China Research

In 2012, China will almost certainly achieve another first, surpassing the U.S. to become the largest national market for computers as measured in units.  In 1995, China consumers, businesses, and government enterprises purchased 3% of the world’s PCs and servers.[1] The U.S. purchased 39% of the world share that year.   In 2011, each country’s end users purchased 19% share.   How did China catch up with the US in the size of its final market for PC sales in just 16 years?

This account is based on Intel’s best estimates of the demand for PCs over the past 16 years.  I thank Intel Corporation for allowing the use of these numbers, but Intel bears no responsibility for the use or misuse of these results.

Before parsing the growth figures to see what drove the rapid convergence of China to U.S. results, we need to discuss the data issues behind this work.   The data behind PC purchases and uses can be characterized as sparse and poorly conditioned.  Markets for PC components such as CPUs, hard disk drives, and LCD panels as well as the markets for assembled PCs are global.   Laptop computers and a number of desktop PCs generally are assembled in Asia and often in China.    Imports and re-exports of components and assemblies are the rule.  Once PCs are fully assembled, they are sold through a variety of venues.   Outside of the advanced economies, so called “channel” points of sale are favored.   The “channel” points of sale range from formal, permanent stores with large displays, branded systems, and documented transactions to small, temporary kiosks with undocumented sales, little or no tax collection, and little or no licensed software.   To complicate matters further, PCs are durable goods and, once purchased, can remain in a household or enterprise until “refreshed” and can stay in a country or the world installed base for a time ranging from a few weeks to many years.    Hence, the uncertainty about sales figures pale compared to the uncertainties about the installed base of PCs.    Keeping these difficulties in mind, the reader will appreciate that all of the numbers presented here are estimates; indeed, they are “modeled” more than they are “data” in the popular parlance.

Total PC sales were 15 times higher in the U.S. than in China in 1995 and were at parity by 2011.   Convergence to total parity does not mean parity in all particulars, of course, but the convergence was remarkably uniform.   In 1995, U.S. business PC purchases were 12 times higher than China’s.   By 2011, the U.S. advantage in business purchases was only 1.1 with parity likely in 2012.   As befits the large public domain in China, U.S. government plus education purchases were only 7 times higher in the U.S. than in China in 1995 but China public purchases exceeded U.S. purchases by 5% by 2011.   The most dramatic change occurred in consumer demand.    The U.S. advantage shrank from 49 to 1 in 1995 to a 4% deficit by 2011.   The early China consumer sales numbers, however, should be regarded as very rough estimates.

While the various demand segments converged, there was also convergence among devices.   Desktop PC purchases in the U.S. outnumbered those in China by nearly 14 to 1 in 1995 but by 2011, China had a 62% advantage.   Parity in desktop purchases was obtained in 2006.   Notebook purchases in the U.S. were 44 times greater than those in China in 1995 but only 39% greater by 2011.   Convergence has been very rapid even for notebooks, once considered luxury items compared to desktops.   Servers were still purchased in the U.S. at 2.5 times the rate in China in 2011 but even this has converged from the nearly 7.5 multiple in 1995.    Server purchases are expected to be greater in the U.S. than in China for several more years as data centers expand quickly in the revitalized U.S. market.   However, China government and business leaders have demonstrated strong interest in an expanding Chinese Internet presence that will likely lead to more domestically hosted data centers and eventual parity with U.S. unit purchases.

Parity in PC and server purchases between China and the U.S. will inevitably be followed by outsized Chinese purchases.   China’s four to one population advantage over the U.S. has and will overwhelm the very large (but narrowing) U.S. lead in per capita income.   While there are significant differences in the Chinese adoption of technology, the importance of rising GDP and falling computer prices has made convergence with the U.S. numbers inevitable.   The driver of this convergence can be seen in an affordability measure: how many weeks of average per capita income did it take to buy the average purchased PC in China and the U.S. in the past and the present?    We’ve calculated these figures for 1995 and for 2010.   While it took an estimated 100 weeks of average income in China to purchase the average consumer desktop in 1995, and 176 weeks of average income to purchase the average consumer notebook, it took only 5 weeks to buy the average desktop and 7 weeks to buy the average notebook in 2010.  In the U.S., the 5 weeks of income for the average U.S. consumer notebook and the 3.4 weeks for the average U.S. consumer desktop narrowed to only 5 days of income for a desktop or a notebook by 2010.   Even in the U.S. the falling prices for rapidly improving PCs enabled by Moore’s Law have driven robust purchases for decades now.   In China, the steady decrease in prices for better and better PCs has combined with rapidly increasing incomes to lead to the displacement of the U.S. as the largest national market by the still rapidly growing China market.

Paul Thomas
Chief Economist, Intel