Chapter 4 Chapter Two How fast is China's economic growth?

As a researcher of the Chinese economy, I often fly around the world to meet fund managers and business people.I am often asked the question "How much do you believe in so many Chinese economic data in your report?" Sometimes, it is the same question, but it is asked more sharply and directly.In China, I also sometimes hear people question the fact that the official data seems to be different from the reality.The Chinese call this data "moisture", which means that it has been mixed with water.For example, was inflation really that low before 2008?Why do people in the city feel that the cost of living is soaring? Is it true that the increase in housing prices in 2007-2008 was only as small as the official data revealed?Many people's experience is much deeper.Is the urban unemployment rate really less than 5% in the past five years?The global economy has fallen into a severe recession that has not been seen in decades. Many Americans and Europeans are very puzzled. In this environment, how can the Chinese economy maintain a year-on-year growth rate of more than 6% in the fourth quarter of 2008 and the first quarter of 2009. ? In the fourth quarter of 2008 and early 2009, I was often asked "Is China's economy really growing—or is it in recession?" Some critics were overly critical of the quality of economic data.However, there are also some rigorous scholars who, through rigorous analysis, put forward well-founded and profound reflections on the authenticity of some data.

So in this chapter we'll talk about statistics—hopefully, I've been able to take an unbiased stance.The first part of this chapter will analyze the most important economic data GDP.Simply put, GDP is the total value of goods and services produced by an economy within a certain period of time, usually a year.Facing the quality of China's GDP data, we can ask such a basic question: How fast is China's economic growth?Some scholars believe that the GDP data at certain times are "watery", especially when the economic growth is slow, such as now.The most influential person holding this view is Thomas Rawski, a professor of economics at the University of Pittsburgh in the United States, and he is also a well-known expert on China issues.Another influential scholar is Carsten Holz of the Hong Kong University of Science and Technology. He has devoted a lot of time and energy to the study of Chinese economic data beyond your imagination.

When economists suspect that official data may contain "moisture," their usual practice is to find as many "proxy" indicators of economic growth as possible, including domestic freight volumes and electricity consumption, that are representative of overall economic activity. situation.However, as I have explained, since these indicators often have their own problems or tendencies, they should be treated with great caution.But I still think it is reasonable to believe that the Chinese economy will not grow more than 6% in the fourth quarter of 2008 and the first quarter of 2009.

Due to the significant significance of China's economic growth to global economic growth, many people are eager to understand the actual situation of China's economy.If they think the statistics are miscounted, or worse, falsified, then there is a problem.Not only may China's image be damaged, but it will also interfere with the formulation and implementation of other countries' economic policies.More importantly, China's policy makers need to make important decisions about cooling or fueling the economy, and they need to get the data right, or it will spell trouble for us all.

It is a well-known fact that sometimes local governments may over-report or under-report data due to certain motives and considerations.The performance appraisal of local officials often depends on how much investment they have attracted (especially foreign capital), how much tax revenue they have realized, and most importantly, what economic growth data they have reported.Therefore, it is not difficult to understand that driven by these incentives, local officials generally report better data.But this is not the only reason for data problems, there are other common objective reasons that need to be fully recognized:

1. China is still a developing country.Accompanied by this fact, there are several problems: First, China still lacks the mature data reporting system that developed countries usually have.In fact, China is far better than other developing countries in broadening the coverage of data.Some developing countries can't even calculate GDP in time.For example, my colleague in the Middle East has been doing research without official inflation data for several years. He often has to wait until half a year has passed before he can see the GDP data for the previous year. What's more, two The government department released two completely different annual GDP, he was almost dizzy!

2. At the same time, everyone knows that China is a rapidly changing economy.The company that provided you with data yesterday may disappear today; the migrant workers who received your research yesterday may disappear tomorrow.The department that publishes the data is constantly improving the way it counts—of course that’s a good thing, you get better data; but it’s also bad because you can’t compare the new data with the old data, so you can’t get a clear picture. trend.It is relatively easy to do statistics in a mature economy with stable economic growth and no structural changes, but it is not easy to do statistics in an economy like China.

3. China's statistical system also has to face other problems left over from history.In fact, between 1968 and 1972, the government statistics department was closed, and it is particularly regrettable that data collection was also interrupted during this period.It is even more pity that a large number of statistical talents were swept out together with their statistical knowledge.When the statistical agencies resumed, it was difficult to find qualified statistical professionals.It took a long time to retrain statistical experience and statistical talent. 4. Another problem is also related to historical factors - the current statistical system is very good at grasping the production data of enterprises, especially state-owned enterprises, but poor in service industry statistics.The service industry is mainly composed of small-scale private enterprises, such as the catering industry, which is an important part of China's economy.The reason for the weak services statistics is that during the Mao era, industrial production dominated the economy and services were insignificant.The basic accounting principle of statistics is the Material Product System (MPS), and service industry accounting has never been carried out. In the 1980s, the accounting principle was transferred to the System of National Accounts (SNA), which gave due importance to the service industry.However, historical legacies persist in a number of ways.For example, there are still no monthly estimates of service sector activity, and retail sales figures reflect only a subset of consumer behavior—they include a large number of public purchases by government and business entities, and thus cannot be called a reliable indicator of private consumption .

5. China has a vast territory and is therefore very diverse. Simple averages of such diversification (many official data are usually processed in this way) tend to cover up a lot of content.Living in a big city like Shanghai or Beijing means you deal with a very different economic environment than rural Sichuan or Harbin in the northeast.It also means that sometimes the stats don't mean much to you and don't seem to match your actual experience.This may be a problem with the data, or it may indicate that the economic development of your place is different from other places in China.This was especially evident in 2009, when people living in Shanghai felt chills, while Tianjin posted an astonishing double-digit economic growth rate.

All these problems are also fully understood by people in statistical departments across the country. They spend a lot of time and energy improving statistical work, which is very positive and gratifying.For example, they have conducted a large number of investigations by themselves to verify the accuracy of the data reported by various places.Below we will mention one by one the progress made by the statistical department in improving data statistics. But first, let's revisit the great international debate a few years ago about the veracity of economic data the last time China's growth slowed.

In 1997, more than ten years ago, the Asian financial crisis broke out.Under the violent impact of international and domestic factors, the economies of Hong Kong, China, Thailand, South Korea and Indonesia contracted sharply.The epicenter of this crisis is in Asia, and it has affected developing economies such as Russia and Brazil, as well as developed countries.China's export markets, including the United States and Europe, have experienced an economic slowdown, and foreign investment inflows have also slowed.What is particularly important for China is that the domestic economy is still under the influence of macro-control measures to cool down the overheated economy in 1994 and is facing great challenges.At the same time, the mess left by the reform of state-owned enterprises needs to make great efforts to clean up. In the mid-to-late 1990s, millions of people were laid off due to the reform of state-owned enterprises, and the private sector could only absorb a part of them.Under the attack of internal and external factors, the Chinese economy struggled from 1998 to 1999.Around 1999-2000, world trade picked up, the climax of state-owned enterprise reform was over, and private enterprises received more policy support.The housing system has undergone major reforms since 1998, with the abolition of the welfare-allocated housing system. The commercial housing market has been gradually established, and household balance sheets have been greatly boosted, laying the foundation for accelerated consumption growth in the next few decades. However, in the eyes of economists, 1998 appears to be very mysterious.According to official data, China's economy actually grew by 7.8% in 1998.In the eyes of some scholars, this number seems extremely high compared with the pain that people actually felt at that time.Roski, a professor of economics at the University of Pittsburgh in the United States, believes that China's official GDP data in 1998 and 1999 "completely deviates from reality."According to his estimates, the annual economic growth rate in the past two years was only 2% at most, and the cumulative economic growth between 1998 and 2001 was between 0.4% and 11.4%, while the official data (data before the 2004 economic census was revised) was 34.5%. . "What's Happening to China's GDP Statistics?", 2001, and "China by the Numbers: How Reform Affected China's Economic Statistics", China Perspectives, No.33, June 2001 To prove his point, Roski cited the stagnation of rural income, laid-off workers Reports of sudden increases, inventory backlogs, and declines in steel production pointed out that household spending growth was far below official retail sales growth, and further suggested that local governments under pressure to catch up with the central growth target (8%) reported higher than actual levels economic data.All of Rawski's above opinions are published and submitted to the National Bureau of Statistics of China. To support his point, Rawski also compares other measures of economic activity, which we call "growth proxies." Rawski first analyzed energy production. In 1998, the average growth rate of power generation dropped from 4.5% in the previous year to only 2.5% (but rebounded in the second half of 1999, bringing the average growth rate of power generation in 1999 to 6.8%).Rawski also examined the number of air passengers. In 1998, the number of air passengers increased by only 3.4%. According to his calculation, if the actual economic growth in that year was really close to 8%, the growth rate of air passengers should be close to 19%.Another indicator Roski uses is imports.If a country has a large demand for imports, the currency value of the country is stable (the renminbi was stable at that time), and the import trade barriers did not change (the import tariff barriers in China did not change at the time), then imports should be a good indicator of domestic demand growth.We can divide imports into processing trade imports (imports of parts and components used for assembly and then export) and general trade imports.General trade imports can better track and reflect the status of domestic economic activities.General trade imports fell by 5% year-on-year in 1997, almost zero growth in 1998, and quickly rebounded to 31% in 1999.Similarly, import indicators also indicate that China's economic growth in 1998 was not good. Not long ago, Harry Wu of the Hong Kong Polytechnic University raised new questions about China's economic growth data. "The Chinese GDP Growth Puzzle: How Fast has the Chinese Economy Grown?", 2007 He spent a lot of effort researching historical industrial data, and spent months (if not years) of time and effort to create China by industry Output data from 1949 to 1997. "How Fast has Chinese Industry Grown? Scepticism about the Official Statistics", Review of Income and Wealth, June 2002, he mainly calculates the quantity of products produced, while the official data records the value of output (so after the price changes meaning will be weakened).Wu Xiaoying found that since the 1970s, official data had overestimated industrial output, especially in the 1980s.His paper is the first systematic attempt to reconstruct more accurate Chinese industrial data. Wu Xiaoying also concentrated on studying the 2004 national economic census data.Through this census, the National Bureau of Statistics verified the accuracy of statistical samples and data used.One of the key findings of the economic census is that the service sector is much larger than expected.And, the census also appears to have found that previous data had significantly underestimated the magnitude of price increases.Therefore, the National Bureau of Statistics raised the real GDP between 1994 and 2004 by an average of 0.5 percentage points per year.The acceleration of economic growth during this 10-year period has a great impact, which means that the statistical department must increase the total economic output in 2004 by about 17%, from 13.7 trillion yuan to 16.1 trillion yuan.This is like discovering that Indonesia is hiding in the desert of the Gobi Desert in Xinjiang. Wu Xiaoying questioned this amendment.The National Bureau of Statistics did not carry out price statistics between 1994 and 2004.But the method of revising real GDP should be based on the availability of new and accurate price data.Wu Xiaoying therefore infers that the National Bureau of Statistics has outlined a GDP growth trend that it considers reasonable, and then adjusted the price deflator (that is, the overall change in prices each year) accordingly.That is to say, the NBS has to make certain assumptions about the price changes in this 10-year period.Of course, the statistical department can draw a general price trend from a large amount of data, but how to get the price deflator is not known to the outside world. After the 2004 national economic census, as mentioned above, the National Bureau of Statistics raised the GDP data for the years from 1994 to 2004, except that the GDP growth rate in 1998 remained unchanged at 7.8%.Like Rawski, Wu Xiaoying also believes that this data showing a weak economic slowdown is a bit strange compared with other situations that actually happened in the Chinese economy at that time. Therefore, he once again used the self-built industrial growth data system to conduct calculations.As we mentioned earlier, Wu Xiaoying’s data is concerned with the quantity of output, while the statistical method of the National Bureau of Statistics focuses on the value of output. Both ways have pros and cons.Focusing on volume avoids the impact of yearly price changes (trends are difficult to track) and prevents local governments from overstating the value of output.Since it is easier to examine the quantity of output than its value, Wu Xiaoying's method has important implications.However, the weakness of this method is that he has to make an estimate of the price.This is very difficult.According to Wu Xiaoying's estimation, the average year-on-year growth rate of industrial output from 1978 to 2002 was 16.2%, lower than the official 20.4%.He believes that in 1998, China's industrial contraction (negative growth) was nearly 7%, and in the previous 1996/1997, there was basically no growth.In other words, Wu Xiaoying's data support Roski's view that China's economic growth rate in 1998/99 was much lower than that indicated by official data. Wu Xiaoying also worked with Madison to reconstruct China's GDP growth rate since the 1950s.They used Wu Xiaoying's estimated industrial output data, as well as agricultural and service industry data from other sources, to obtain overall economic data.According to their estimates, from 1978 to 2003, the average annual growth rate of China's economy was 7.9%, which was lower than the official data of 9.6% (before the revision of the 2004 economic census).Their point of view on economic growth in 1998 is that China's economic growth rate was not large that year.They believe that China's economic growth rate was 0.3% in 1998, and the real economic growth rate recovered to 6.6% in 1999. "Measuring China's Economic Performance", World Economics, April 2009. There are always at least two schools of thought on most controversial issues.Naturally, the viewpoints of Rawski and Wu Xiaoying have also been criticized by many economists.Some internationally renowned scholars believe that the degree of China's economic slowdown in 1998 was not as drastic as Roski and others claimed. Nicholas Lardy, a China expert at the Peterson Institute for International Economics in the United States, also observed a variety of economic growth indicators. "Evaluating Economic Indicators in Post WTO China", the year-on-year growth rates of government tax revenue from 1997 to 1999 were 22.5%, 10.7% and 19.4% respectively.Since tax revenue is mainly contributed by industrial value-added activities, and GDP is also a value-added concept, tax data can be a good indicator of economic growth.Tax data show that since the second half of 1998, China's economy has stepped out of a V-shaped rapid recovery trend.This shows that the situation in 1998 was not as pessimistic as Roski said.Lardy also studied import data, which began to pick up in 1999. Based on this, Lardy refuted Roski's claim that the Chinese economy continued to slump throughout 1998-2001. He believed that the actual growth rate in 1998 was consistent with official data. There is a certain distance than that, but it is definitely not as big as Roski said. A true academic, Rawski takes the criticism seriously: official tax data may not be a reliable indicator of value-added activity, especially in times of recession.This is because the ability to generate income is an important indicator for the performance evaluation of local officials, and the local government needs to complete certain tax indicators every year.We know that in the past (and some guess even now), local governments kept a portion of their revenue off-budget. In the 1990s, this portion of revenue was referred to as "extrabudgetary" revenue.In recent years, the central government has carried out a series of reforms to the national budget system to standardize local revenue and expenditure accounting.What happened in 1998/1999 may be that the local government raised tax revenues to meet the annual target by transferring extra-budgetary revenues into the budget.Transferring extra-budgetary revenue into the budget is a clever way to inflate fiscal revenue and smooth fiscal revenue fluctuations.In 1998/1999, Roski asserted, this practice may have been more frequent. As for the import data, Rawski believes that the increase in imports through official ports under the anti-smuggling measures in the late 1990s means that the import data is not a reliable indicator of economic growth. In addition, there may be problems in the statistics of power generation data. At that time, many independent power generation companies were not included in the official sample, and many companies generated their own power to make up for the shortage of power supply from the grid.As a result, electricity generation and consumption data lose the ability to gauge overall economic trends. In short, Roski persisted, and Lardy's doubts persisted. Mu Jia of the Hong Kong University of Science and Technology also joined the debate. "Fast, Clear and Accurate: How Reliableare Chinese Output and Economic Growth Statistics?", in 2003, he devoted a lot of energy to researching the Chinese statistical system, and wrote many amazing discoveries.In a paper he published in 2003, he defended the veracity of Chinese statistics. He noted that the National Bureau of Statistics has taken steps to improve data collection by allowing large companies to report data directly online.In addition, the Bureau of Statistics also conducts surveys on urban and rural household income and expenditure through urban survey teams and rural survey teams, as well as industrial surveys.These two methods can effectively avoid the incentives of local governments to make false reports caused by the reporting system (the data reported by local officials is an important part of their performance).The investigative brigade is managed vertically, and personnel salaries and promotions are not subject to official government control.It is a well-known fact that in recent years, the official national GDP has been significantly lower than the data reported by the provinces.The reason is that in addition to the data reported by local governments, the National Bureau of Statistics also uses its own research data to verify the actual situation.This was not possible a few years ago. After the above remarks, Mu Jia then expressed his concerns.He studied the revision of economic growth data over the years after the 2004 economic census, "China's 2004 Economic Census and 2006 Benchmark Revision of GDP Statistics: More Questions Than Answers", and tried to recalculate these years with the calculation method adopted by the National Bureau of Statistics GDP data, and found several inconsistencies.He believes this is a problem caused by data correction.His conclusion is that official GDP figures have a tendency to contain "moisture" when economic growth slows. More importantly, similar to Rawski, Mujia also pointed out the problems of some economic growth indicators.This lesson needs to be heeded. At the end of 2008, some analysts questioned the 6.8% GDP growth rate in the fourth quarter because electricity generation fell to zero growth in the quarter.They believe that China's economy has stagnated in terms of power generation.But in fact, there are also relatively large structural tendencies in the power generation data in recent years.Steel, copper and aluminum producers are big consumers of electricity, and if production in these industries stops, the generation figures will quickly and sharply fall.But that doesn't mean the entire Chinese economy has come to a standstill, too. There are other indicators that better reflect the health of the economy, and freight volumes are one of them.The Ministry of Transport and the National Bureau of Statistics collect data directly from railways, ports and large transport companies, while also conducting some research themselves.An obvious positive meaning is that the freight transport of many private enterprises should also be included in freight statistics.What is still insufficient is that, compared with rail transport, since the operators of river transport and road transport are mostly private, they are not included in the official statistics, and there may be underreporting. He further pointed out that the problem lies in many economic indicators themselves.Taking energy consumption as an example, Mu Jia believes that its survey coverage is too narrow to be a reliable indicator of economic growth.Other studies have also pointed out that the closure of small coal mines, the expansion of small power plants and the increase in the company's own power generation have not entered the statistical system and have affected the data.An independent study of petrol consumption by a road transport company found that petrol consumption rose much faster in the 1990s than official freight figures suggested, meaning it was lower than it actually was.After careful comparison of the data, Mu Jia also found that there was indeed a positive correlation between changes in freight data and industrial output in the 1990s (but not in the 1980s). The Standard Chartered China Freight Index we created covers road, rail and air transport data.In our view, this indicator provides a good indicator of the state of industrial activity.It is a pity that the data we have at present starts from 1999, and there is no previous historical data.It can be clearly seen from the freight index that China's economy slowed down in 2001, and then a new round of growth cycle began. The outbreak of SARS in early 2003 temporarily disrupted this round of growth, and cargo volumes contracted for about a quarter compared with the same period of the previous year.But in 2004 it showed a strong rebound.Concerned about the overheating of the economy, the central government began to implement macro-control measures to cool down the economy.From the second half of 2004 to the first half of 2008, the freight economy grew rapidly and steadily.This period was also the golden age of China's industrial economy.But in the second half of 2008, China's economic growth slowed down sharply, and the growth rate of freight volume fell into the negative growth range again at the end of 2008.Since around the Spring Festival in 2009, China's economy has shown a strong V-shaped recovery trend. Returning to the topic of some indicators reflecting economic growth, it needs to be clear that these indicators can only tell us part of the story.Most of these indicators reflect the industrial economy - no indicator has been found to track the state of the service sector well.Moreover, the first national economic census found that the service industry was developing so rapidly that a large number of service industry activities were not included in the official statistical system.The service sector has become an important part of the informal economy (the economy that does not pay taxes).Also for this reason, we need to be very cautious when applying various economic indicators to the overall economy. So, how should we evaluate the academic controversy caused by the GDP data in 1998?The following are my summary points: 1. The bone of contention is official GDP in 1998.Whatever the respective problems and limitations of the various economic indicators, one cannot ignore the fact that all of them pointed to a deeper slowdown than official data suggested at the time. 2. Fiscal revenue and import data show that China’s economy did not stop growing in 1998, but an economic growth rate of 3% to 4% seemed more reasonable, although research by Wu Xiaoying and Madison showed that the economic growth rate at that time was close to zero. 3. Judging from the rebound of most economic indicators, the official GDP data from 1999 to 2001 are relatively healthy and realistic.Lowesky sees little evidence of continued economic weakness. 4. Statistical departments have made great progress in sample collection, but they still need to be cautious about some data, especially macro data such as GDP. 5. Economic indicators that reflect the economic situation are useful, but they need to be used carefully. It is better to compare multiple indicators comprehensively and confirm each other than to use them alone. In 2008, the pace of China's economic growth had slowed from the golden years that began in 2002, and now it is picking up momentum again, and the momentum is strong - far exceeding my expectations. What the year-on-year data tells us is how much it has grown compared to the same period last year.The year-over-year approach can eliminate seasonal effects.There are regular changes in economic activities in many countries throughout the year.China shows a greater seasonal effect around the Spring Festival or at the end of the fiscal year.Banks are generally willing to lend money in the first few months of the year, and the government seems to like to spend money suddenly at the end of the year.But if we have been comparing the data of a certain period with the same period of history, we don't have to worry about the interference caused by these factors.Usually we average the data of January and February together to remove the influence of the Spring Festival factor. However, year-on-year data cannot tell us too much about the situation and momentum of economic growth in the short term.Therefore, we still need to look at the quarter-on-quarter GDP. If we want to get updated information, we can also look at the month-on-month data.But a big problem with the chain data is the seasonal effect.Seasonal factors, especially the Spring Festival, will bring great interference to the data, resulting in data distortion.The solution to this problem is to study the long-term seasonal effects and adjust the data accordingly, ie seasonally adjusted.The problem with this treatment is that many times the seasonal effect is very large and thus the margin of error in making the adjustment is large. This is the case with the GDP data - the potential for error is large.Therefore, the quarter-on-quarter GDP data, such as the quarter-on-quarter growth rates of 1.3%, 3% and 2% in the first three quarters of 2009, still need to be treated with caution. The central government has clearly set "guaranteeing eight" as the economic growth target for 2009, and it may also be the target for 2010.It now appears that the growth target for 2009 is still achievable.It's just that what we should pay attention to is not only the statistical 8%, but more importantly, real growth and employment.Knowing the real economic situation is very important. In order to verify whether the overall economic growth data is accurate, we need to compare and confirm various economic indicators.Care must be taken in the comparison process.It is not a scientific attitude to conclude that China's economy has fallen into recession based on the continuous decline in power generation.In the second half of this section, we observe various available economic indicators separately, and analyze the limitations of these indicators reflecting the macroeconomic situation one by one.But before we get started, three caveats: First, the various economic indicators discussed here are coincident indicators, and they promise to tell us what happened today, not what will happen tomorrow.In fact, we hope to find some solid economic leading indicators that can tell us what will happen tomorrow.New bank credit is one of the best options, but it's not always reliable. Second, this point was also mentioned above.Most of these indicators tell us how industrial activity is, but very little about consumption and services, thus ignoring at least a third of China's economy. Third, most of these indicators describe industrial activity rather than value-added data.And GDP is an incremental concept.In general, it shouldn't matter, but it is necessary to be aware of it.Enterprises may always be busy, but they are not necessarily creating new value, so the actual GDP growth rate created is limited. The indicator that the market pays more attention to is the Purchasing Managers Index (PMI), which is generated by the results of a questionnaire survey of corporate purchasing managers on issues such as corporate orders, investment plans, employment plans, and inventory.A reading above 50 indicates that most business managers report better conditions this month than last, while a reading below 50 indicates worsening conditions.The official manufacturing PMI dropped below 50 in July 2008, continued to deteriorate, and returned to above 50 in February 2009. October 2009.The decline in the non-manufacturing PMI index in 2008 was weaker, and there were signs of improvement after the Spring Festival in 2009.It can be seen that the market reacted very strongly to the improvement of the PMI data, which is believed to herald a V-shaped rebound.The market generally believes that there is no artificial factor in PMI data.The only problem is that the proportion of state-owned enterprises in the sample is too large, and it is unclear to what extent it can accurately reflect the economic activities of the private sector.The state-owned industry is of course a huge beneficiary of the stimulus package.Another unofficial PMI (the Purchasing Managers' Index published by HSBC and data provider Markit) has also turned better since April 2009, suggesting that private sector manufacturing activity is also recovering - although the export sector is struggling. Mostly not so good. Next, let's look at the fiscal revenue data.We know that taxation is based on activities that add value, so taxation should be an indicator that can better reflect economic growth. In the fourth quarter of 2008, the national tax revenue growth rate fell to zero, and then fell again in the first quarter of 2009, down 8% year-on-year, and increased by 3% year-on-year in the second quarter. In the first quarter of 2009, corporate income tax dropped by 27%, and corporate tax revenue continued to decline in the second quarter.The increase in personal income tax was flat in the first quarter and rose slightly in the second quarter.Does this mean that the Chinese economy has stopped growing?Of course, there is no doubt that economic growth has slowed down significantly, but we still need to be cautious about fiscal revenue data because: 1. The source of government revenue has changed. In 2007 and into early 2008, booming real estate and stock market deals generated a lot of related tax revenue.These revenues should be reflected in the official fiscal revenue data and drive the rapid growth of fiscal revenue (the land transfer revenue in some places may still not be included in the fiscal revenue data, but into "extra-budgetary revenue", but the scale is difficult to determine).With the sharp shrinkage of real estate market turnover, related fiscal revenue dragged down the overall fiscal revenue decline.Moreover, in 2008, corporate profits changed dramatically, and loss-making enterprises had a greater impact on the growth rate of overall corporate profits. 2. Part of fiscal revenue may be included in extrabudgetary accounts.After the local government completes the fiscal revenue task for the whole year, it hopes to keep as much revenue as possible quietly in the local government and put it in the extra-budgetary account.When the "rainy day" really comes in 2009, leave room for stimulating the local economy. 3. Due to cash flow reasons, the time for enterprises to pay income tax may be postponed - the taxation authority's 2008 tax collection target has been completed in September, so the end of the year will be more relaxed for enterprises.The purpose of tax authorities doing this may be to realize the tax collection task in 2009, and everyone knows that 2009 is a very difficult year. 4. Tax rate changes. In January 2008, the domestic corporate income tax rate was reduced from 33% to 25% (foreign-invested enterprises will implement a 25% tax rate during the five-year transition period).The lower tax rate resulted in lower revenue for the full year of 2008, not just the fourth quarter. From January 1, 2009, the value-added tax transformation reform will be implemented in all industries, from production-type value-added tax to consumption-type value-added tax.The input tax included in the purchase of equipment by enterprises can be deducted.The resulting reduction in value-added tax may affect the tax growth trend in 2009.Since value-added tax accounts for a relatively large proportion of various government tax revenues, this factor should be taken into account when observing tax data after January 2009. For these reasons, I believe that the lower tax growth rate may have exaggerated the magnitude of the economic slowdown from the fourth quarter of 2008 to 2009.Similarly, when the economy recovered in 2009, tax data were lagging.At the same time, we must be very careful when looking at the overall fiscal revenue data, because it includes non-tax revenue. In June 2009, the overall financial revenue increased by 20% year-on-year. The rebound momentum is very surprising, but the tax revenue growth in the same period was only 12%. It may be that the land sales revenue obtained by the local government in 2007-2008 was handed over to the central government in mid-2009 to meet the fiscal revenue growth target.But by the end of the third quarter, the recovery in the broader economy may have been reflected in better tax revenues. Let's look at the power generation data.After years of growth rates of more than 10 percent, power generation has plummeted since the summer of 2008.Does this indicate a sharp decline in China's economic growth?Yes.We think the rate of economic growth has slowed somewhat, albeit briefly.As mentioned earlier, industry uses two-thirds of electricity, so generation figures only show us a fraction of the overall economy.Among various industrial sectors, heavy industry is also a large consumer of electricity, and the vast majority of electricity consumption is concentrated in aluminum, steel, copper and other production enterprises.这些行业受建筑原料需求大幅下降和2008年第三季度大宗商品价格突然调整的冲击，存货大量积压，很多企业不得不减产或停产。因此，这些行业的发电量需求大幅下降。但同时如果看非能源密集型工业，我们会发现2008年第四季度到2009年第一季度它们的产量出现剧烈下滑。换而言之，整个工业部门的增长大范围放缓，同期农业和服务业没有受到冲击。这些数据暗示2008年第四季度工业部门进入负增长，2009年第一季度以来出现微弱的恢复。 2009年6月，发电量开始上升，同比增长5.5%，表明重工业和轻工业都开始恢复。进入第三季度后恢复势头继续得以巩固。 货运量的变化可以较好地体现工业增长。2008年下半年至2009年第一季度，水路运输出现明显回落。水路运输具有成本低、安全性高等优势，非常适合煤炭、钢材等货物的运输。自全球金融危机以来，需求受到巨大影响，煤炭和铁矿石库存居高，这两个行业下滑幅度尤甚于其他行业。直到进入第三季度，水路运输似乎仍未表现出复苏迹象。 铁路运输方面，2008年第四季度也出现下滑。铁路运输数据可能是质量最有保证的货运指标（铁路全部为公有，数据集中上报），但铁路运力不足影响着数据质量。以2004年上半年为例，当时经济活动十分活跃，水路、公路运能利用率急剧扩张，但铁路运输没有扩张。原因是运力不足——车厢和铁轨不够用，到现在也一样不够用。2008年初我在新疆调研时发现，令很多企业头痛的是找不到车皮运出产品，运回化肥和饲料。一位挺有能耐的老总以不无得意又疲惫不堪的口吻说，他一路拜佛拜到地方最高党政领导那里，才求来几节车皮把生产的棉花运出去。这意味着铁路货运增速从来赶不上公路和水路货运增速，但在2008年11月，铁路货运增速也陷入负增长，到2009年第二季度末，才恢复到国际金融危机之前的水平。 公路运输数据保持强劲增势，未受太大影响，但这一指标存在的问题是，可能没有反映只拥有一两台车的“作坊式”私人运输，以及制造企业自行组织货物运输的情况（数据样本只包含特定的运输企业）。全球金融危机集中爆发以来，小运输企业的经营受到很大影响。公路运输数据表现出的弹性值得关注。 2008年第四季度，航空客运业进入寒冬，航空旅客周转量出现负增长。随着各航线票价全面降低，每客位收入直线下降。航空业首当其冲，受到企业、政府大幅削减开支的冲击，因而对于经济放缓十分敏感。铁路仍是最为大众化的出行方式。然而有意思的是，2009年第二和第三季度，铁路客运量没有回升，而是基本与上年持平。 接下来看进口数据。这里需要将两类进口加以区分。2008年第四季度加工贸易进口名义增长率大幅滑坡，这一类进口主要反映了全球经济活动状况，而不是中国国内经济活动状况。2009年第一季度，全球经济仍处于收缩之中，因此加工贸易进口继续受到影响。而同期一般贸易进口滑坡出现得并不像加工贸易进口那么早，但也在2008年第四季度进入了负增长。这是否暗示工业经济走入衰退？ 得出这样的结论存在很大的问题：我们所看到的是名义进口增长率。2008年下半年大宗商品价格平均从高点跌落40%～50%。因此，即便进口数量未变，进口金额也会减少。我们没有好的办法得出一般贸易出口的实际增长率，但可以看到进口，包括一般贸易和加工贸易进口的实际增长率。我将结果显示在图22和图214中，同样可以看出2008年第四季度进口明显收缩，其原因不仅仅是加工贸易滑坡，也是因为第四季度工业经济遭遇严重收缩。到2009年第二和第三季度，实际进口明显回升，进口增长率基本回到零的位置，换句话说，接近于国际金融危机之前的水平。这是非常值得注意的变化。我们也可以从主要进口商品的数量来观察，2008年第四季度进口量增长急剧放缓，到2009年第一季度，随着钢铁生产企业抓住原材料价格低廉的时机，并在经济刺激方案将大力提振钢材需求的预期之下大量补进原材料，铁矿石进口大幅增加。2008年第四季度，随着整体经济增速放缓，原油进口同比增长也变为负值，2009年第一季度恢复了一些，第二季度大幅增加。再次体现了工业部门以放缓的状态进入2009年，然后开始较为强劲的反弹。 如上所述，这些指标不能很好地反映服务业和私人消费的情况，因此也就漏掉了农业和服务业（占整体经济活动的55%）。那么，2008年第四季度和2009年第一季度消费发生了什么变化？这里也有点让人困惑。这一期间零售额保持旺盛态势，尤其是进行物价调整之后（物价自2008年开始回落），似乎中国消费者并未受到乱成一团的全球经济的影响。消费者信心数据却告诉我们另一个故事：似乎很多中国家庭的信心被发生的事实所撼动，直到2009年2月仍未恢复。从快餐业到服装业到移动通讯行业，很多商家在2008年第三季度左右就感觉到生意难做了。而销售额数据未能体现出家庭支出的放缓，我想，一个重要的原因在于，零售额数据中包含了政府和企业的团体采购，而且也未捕捉到城市新兴中产阶级的很多消费行为。消费者信心的低潮似乎出现在2009年第一季度，从第二和第三季度开始，央行的消费者信心指数开始逐步攀升。 其他数据显示，2008年第四季度消费受到冲击。商品房交易量和汽车销量下降。快餐、批发零售等消费行业也出现了生意转淡的情形。换而言之，2008年第四季度不仅工业部门出现剧烈收缩（出口跌落是部分原因），而且整体消费也趋缓。 2009年第一季度，消费者似乎恢复了活力。春节过后房地产市场开始回暖，3月商品房成交额重回增长态势（同比增长15%）。第二季度已售房屋面积大幅飙升。房地产市场的回暖是很多原因促成的，包括税收优惠、按揭利率降低以及房价走低。所有这些当然都是振兴房地产市场的因素，但人们对于刺激措施作出的反应释放出令人深思的信息。 到2009年10月，多个城市商品房销量下降，一个重要原因是房价冲高，挑战着很多人的支付能力。很多城市的房价在2009年的涨幅达到30%～40%。但是对于开发商而言，时机大好：它们已清空了库存，现金流充裕，有能力再度增加土地储备。我预计2010年随着新房供应上市以及信贷收紧，房价将出现一些波动，但我并不认为房地产整体处于泡沫区间，房价跌幅也不会超过20%。 2009年第一季度汽车销售额继一段时期的低迷之后出现反弹。购置汽车属于大额消费，因此汽车市场回暖是积极的迹象。有观点将汽车销售回升视为车辆购置税下调的短期刺激作用。自2009年1月20日起至当年年底，1.6升排量以下车辆购置税由10%降至5%。税收优惠的效应应该已经消失了（得要承认我也这么认为）。但是截至2009年9月汽车销量一路攀升，2009年前9个月共售出了约720万辆小汽车。 零售额持续走强显示，背后起作用的不仅是减税——人们拿出积蓄和增加的收入购车。 2009年初，当美欧消费者为行将不保或已经失去的饭碗、所剩无几的存款上街游行时，中国消费者似乎还在为有工作而庆幸（尽管这一年加薪无望，甚至荷包还会缩水），为无债压身而备感轻松。并且，中国人似乎还愿意花掉一部分储蓄，尽管中国人不如美国人那么富裕，还不能在规模上取而代之，但对于2009年的中国经济，仍不失为重要的积极因素。 现在我们所看到的各类指标绘出了一幅喜忧掺杂的晦涩图景，但有一点是十分明确的，即2008年第四季度经济状况非常糟糕。为将多种经济指标放在一张图中加以综合比较，我和同事李炜尝试将多项指标糅合成中国工业活动指数（China Industrial Activity Index，CIAI）。我们认为，这一指标是衡量工业活动的良好指标（但还没有发现衡量消费者活动的理想指标）。中国工业活动指数是由以下四项指标构成的： 1.工业产出。该指标追踪矿业、制造业以及公用事业（电力、水、燃气）等行业的工业产品生产总量。覆盖产品范围广泛，从原油加工品、电力产品，到盐、服装等。工业产出以实际数量单位来衡量。2007年工业部门对GDP的贡献率高达43%，因此工业产出是体现整体经济活动的一个重要指标。 2.渣打中国货运指数。我们的这一指标包括铁路、航空、公路和水路货运量。经过一段时间的实践，我们发现这一指标的走势与整体经济活动的相关性很强。 3.银行信贷增速。由于中国经济增长严重依赖信贷扩张，经济反弹的任何迹象极有可能从这一指标中显现出来。但只有银行信贷的实际购买力才具有价值，因此我们以生产者价格指数（PPI）作为通胀指标对银行贷款的名义增长率进行了调整。2009年上半年银行信贷的迅猛增长显然为中国经济提供了巨大刺激，这一时期中国货币供应增速在全世界最为可观。 4.主要大宗商品进口量。该指标衡量中国对主要大宗商品的进口数量，我们视其为体现整体国内需求增长的良好指标。如果一个国家在全球市场上的购买力相对稳定，并且关税壁垒未发生变化，该国进口增多则表明国内需求增加。但该指标会受到实际有效汇率变动的影响。 我们分三个步骤测算工业活动指数。首先计算各构成指标的同比变化率。对那些以不同形式表现的指标（如景气指数），以当期（t）与前期（t-12）数据间的算术差作为近似同比变化率；其次，将各构成指标的同比变化率除以其标准差，这样可以使各构成指标对综合指标变化的贡献率平均化；第三步，对调整后的综合指标同比变化率取平均值。它的波动性大于GDP数据。近些年工业活动指数表现出的特点包括： 1.1997年下半年，工业活动指数急剧下滑，原因是亚洲金融危机导致亚洲经济减速，1998年指数反弹，表明1998年经济并非近乎零增长。 2.2000～2001年，工业活动指数增长缓慢，之后在2003年SARS疫情爆发之前出现反弹。 3.SARS疫情过后，2003年下半年开始，工业活动指数显著加速。2004年上半年开始受国家宏观调控政策的影响，突然走低。 4.2005～2006年，中国经济开始重拾升势，到2007年增长加速。2008年经济活动减速，原因主要是信贷控制和美国经济危机造成的出口滑坡。 5.2008年最后两个月，工业活动指数进一步恶化，降至负增长，为自20世纪90年代初以来最低工业增速。 6.2009年前两个月，工业活动指数开始回升，银行信贷迅猛增长可能是原因之一。 7.2009年第二和第三季度，工业活动指数出现清晰的V字型反弹。反弹之强劲，反弹时间之早超乎我（和大多数人）的预期。 由此可以得出什么结论？从各种指标和迹象看，我很难相信2008年第四季度经济增长能达到6.8%，2009年第一季度能达6.1%。 我猜想实际数字要低很多，因为此次经济放缓的程度超过了1998年。估计GDP同比增幅在3%以内。但从2009年3月开始，工业部门开始稳固，消费者信心和消费活动似乎也恢复了元气。采购经理指数、货运量、发电量和其他经济指标都显示，2009年第二季度经济增长非常强劲，继上季度之后出现V字型反弹。无疑，年初以来银行信贷爆炸式增长是一个重要的推动因素。工业领域很多行业出现了V字型复苏，同样出现强劲复苏的还有主要消费领域，尤其是汽车和房地产市场。增长动能有望延续到2009年底，还有很多制造业企业客户反映，2010年订单形势不错。 从很多行业的增长动力，以及从企业客户反馈的信息看，到2009年第四季度和2010年第一季度之后，关于GDP数据的风险不是高报，而是低估。而在2009年初的时候，很少有人，如果不是完全没有的话，能够预想到还有这种可能。 中国经济增长高度依赖信贷增长，2009年上半年新增人民币信贷超过7万亿元，规模之大实属罕见。新增信贷是2009年经济复苏的必要但不充分条件。与信贷井喷相关的两个问题当然是，第一，信贷流向了哪里？用于生产，还是流进股市和存款账户？诚实的答案是无从知晓——货币周转过于容易，一旦资金离开了客户的账户，我们这些银行宏观分析师很难有更多的渠道去跟踪资金流向。或许你听说过一些猜测，例如新增信贷的30%进入了房市和股市。中长期贷款中相当一部分进入了基建项目（或至少被分配进基建项目），其中大多数项目应该有益于长期的经济增长。然而，部分短期贷款可能进入股市和房地产市场，或只是作为存款仍然存放在银行体系中。这种状况自然非常不健康，并且令人担心正在形成资产泡沫。还有部分资金可能在经济强劲反弹的预期之下，用于增加原材料库存。这对于未来经济增长具有信号传递效应。但是当然，企业也可能判断失误。 关于信贷猛增的第二个问题是，有多少贷款被浪费掉了？一些信贷资金伴随着对未来收益的不确定性进入一些基建项目。一些项目的背后，影绰着地方政府的身影。未来几年地方政府将背负巨大的税收压力。信贷扩张过快可能导致不良贷款增多。但更重要的是，这暗示中国银行业的商业化改革出现倒退。这不是商业性放贷，而是政策性放贷。尽管好处显而易见——可以非常快地刺激经济复苏；但是，从长期看存在着消极后果，本书的第九章将对这一问题进行深入探讨。几年以后，这些问题中的一部分又会跑来困扰我们，我们也不得不再次回到GDP增长率的问题上来。