But can we find a transitional work and point at emerging themes and say, here, this is the inflection point? Doing this with culture is not quite as easy as in biology, where there must be a clear linear descent.

Defining an age and trying to find its joints is necessarily a sloppy business, but this doesn't stop us. John McWhorter is a linguist, formerly of UC Berkeley and now with the Manhattan Institute, who wrote Doing Our Own Thing: The Degradation of Language and Music and Why We Should, Like, Care. McWhorter is a lover of the English language and the pages of this book mostly take the form of an elegy for a formal style of rhetoric (or really, the existence of rhetoric as such) that has passed into history in the United States, evidenced by the lesser demands placed on modern music and public speech. Mass media provide sensible landmarks of public taste for these kinds of discussions because they're a shared experience. McWhorter expounds on why the shift might have occurred and repeatedly comes back to 1965 as the inflection point, going so far as to find a "transition species", a so-called cultural archaeopteryx, in a live performance by Sammy Davis Jr. that had one foot in the old, formal style and one in the new, structureless, self-indulgent informality. McWhorter argues that a host of values and attitudes shifted along with this sharply punctuated 1965 transition.

Thoughtful people interested in the cultural changes of their country (and where they fit into it) can't help but find these speculations engaging. Probably the most famous treatment of the shifting of attitudes is Strauss and Howe's Generations. They attempt to explain history in cyclic terms with 4 recurring generational types, each determined by the nurturing patterns of the previous generation.

Without addressing Strauss and Howe's generational types, I've often speculated that a more recent and perhaps less profound cultural transition took place around my own coming of age, and there's a link to McWhorter's 1965. The 1980s in the U.S. - when I passed from kindergarten to tenth grade - in retrospect seem an oddly conservative island, a repeat of the 50s sandwiched between the era of disco, drugs and Vietnam on one side and grunge and the early internet on the other. Why? The kids of the post-1965, post-formal generation weren't yet out in the world on their own, independently interacting with the world, spreading those post-formal values. If you got married in 1963, had your first kid in 1965, she would have started college in 1983, carrying forward her parents' pre-1965-transition values. On the other hand, if you met at Woodstock and had your first kid two years later, she would start college in 1989.

What Happened in 1990?

My inspiration to collapse my thoughts into this blog post was a post on Andrew Sullivan's blogs, showing a sharp positive change in public perception of gay people in 1990. On this specific topic, try watching a few "socially conscious" 1980s movies that wear their values on their sleeve; they're recent enough that you expect their values to be the same as yours, but they're not. (The same argument can be made for why I am annoyed by the characters' values as they relate to gender roles in Bronte and Austen novels in a way that I am not by, say, Chaucer.) But the pro-gay attitude shift is just a canary in the coal-mine. In the early 90s, suddenly kids were growing their hair long on masse again and wearing lots of black and talking about conformity, there was loud angry music and grunge everywhere, and pot use skyrocketed. Coincidence? Or the coordinated coming of age of the first post-formal generations' kids?

There's no one archaeopteryx for the 1990 shift, which in any event wasn't quite as dramatic as 1965, but here are a few: 1989 had Batman, which celebrated "dark" (new to mainstream American film audiences); 1990's Dances with Wolves had the first naturalistic and positive treatment of Native Americans (imagine that in 1985!) and in 1991, Terminator 2 shows us the badly-behaved punk kid (the young John Connor) whose criminal sensibilities end up serving him well. Imagine a pubescent criminal as protagonist and hero on the mainstream big screen even in 1988. Musically, in rock, late 80s thrash (underground, no airtime and little MTV exposure) gave way fully to grunge by 1992 (on MTV you couldn't get away from it).

Is It a 25-Year Cycle?

If the pattern is real, then we're due in 2015 for another shift. But I have my doubts that it will remain cohesive. The use of mass media as milestones is becoming potentially problematic, because the way we consume media (and create it) has changed so much. On the other hand, it's the spread of values that shapes these shifts, and thanks to technology that process has never moved so quickly. Unfortunately I can't make a prediction because I don't have a sense of which values will carry over from the early 90s and dominate the new zeitgeist, just as it would have been difficult in 1989 to make the same kind of call. Check back in 2017; by then such a shift should be obvious.

[Added later: Razib Khan separately notes an inflection point also in 1990 for another sexual more, black-white dating.]

High Growth Rates: Nature Abhors a Vacuum

When confronted with China's recent brilliant growth rates, a cynic might say China had an unfair advantage: it had room to grow. That is, it's easy to grow your GDP at 6.46% annually since 1980 if you start out with a per capita income of US$305.46. Labor is cheap, you have no legacy infrastructure to deal with, and your exports are extremely competitive. Of course, this begs the question that there are lots of countries with low PCI, and not all of them grow at such robust rates - but let's come back to that. I was also once challenged by a European that the U.S. grew slightly faster than Europe not because of any decision we've made to embrace free markets, but because of our good land and wide open spaces which are cheaper to develop. Because of population density rather than PCI, we have room to grow.

This interested me, so I pulled together some IMF and UN figures for 179 countries and territories; most growth rates are annualized since 1980. First let's look at population density's relationship to income growth, if any (source for population and area data here and here resp.) For viewability purposes, the scatter plot below excludes the 11 most densely populated countries/territories (Singapore, Hong Kong, Malta, Bangladesh, Bahrain, Maldives, Taiwan, Mauritius, Barbados, Korea, and Lebanon, all > 400 people/km^2).



The red circle contains 18 countries, all of which have had at least 10% annual growth since 1992: Armenia, Kazakhstan, Estonia, Latvia, Lithuania, Turkmenistan, Bosnia and Herzegovina, Equatorial Guinea, Russia, Azerbaijan, Belarus, Tajikistan, Cambodia, Ukraine, Slovakia, Moldova, Czech Republic, and Croatia (Bosnia-Herzegovina data since 1994, Equatorial Guinea data since 1980, Cambodia data since 1986). There's a trend there: of these 18 countries, fully 17 transitioned from a closed communist in the last decade; 16 of 18 were Soviet satellites. The trend on display is the effect of markets, not low population density. Not the effect I was looking to call out, but interesting that it's so apparent here.

It's worth pointing out that, for the 12 countries that grew at greater than 15%, the average density was 45/km^2; for the rest that grew at less than 15%, the average population density was 218/km^2. The same statistics using 10% as the break point are 62/km^2 for 18 countries >10% growth and 222 for the rest. Breaking the other way, the 58 countries with greater than 120 people/km^2 density grew at 4.15%; the rest that have less than 120 people/km^2 grew at 4.34%. There does seem to be some effect. (These figures include the 11 densest countries taken out of the scatter plot.)

The picture for growth rate and PCI could fool you into thinking it's some sort of normal distribution, but it's not. PCI is taken from 1999 for all countries because it was the first year that IMF had data for all 179 countries.



Interestingly, the vast majority of high-PCI countries have a middle-road growth rate of around 5%. Low PCI countries are more widely distributed. The 18 countries with growth above 10% have average PCI of US$1,836.08; those with growth less than 10%, US$6,505.82. Then again, the 32 countries with negative growth rates clocked in at average PCI US$1,703.99, vs. positive growers at US$6,979.31. Breaking the other way, those with 1999 PCI below US$5,000 had average growth of 4.17% (131 countries), while those above US$5,000 PCI had average growth of 4.55%. Of course, again the low PCI, high growth countries were all ex-communist but one. Who are the low PCI low-growth countries (i.e. < $5,000 PCI and negative growth)? Georgia, Congo-Zaire, Ghana, Mongolia, Yemen, Niger, Sierra Leone, Burundi, Eritrea, Madagascar, Papua New Guinea, Nigeria, Ivory Coast, Gambia, Solomon Islands, Namibia, Zambia, Myanmar, Malawi, Paraguay, Syria, Togo, Uganda, Ethiopia, Suriname, Guyana, Rwanda, and Guinea. This is a grab-bag, but many of the countries were victims of civil wars (10 of 28) and a few resource-cursed ones.

There is a weak inverse correlation between growth and both population density and per capita income, although it is swamped by the signal from the post-communist states. The lesson here? Those states were left with strong institutions, which visibly benefit them (particularly in the case of the growth-PCI comparison). So perhaps it's true: China's low initial PCI and its strong institutions, as well as the U.S.'s open spaces, are both advantages to growth.

Mexico's Resource Curse: The United States

"¡Pobre Mexico! ¡Tan lejos de Dios y tan cerca de los Estados Unidos!"

– Porfirio Diaz


To a developing country, a long border with a wealthy, industrialized neighbor might seem like a blessing. But we have at least one pairing where this is anything but obvious: Mexico and the United States.

There are other cases where a national asset that seems on its face to be a big advantage turns out to be anything but; the famous example is the resource curse. You would think that a developing country fortunate enough to be sitting on mineral wealth would be able to use that wealth to its advantage – especially if it's oil. Nigeria is the textbook case.

These countries experience a vicious cycle of incredibly corrupt juntas uninterested in developing other industries or indeed doing anything except pocketing the proceeds from mineral extraction being conducted by foreign companies. In these institutionless post-colonial kleptocracies, the only options for the ambitious are to get out, or try to get in on the next coup. It's hard to believe that these governments could keep the figurative lights on for even one minute after the oil and diamonds stopped coming out of the ground. These are failed states with an allowance. They're Somalias waiting to happen.

What started me thinking about this was the recent arguments I've seen from several quarters that Mexico is increasingly showing alarming characteristics of a failed state. Combine this speculation with the interesting observation that the U.S.-Mexican border is easily the longest one in the world between a developing and an industrialized country, and you may begin to wonder if there's a connection. If Mexico had started off with a decent-sized middle class and relatively transparent institutions, they may have joined in the ongoing growth that the Anglophone parts of the continent have enjoyed since the industrial revolution.

Among the generally agreed upon characteristics of failed states are these: economic decline; loss of control of and security in their territory; deterioration of basic services; inability to make and execute decisions, in both domestic and international arenas. I don't think Mexico is there yet. It has elections, the lights shine and the toilets flush, and it's a productive member of the international community. It's even tied for 72 out of 180 in Transparency International's 2008 index – not great, but pretty good for a supposedly failing state. Still, the increasingly brazen coordinated paramilitary attacks on the police are a bad sign. Mexico is, in fact, losing control of and security in large stretches of its territory. Where? The states closest to the U.S. To whom? Drug cartels. Coincidence?

The truth is that Mexico is resource-cursed, and the resource is drug-consuming Americans. To be more accurate, the resource is drugs and the market is the U.S., but the situation is in some ways worse than in Africa's resource-cursed states. Imagine that Nigeria had a long land border with the EU. Now imagine that oil is outlawed as a result of global-warming legislation. The oil would still flow north - but it would become contraband, and the trade would be an entirely criminal activity. Because the drug trade is internationally sanctioned, the Mexican government (unlike Nigeria's) can't openly profit from the trade as it does with legal oil in Nigeria - because even graft and kickbacks are parasitized from activities that are at least legal in and of themselves. So the business becomes the domain of paramilitary drug cartels and some corrupt officials that allow them to flourish. It's worth pointing out that, although it doesn't border the United States, Colombia is the other perilously-close-to-failing state in Latin America, though it's improved in recent years. Still, it's had large tracts of its territory not under its control for years on end – and those tracts were controlled by organizations in the same trade as the paramilitary groups in Mexico. And they had the same end market.

A reasonable objection is that there are income disparities across other borders in the world; surely Mexico and the U.S. aren't the only odd couple, yet there are no paramilitary drug groups forming elsewhere. I suspected there were reasons why this didn't happen elsewhere, so I compiled a list of 279 sets of two-country shared land borders, and ordered it in terms of absolute nominal per capita income difference. Out of 279, here are the top ten:

Rank	Country1	Country2	PCI1	PCI2	PCI Diff
1	Austria	Liechtenstein	50,098	145,734	95,636
2	Norway	Russia	95,062	11,807	83,255
3	Switzerland	Liechtenstein	67,385	145,734	78,349
4	Saudi Arabia	Qatar	19,345	93,204	73,859
5	Finland	Norway	51,989	95,062	43,073
6	Iraq	Kuwait	2,989	45,920	42,931
7	Norway	Sweden	95,062	52,790	42,272
8	Finland	Russia	51,989	11,807	40,182
9	USA	Mexico	46,859	10,235	36,624
10	UAE	Oman	54,607	18,988	35,619

Source: IMF World Economic Outlook Database 2009, except Liechtenstein from CIA World Factbook April 2009. Border ranking process did not include exclaves (e.g. Ceuta, Kaliningrad)


It's immediately interesting that the U.S.-Mexico border pops up as 9th out of 279, and is one of the longest on this top-ten list. Three qualifiers in order here. a) There is less incentive to engage in risky activities if basic needs are met. An Austrian might know his neighbors are wealthier, but it's different when your PCI is over US$50,000 as opposed to just over US$10,000. If you're comfortable, you're probably less likely to consider running drugs to Liechtenstein. b) Not all borders are as easy to cross (legally or otherwise) as the one between the U.S. and Mexico. Many countries don't have the same freedom of movement (like Russia), many countries don't have as well-developed a road system as the U.S. or Mexico, and even if a border is about as long as the U.S. Mexican border (as in Finland and Russia), it may be even less hospitable than the often-mild desert between the U.S. and Mexico. c) Many of the extremes of per capita income reported by the IMF would be flattened if a median calculation were used instead of a mean. Few subjects of the UAE come close to the reported per capita incomes listed here.

What's the solution? I don't anticipate Americans' consumption of drugs will stop any time soon, nor will Mexicans' willingness to supply them; after all, markets are markets. The part of the equation we can control is a choice that we've made which forces the profits from the drug trade underground. That is to say, if the United States decriminalizes, suddenly Mexico's unique resource curse can at least benefit Mexicans and their institutions openly. Sounds like a pie-in-the-sky solution, right? Wrong. One country – Portugal, a civilized EU country no less – has already done exactly this, and "judging by every metric, decriminalization in Portugal has been a resounding success."