Sunday, July 24, 2011

Shanghai housing density

This morning I went for a walk in Heping Park, which is on the border between the Hongkou and Yangpu districts and about a mile from where we are staying. To get there, rather than taking the subway to the Siping Road stop, I walked along Fuxin Road, which runs through a residential neighborhood which was probably built in the 1950's or early 1960's when Shanghai's population was skyrocketing.




The park itself is about 240 acres and was built in 1958, at the same time as Yangpu Park two miles or so to its east. It was crammed with people. The occupations were the usual: old men chatting and fanning themselves, others airing their songbirds, people doing solo or group Taiqi and exercises, group dances, people practicing instruments or singing, mothers or grandmothers with children, and people just sitting around. Almost every potential seat on benches, walls, stones, and elsewhere was occupied. I only walked around half of it, and informally counted thousands of people. My guess is that there were well over 5,000, possible twice that. The sidewalks around the gates were so dense with parked bicycles that pedestrians had to walk in the street.

The reason is not hard to guess -- Shanghai in late July is sweltering, even at 8:00 in the morning, and most of the apartments surrounding the park do not have air conditioning. While those along the main roads are modern high-rises, those behind them are brick and concrete apartments blocks from 60 or so years ago. I wondered, how many people did the park serve?

Shanghai's overall population density is between 2700 and 3700 people per square kilometer depending whose statistics one uses. But much of this is not densely populated. The Hongkou district where the park is has an average population density of 34,000 people per square kilometer. The downtown areas of the old foreign concessions, where much of the old Shikumen housing had densities of 7000 people per block have an average density of around 50,000 people per square kilometer. Looking at the mass of apartments which filled the blocks near the park, that did not seem high.

Some of the newer apartment blocks had window air conditioners, so I used them to count the probable number of apartments per floor. 12 per side, so 24 per floor. The buildings have between 5 and 7 floors, so that means 120 to 168 apartments per building. Shanghai has an average occupation rate of a bit over 3 people per apartment, so each of the buildings would house on average 360 to 500 people. Assuming that 6 stories is the average, that means about 432 per building. I couldn't count the number of apartments per block due to the density. So I went to Google Maps and used the satellite view. The neighborhoods through which I was walking averaged 126 apartment buildings per square half kilometer! That gives a density of 226,800 people per square kilometer. Subtract a bit for streets and shops, and there are probably about 225,000.

That makes sense given the overall data. Dense housing blocks make up only a fraction of the area (if nothing else, the universities, parks, business and public buildings need lots of space), but the housing areas are dense indeed. Given that the apartment blocks are at least twice the height of the old Shikumen slums, the density should be comparable even with more space per person. But not much -- I gather the apartments average no more than 500 square feet.

At any rate, given that there are several such dense housing areas on each side of the park, the catchment area has a population of over a million people. Five or ten thousand people going to the park on a stifling day would not be surprising.

Wednesday, July 20, 2011

Is it better to learn Chinese in Beijing or in Shanghai?

Before the second world war, a foreigner learning Chinese had a series of difficult decisions to make. The few American (or European) universities which taught Chinese taught only the classical language (like Latin to modern French or Spanish speakers) -- the textbooks disparagingly noted that the student who went to China would pick up the spoken language on the street once he got there (the exclusive male gender is in my father's 1930's textbook from which I took this comment). But the language on the street was anything but standardized. Until the late 1920's, there was effectively no national standard language -- the Republic of China official standard called 國語 was an artificial creation which no one spoke. Starting in the late 1920's, the Beijing dialect was adapted as standard, but outside of the Beijing area, only a few educated people spoke it. Many more spoke some regional variety of what we now call Mandarin, but these dialects were not spoken in any of the prosperous trading ports. In places like Shanghai, Ningbo, Fujian, Guangzhou and Hunan, ordinary people spoke "dialects" as different from Mandarin as French is from Spanish; educated people spoke Mandarin with a heavy regional accent. Anyone listening to recorded speeches by Sun Yatsen, Chiang Kaishek and Mao Zedong finds their Cantonese, Zhejiang, and Hunan accents almost incomprehensible. So if one learned Chinese, he learned the dialect of the area where he worked and lived. The very few students who went to China for academic reasons rather than to work learned either Hunan-accented Mandarin at the Yale in China program or Beijing dialect at one of the programs there.

After 1949, things became less complicated, simply because there were so few choices. With American students banned from mainland China, the only alternatives in the 1950's were Yale in China's Hong Kong program or the Mandarin Training Center at National Taiwan Normal University (disclosure: I am a graduate of the latter). Starting in the early 1960's, the Inter-university program in Taipei usually known as the Stanford Center became the preferred alternative. From the standpoint of total immersion in the spoken language, none of these programs were ideal. In Hong Kong, the language of the street was Cantonese (with quite a bit of Shanghaiese in those days), and in Taipei most people's first language was Fujianese, and most Mandarin speakers had learned it as their second Chinese dialect. So American students who studied in those programs tended to come back either with a noticeable Cantonese accent or a Fujianese one. In my case, the University of Michigan instructor of record for my Chinese classes had studied in Hong Kong and was married to a Cantonese, while my native-speaker spoken lab teacher was from Shanghai. The family I lived with during my first year as a student in Taipei was also from Shanghai, and so to this day in spite of frequent residence in Beijing, I speak Mandarin with a Shanghaiese accent (which turns out to be useful now that I am living here).

Today students can study Chinese anywhere in China, and Chinese universities from all over China actively recruit American language students. There is a nationally-standardized language (not quite the same as the Taiwan 國語 which I learned), Chinese students everywhere speak an approximation of the national standard, so it should make no difference. There are, however, some things to consider: Mandarin regionalisms and accent.

There are quite a few vocabulary and grammatical differences between northern and southern Mandarin. A student in Beijing (and to a lessor extent in Sichuan) will hear and learn the northern versions; a student in Shanghai or Nanjing will hear and learn the southern version from both teachers and other students.

In terms of accent, an American student who studies in China will pick up speech habits from three sources. First, Chinese language teachers. Any Chinese university Chinese language teacher will probably pronounce individual syllables in perfect Beijing standard. However, sentence intonation varies with regional dialects -- a native Shanghaiese speaker will combine Beijing syllables with Shanghaiese sentence intonation (that is the way I speak). Second, fellow Chinese students. While major universities draw students from all over China, the number of local students will be much higher. Third, people on the street. In Beijing, they will speak with a strong non-standard Beijing local accent. A downside of studying Mandarin in Shanghai is that much of what one hears in public is Shanghaiese and not Mandarin. The locals will normally speak Shanghaiese with each other, but will speak accented Mandarin to out-of-towners and foreigners. The bottom line is that those who study in Beijing will pick up a strong Beijing accent, and those who study in Shanghai will pick up a mild Shanghai accent. Those who study elsewhere in China will likewise pick up some of the local accent, vocabulary and grammar.

Of course, most Westerners speak Chinese so badly that it doesn't matter -- the American, Australian, or German accent will drown out any regional Chinese. But let's confine ourselves to the minority with good ears, flexible tongues, and who take the time and effort to mimic and practice and learn to speak well. Does it matter what regional accent you pick up?

A strong local Beijing accent is no problem -- the hyperlocal Beijingisms will disappear if you live elsewhere in China and leave a perfectly standard accent. Since that is the language of the official media, everyone can understand it. The one downside is that a Beijing accent provokes resentment in some parts of south China. This probably applies less to foreigners than to native Chinese speakers.

Since Shanghai is the business capital of China, Mandarin with a Shanghai accent will probably do no harm for a business career. Or any other, given the large number of national political officials who come from Shanghai (there have at times been political jokes about the Politburo conducting its business in Shanghaiese). Since the southern variety of Mandarin is the version spoken in all of the prosperous coastal areas south of the Yangzi River (including Taiwan), speaking that variety helps fit in in most of the economically-advanced areas. This would also argue for learning Mandarin in Nanjing.

It would, however, probably be best to avoid Chinese language study in other areas of China where the local dialect is not a form of Mandarin (like Zhejiang, Fujian, Guangzhou and Hunan). More controversially, there are Mandarin-speaking areas where the accent is looked down upon in the rest of China (Anhui and Shandong for example). A Sichuan accept would probably give one a leg up in much of western China, which is developing rapidly.

So where to study? If I were 19 again, I would spend my junior year in Beijing. Then I would come to Shanghai for my MBA.

Tuesday, July 19, 2011

Real estate prices in low-corner Shanghai

Most discussions of the soaring real estate prices in China (and whether or not there is an incipient real estate bubble) concentrate on the expensive housing developments where the Chinese upper middle class and foreign expats live. In Shanghai, that would be the former foreign concessions and the Changning and Gubei districts to the west of them in Puxi (the older section of Shanghai west of the river) as well as the more trendy areas of Pudong (the new city east of the river). The sections west of the river (and south of Suzhou Creek) are the area referred to in Shanghaiese as "high corner" -- the "good" areas to live.

I want to talk about real estate prices in "low corner" Shanghai -- the wrong side of the tracks as it were. Specifically, a neighborhood dating from the 1950's in Yangpu district in the Northeast.

The area north of the Huangpu River and east of Suzhou Creek is the Hongkou district. In the mid-19th century it was the American concession, and with the formation of the international concession in the 1860's, it gradually became the defacto Japanese residential area. Much of it (along with the adjacent slums of Zhabei north of Suzhou Creek) was destroyed during the first invasion of Shanghai by the Japanese in the 1930's. It became the Jewish ghetto after 1941, and then became an industrial area after the Revolution. Most of the housing is concrete block work-unit apartments built during the 1950's and 1960's. Joseph Gamble in his book Shanghai In Transition : Changing Perspectives And Social Contours Of A Chinese Metropolis devotes much attention to it, since his wife's family lived there, and he lived with them for some time while doing his social research.

The Yangpu district is directly north of Hangkou. It became part of the city only in the 1950's, when the campuses of many of the newly reorganized universities were moved there. The southern part of the district, which I will be talking about, was built up with the usual 5-6 floor concrete worker housing, features a large park (Yangpu Park) and is served by subway line 8. Over the last 15 years, there have been hundreds of new high-rise apartment (condo) towers built there. Chinese condo complexes seem to have problems with maintenance, and so those towers older than a few years definitely show their age.
Older housing block

New highrise housing

The condo complexes which I priced are located between Kangjiang Road and Benxi Road east of Jiangpu Road towards Yangpu Park. They are within a few blocks of a subway stop, and 8 stops (about 20 minutes) from downtown. They are in fairly new buildings (illustrated) and are almost all between 625 square feet and 1050 square feet. It should be noted that Chinese apartment areas include some of what would be considered public space in the US, so quoted figures should be reduced by about 5-10% to make them strictly equivalent. Ignoring this detail, the average price per square foot of the apartments in this area is US$377 at the official rate of exchange.

One of the condo blocks priced


Another of the condo blocks priced

This is not a neighborhood where people live cosmopolitan global lifestyles. On the street, almost everyone speaks Shanghaiese and not Mandarin -- some of the older people appear not to be able to speak Mandarin at all. Many people still go about in their pajamas, and some men in their underwear. Clothes are anything but trendy -- many of the older men go around in shorts and singlets, and few outfits can have cost more than US$15. The shops sell only the basics. There are lots of bicycles and few cars; the cars are mostly older VW's or low-end Chinese brands. The only foreign face I have seen there is my wife's. The trendy brand-conscious crowds of high corner Shanghai are a world away.

In other words, US$377 per square foot for an apartment in this neighborhood on the outskirts of Shanghai is an incredible amount of money. Ignoring purchase power parity, the condos here are priced the same as in the most trendy areas of Portland, and twice what you would pay in the far outskirts of town. Given Chinese middle-class incomes and other prices, a small (650-750 square foot) apartment for US$300,000 is equivalent to over a million dollars in the US. And I suspect that few of the people in the neighborhood in question have middle-class incomes.

Thursday, July 14, 2011

The real cost of living in Shanghai

One of the problems in comparing the cost of living in different countries is that even if exchange rates are not manipulated, exchange rate equilibrium between two countries represents only tradable assets, which are in theory subject to arbitrage. Thus, in the long run the price of an iPhone purchased without a contract in the US with US dollars and an identical iPhone purchased in China with Yuan should differ only based on value added and other taxes and transportation costs. Otherwise, cheaper iPhones in the US will be purchased and shipped to China, bringing the two markets into equilibrium (something which Apple tries to prevent of course).

The problem is that most goods and services are not tradable. While in theory a big mac should cost the same in both places, one cannot buy cheaper big macs in Portland and resell them in Shanghai. The same with Starbucks lattes, which are also used for a tongue-in-cheek price index. Similarly, real estate, haircuts, and karaoke. So what is the real cost of living in Shanghai?

The official figures are based on comparable market baskets of goods and services, in the same way as the consumer price index. The problem is, a middle class resident of the US will have a different mix of goods and services than his/her counterpart in Shanghai. What Shanghai market basket should be used to compare price levels?

An identical top-end global lifestyle costs about the same, perhaps more in Shanghai because of value added taxes and inflated real estate prices. If you buy premium western brands, eat western food, eat in western restaurants and bars, drink French wine, drive a German car, and own a condo in a fashionable neighborhood, your cost of living in Shanghai will be about the same as in New York.

If you live an upper middle class westernized lifestyle but adjust for local products and food, Shanghai is cheaper. But how much? Many Chinese argue that the cost of living in Beijing or Shanghai is about 1/6 that of the US, but that is probably due to confusing purchase power parity with currency exchange rates. Based on salary levels and equivalent market baskets, 1/3 might be more accurate. The International Monetary Fund estimate is that in terms of purchasing power parity, one US dollar is worth 3.8 Yuan, which would make the cost of living here somewhat more than half the US level. My personal estimate is that for a childless Chinese-speaking couple who rent and have no car, the cost of living in Shanghai is about half what it is in Portland for an equivalent lifestyle. If it weren’t for rent (which is about half the cost of a Portland equivalent and dominates the market basket due to the amount), the Shanghai cost would be closer to 33% of the Portland cost.

Food is much cheaper (if you don’t buy too many imported products), clothing is about half if you buy (and can fit into) Asian brands, medical costs are much lower, public transportation is much lower, but cars, and real estate are much higher. Cell phone usage is much cheaper; the phones themselves are much cheaper for low-end models and more expensive for top-of-the-line models (with plenty of people willing to pay a substantial premium for an iPhone or a Galaxy S). Rent in Shanghai is about half what it would be in Portland (and thus probably a quarter of the non-rent-controlled price for NYC). Chinese restaurants are quite cheap, but eating out in Western style restaurants is more expensive for what you get. Imported wines are very expensive; domestic wines and beer are cheap.

There are confounding issues. One is family costs. Chinese families need to pay for children’s educational expenses, and to support their parents. American families (at least among our friends) are much more likely not to have children, and are not expected to support their parents. On the other hand, US taxes are much higher than Chinese ones. Finally, there is the issue of face and conspicuous consumption – living below your means (at least in public) is just not done here.

Wednesday, July 13, 2011

Parallel economies in Shanghai and purchase power parity

One reads quite a bit about the disparities in economic development in China between the coasts and the interior, and about the rising income disparities between those who have benefited from the new economy and those who have not. One thing that doesn’t really show up in the statistics are the parallel economies in Shanghai (and Beijing, and probably most other large cities) between people living a fairly traditional Chinese lifestyle and those living a cosmopolitan global lifestyle. They are almost in different countries.

We are living in “low corner” Shanghai (the wrong side of the tracks) since many of the universities were moved to the far northeast Yangpu district after the revolution. This is a very high density neighborhood, made up mostly of crumbling 5-6 story concrete block apartments which usually have no air conditioning or heat. Many of the local people are elderly; I suspect that most of them were moved here after their communal apartments towards the city center were torn down and redeveloped. There are many more rusting bicycles than Mercedes. With the new line-10 subway line and a huge shopping and entertainment complex near here, the neighborhood will probably be changing quite a bit over the next few years.

A block or two around the corner there is a local supermarket, a wet-market, and lots of food stalls and stores. There are quite a few small restaurants in the area, along with clothing shops, etc. The economics of living a Chinese lifestyle here are orders of magnitude from the Bund and the former French concession – I estimate about a 1:50 difference!

Lunch for two from a neighborhood food stall consistently costs the two of us under 5 Yuan (US 75 cents). Dinner for two with beer at a modest Chinese restaurant costs under 40 Yuan. By comparison, two lattes at one of the hundreds of packed local Starbucks cost 50 Yuan; a western-style lunch for two with coffee at a trendy restaurant in the former French concession costs almost 300. A complete suit of traditional women’s clothing at a neighborhood shop costs 39 Yuan (US$6); a Chinese-made polo shirt from the crowded Uniqlo shop at the nearby mall costs 149 Yuan ($23); Asian-brand slacks at the mall are 150-250 Yuan. For those who prefer Western designer brands, they are a bit more expensive than in the US due to value added tax, and sell like hotcakes.

In short, there are parallel economies here which have little if anything in common and confuse the meaning of the economic statistics, especially when it comes to purchase power parity. Someone who eats traditional plain food, wears traditional clothes, lives in an older-style apartment, and drinks local beer and ordinary tea has a cost of living about one tenth that of a middle class consumer living a cosmopolitan lifestyle in a modern condo in an outer district who shops at mid-market Chinese or western/Japanese chains. Add a condo in an up-market neighborhood (high corner Puxi or a fashionable area of Pudong where real estate costs are at NYC levels), a western brand of car and designer clothes and accessories, and the cost of living jumps by over 500% above the “ordinary” middle class.

The question then is: when calculating purchase power parity for life in China, what life do we choose? Probably not our many neighbors who appear to live on a thousand or two US dollars per person per year. On the other hand, for those expats who want to live exactly as they did in the US or Europe (and their Chinese counterparts who want to do the same), the cost of living in Shanghai is about the same as in New York City or San Francisco, so there is no purchase power parity adjustment. One dilemma is that choosing either the “ordinary” or the high-end middle class lifestyle makes the economic statistics corrected for purchase power parity seriously understate the quality of life of the majority of the population, for whom the correction would be much greater.

Shikumen (石库门) houses in Shanghai

Shikumen () houses are a fusion of Chinese and European design unique to Shanghai. They were the standard type of housing in Shanghai from the 1860's through the 1930's. Originally they were built to house well-off Chinese refugees fleeing from the Small Swords and Taiping Rebellions in the 19th century.  As more and more Chinese moved to the former foreign concessions in Shanghai, tens of thousands of them were built by real estate investors.  They generally took up the entire interior of a city block, with access by alleys called Longtong.  The alleys generally had elaborate stone gates () from which the housing type took its name.

Originally, a single family lived in a townhouse with two floors and 8 or more rooms depending on the size.   The smaller ones are around 1000 square feet, with a parlor, two other rooms and a kitchen on the ground floor, a mezzanine room above the kitchen (originally for a servant), and three rooms on the second floor.  Often an attic floor was added for additional servant rooms.  There was no piped water or sewage, both of which were communal facilities in the alley.

Under the legal arrangements between the foreign powers and the Qing Dynasty, only foreign residents could own real property in the concessions.  Since the tenants were Chinese, the owners were not able to deal with them directly, but hired Chinese partners who handled all dealings with the tenants, and turned over an agreed amount to the owners.  With the high turnover of foreign residents in Shanghai, most of the actual owners lost all connection with the buildings which they owned other than the cash flows from rent.  That left the Chinese property managers as the effective owners, and led to two developments, the effects of which are still present.

The first was extensive subdivision of the townhouses.   With the flood of internal immigrants coming to Shanghai, it turned out to be much more profitable to rent out each floor, and then each room as a separate apartment.  The servants’ rooms on the mezzanine and in the attic were rented out as cheaper one-room apartments, and often extensions were built on the roofs as well.  Thus the Shikumen blocks became (and remain) very high density housing with 7000 people or so per city block.  The second impact was the neglect of maintenance to maximize short-term cash flow, which meant that much of the housing (which had been built as rapidly as possible) became quite run-down.

Because of continued housing shortages after the second world war and because of the diversion of resources to industry during the 1950’s and 60’s, this housing continued to be used. New housing in the areas destroyed by the Japanese or converted from farm land as the city expanded was in the form of concrete block work-unit apartments, but much of the population of Shanghai continued to live in Shikumen apartments without plumbing or heat until the 1980’s.  Since that time, much of the old housing stock has been torn down and replaced by modern high-rise buildings.  Many of the people living in the old housing were moved to newer apartments in the outskirts of the city.  Preservationists (who are often not Chinese) lament the destruction of tradition and the disruption of neighborhoods.  Many of the former residents were happy to get indoor plumbing and more space in the new apartments.

Today, some of the surviving Shikumen housing in the foreign concessions is being converted into expensive modern townhouses.  Sometimes this is done by real estate developers, and sometimes by individuals (we have friends who have done so).  However, many of the old houses were build so cheaply in the first place and have become so run-down after decades of deferred maintenance that they cannot reasonable be salvaged.  In many cases, adding modern utilities is extremely difficult.  Most of the remaining Shikumen neighborhoods can best be described as slums.