Want a business license？The xian will assign some-one to do the walking and talking for you.Want a build-ing permit？They will give you one with a money-backguarantee.Unhappy about that dirty creek passing through the site？They may offer to build a small lake for you.They will help you find architects，findbuilders，and coming to the production phase，they will help you recruit workers for a reasonable fee.Yes，xian shave worker-recruiting teams，which do the hiring forclients.They sell their cheap electricity，sell their parksand entertainment，sell their easy transportation，sell theirwater supply，sell their glorious history，and sell evenhowgood-looking their girls are——no exaggeration!
Competition among localities of a level comparable to that among xians in China is unheard of.Why is this so？One contributing factor is the sharing formula.Let usexplain this formula before turning to other factors in thenext section.
As we have seen，investors coming to a xian would have to pay 17% in value-added tax.Of this the xian isentitled to a cut of 25% ，or 4.25% of valued-added.From the proceeds fromland sales（if positive）75% goesto the xian and 25% to higher levels.This latter percent-age is not uniformfor all xians: the better a xian is locat-ed the lower will be its share of income fromland sales.Of the xian officials I talked to，none of themcared howthe higher levels divide their take.
Land costs are substantial.These consist of two parts.First peasants must be paid to give up their land.Using a5% rate of interest，I estimated in 2006 that such pay-ments come to about 3 - 5 times the discounted marketrental of the land when used for farming.Quarrelsbetween xian officials and peasants occasionallyoccurred，but far less frequently than reported in thepress.It is true that some officials take secret cuts in between，and that a xian short of funds may owe thepeasants for a long period of time.
The costs of developing or improving agricultural land for industrial or business use are even more substantial.In 2006，these costs were in the neighborhood of RMB60,000 per mu（660 square meters），about twice the com-pensation to peasants.Improvements include buildingroads，putting in power，water，gas，and sewers，cablesfor telephones，television and computers，street lighting,and landscaping.These improvements are made beforethe land is sold.Today，even in lower level areas，thequality of newindustrial real estate offered by the xiansis higher than what I was familiar with in the State ofWashington.At the top level，like those in the Suzhouindustrial village，it is the best I have ever seen.Chinesepeasants are superb in planting things——they oftentransplant trees over fifty years old——and you cannotbeat what Arthur Lewis would say is an “unlimited sup-ply” of themworking for landscaping at US 5.00 eachfor one long day.37
【37】This was the figure in 2004，in the middle of the fastest risein peasant income in the history of China.It was a daily wage,and by 2007 the figure had risen to about US8.50.The amountvaries in different areas，and I took rough averages when I touredvillages in China on photography trips.From2003 to 2005，Iproduced enough photographs to publish seven books，so what I say about the livelihood of peasants in China，though deviatessharply fromall other reports，is based on prolonged and inten-sive investigations conducted on site，in exactly the same way asI investigated apples and bees in the State of Washington in1972.See Steven N.S.Cheung,“The Fable of the Bees: AnEconomic Investigation,” Journal of Lawand Economics（April1973），pp.11-33.
I did some rough calculations in 2006，using datafroma xian of moderately above-average performance,to arrive the following results.Assuming that an indus-trial site is built to a plot ratio of 0.8 and a factory oftypical labor intensity，the xian’s 4.25% cut of thevalue-added would come to 12% of the total cost ofacquiring the agricultural land and improving it to agood industrial site.This did not include administra-tive costs，but the figures suggest that the xian couldindeed afford to give away the site freely and subsidizethe investor a little.
There is no question that within a xian and in the saleof virtually identical industrial sites at the same time，theland price may vary greatly.Except for occasionalfavoritismthat leads to scandals，price discrimination isnot implied.Xian officials are selective of the type ofinvestors they want.What they are going after is notonly the value-added tax，but also the image and reputa-tion an investor would bring，compatibility with existingeconomic activity，and so on.It is not unusual that whenan investor’s project is believed to be a big drawfor thelocality，the negative land price would go so far downthat the xian’s cut in the value-added tax would notcover interest on the cost of land.