I think you are right concerning urban environments. We have plenty of examples of societies where the urban environment is very clannish. How well you do is closely tied to how well your extended family does and you also select mates from that family.
To quote from pg. 84 of The Natural history of Inbreeding and Outbreeding:
By definition, altruism is selected against by within-family selection (individual selection) and favored by between-family selection, since the average fitness of a family is directly proportional to the frequency of altruists within the family. Indeed, it can be shown that Hamilton’s rule translates into the condition in which between-family selection overrides within-family selection (Wade 1980)….”
“Breden and Wade pointed out that inbreeding eventually fixes the population for two family types resulting from matings between like homozygotes (AA x AA, aa x aa). Matings between like homozygotes result in families that have no genetic variation in them. Consequently, they concluded that inbreeding generally decreases the average within-family variance. The two types of families produced by matings between like homozygotes are also maximally different from each other in frequency of the altruistic allele and in average fitness. Consequently, inbreeding increases the between-family variance. Both these factors, that is, the decrease of within-family selection and the increase of between-family selection, have the effect of making the evolution of altruism easier. This effect was also emphasized by Boorman and Levitt (1980: 350). They concluded that ‘…a high degree of inbreeding may be expected to favor sib altruism over the random mating case….’
But societies that are very inbreed don't seem to be very "altruistic" in the way the word is used on LW. They seem to be altruistic towards family members but not very much to random strangers. Or it might just be that the well known effect of inbreeding depressing IQs might make their societies more impoverished and thus leaves them less able to help random strangers.
This brings me to three interesting interesting predictions one might make. All else being equal:
1) Societies with a longer history of faster travel tend to be more indiscriminately altruistic (since that presumably means more out-breeding).
2) Societies where the cost of travel over long distances is low will have less inbreeding.
3) Subcultures that maintain high inbreeding despite exposure to 1) and 2) out-compete those groups in a society that don't, because they enjoy the fitness boost of indiscriminate altruists while being highly selective about who is the target of their own altruism.
I do not have a high confidence in these but they seem to make sense.
We may feel sympathy when we read about people killed for protesting in Syria, Bahrain, Libya, and other countries. But tyranny isn't just something happening to unfortunate people somewhere else. It's an existential risk to human civilization.
Civilization - even tribalism - relies on altruism. Altruism is defined as cooperation that is not the happy convergence of interests of rational self-interested agents. That happens too; but we don't call it altruism. Altruism is, roughly, helping others without the expectation of reciprocation or cooperation. And it happens because humans like helping other humans.
Altruism is probably mostly genetic. It's an evolutionary adaptation that instills the desire to help others into a species. Social pressure can install some amount of altruism; but it's my opinion that this would not work at all without a pre-existing genetic basis. Many species exhibit altruism to a level at least as great as that in humans. Some insects, which are incapable of feeling social pressure, are far more altruistic than humans.
Two theories for how this happens are kin selection and group selection. Regardless of which of these you prefer, both of them have two important weaknesses:
It's not known whether humans are still evolving, or have begun devolving due to lack of selective pressure. But in the case of altruism, we can be sure: Even if some selective pressure still exists, most humans today do not live under the necessary conditions for either kin selection or group selection. Humans are living off their evolutionary capital of altruism.
Tyranny, whether it's that of Syria, Iran, North Korea, Nazi Germany, or the Soviet bloc under Stalin, aggressively selects against altruism. The most-altruistic people were among the first executed in all those places. They are the people being shot while protesting in Syria. Social activism under such a government is rarely in your best self-interest. Tyranny selects for self-interest; people who are willing to help the state oppress others are given opportunities for advancement. And it removes altruistic genes quickly from the population, likely undoing hundreds of years of evolution every year. Those genes will never be replaced.
I'm not too worried when this occurs over a few short months or years. But when a people lives under these conditions for generations, you may end up with a large population deficient in altruistic genes.
There's no solution at that point short of gene therapy. The population can stay in place, resulting in a society that is at best hopelessly mired in corruption and poverty, and at worst a danger to the rest of the world. Or it can disperse, and dilute altruistic genes around the globe.
ADDED: Knowing whether this is a real problem or not, would require learning something about how many genes are involved in altruism, and what their distribution in the population is. A legitimate objection to what I wrote is that if genes for altruism are distributed so that killing less than 1% of the population would have a major impact on their abundance, then they probably weren't very important to begin with. Although, sociopaths are only around 1% of the population, and they have a major impact on society. I wonder how much work has been done in studying the maintenance of alleles for which you only need a few members of the population to have them?