论文部分内容阅读
Given its sheer size the Chinese economy is closely scrutinized by the world. But a lack of transparent economic data makes this job difficult， and is a source of much ire among foreign observers. They’re not the only ones struggling， however， says Matthew Crabbe， Asia Pacific research director at consumer insights firm Mintel. The Chinese government is as much in the dark as anybody when it comes to Chinese numbers. In an interview with China Economic Review， Crabbe， who is also author of the recently published book Myth-Busting China’s Numbers， gives his thoughts on proxy measures for China’s economy and explains why we should pay more attention to consumer data.
There is a lot of skepticism among foreign observers of official Chinese data. It can’t all be unreliable， can it？
I don’t think any statistic is ever entirely accurate； it’s just the best estimate. I think one of the things about China is that it’s growing so fast that the expectation that the data could be accurate in such a situation is probably a false one. And I think the problem China presents is that common western expectation is that if the figures are published， that they therefore must be true. So yes there are problems in the data but it’s by order of degrees. As with anywhere， there will be slight inaccuracies in gathering data， and of course in China those get amplified as you go up the different levels towards the national picture.
Yes， there are problems with the GDP figures. But I think there are other common assumptions that I hear such as that the Chinese government knows what the real figures are and they don’t like to publish them. I think the Chinese government is as much as in the dark about what the real numbers as anybody else. If you think about it， the Chinese government needs an accurate picture as possible. They need to tax the economy so they can provide social services as a government. But the depth of that coverage remains very thin. It still needs a lot more money. The only way they can pay for that is tax from the economy， but without accurate figures it’s very difficult to do that.
How should we read economic data given regional disparities in reporting numbers？
Well， here are what we think is a common problem. We often look at a figure， and we assume that’s true. You read the figure， but you don’t investigate it. I think what you have to do is that you have investigate it， and not assume that figure says it what it is. One of the key things with a statistic is investigating how it was created it， why it was created， what the problems might be， what the definitions are. Before we agree on these numbers we have to know where they came from. Investment banks are constantly trying to forecast the country’s growth figures， and in so doing， have relied on an ever-changing array of proxy statistics such as rail freight volume. What are the most reliable proxy statistics for GDP now？
As the economy shifts away from export manufacturing， more towards domestic consumption， the range of proxies definitely have to shift. Before， statistics on cement， power generation had always been useful. Figures on construction and heavy industrial inputs are proxies for how industry is doing， and just looking at that now， its quite limiting. You need to understand what’s happening in consumer markets. You could look at retailers. I know some investment banks that are looking at the growth of KFC and certain retail companies. That could give you an idea of where the consumer market is going. Household spending indicators that come from the government - the rural and urban household consumptions surveys are quite interesting. These figures might be skewed but they give you a good indication of trends. Now， you’ve got big data in the form of the trends that are generated by online retail sales， which can be more time-sensitive and more current. These can be quite useful indicators and proxies as well.
Economists have complained that accurately measuring China’s services economy is close to impossible. How can we go about measuring it， or at the very least， improve methods of measuring it？
I know the National Bureau of Statistics is staring to try and measure services much more than in the past. It might take a while to really develop the figures. The way I did it with retail was to do it from the bottom up. And， in the end， I think that’s the only way you can do it. Split services into their component parts and research each one individually， and get the picture from the bottom. Services is different from retail because retail is volume of sales， whereas services is value-added， in terms of the value of human service， so for example you need to look at wages in the different services sectors to get an idea of growth and the relative strength of different services.
By clamping down on corruption at state-owned enterprises the government could be looking for a way to get better information on corporate revenues. Do you think there is some truth to this？
The government has social security net spending duties to fulfill and to be honest the money has not been enough to cover what it really needs to cover. The only way they’re going to get that [money] is if they can get true figures from companies， and particularly state-owned enterprises. A few months ago， the national oil corporation was caught for under-declaring its profits and not paying enough tax. If flagship state-owned enterprises are avoiding tax， what’s the picture underneath that？ Of course， it’s widespread. That’s a real problem， because this is money that should be going to the government coffers. But the money isn’t there. I think its part of clamping down on this rife tax avoidance. What are the mistakes that foreign companies make when using data to enter or expand in China and how can they try and avoid them？
One of the main mistakes that foreign companies make is reading a figure and believing it. If you really want to operate in China， you’ve really got to know it. It comes back to forensic due diligence into the figures and understanding the granular and the micro markets that you’re dealing with. Ask： What do consumers really want to buy？ Not just assuming what people are buying， when that might be because that’s all there is available to buy. It’s about getting that much better understanding and not assuming what you’re told is true.
If you could have any complete economic data set， what would it be？
Services are the one grey area that people really need to work on. What do people rent？ Who do they hire？ What do they spend doing rather than buying？ I think there’s a big shift in the consumer market more towards buying experiences rather than buying things. It’s about quality of life. We think this kind of thing is becoming much more important， it’s how you use your time， your life to your best advantage. It’s “I don’t want to spend hours doing the laundry， I want to have somebody else do it， or have somebody else collect it， wash it， and then send it back to me”. That’s the kind of thing people want. It’s that lifestyle efficiency if you like. Services are only going to grow and grow. If we have a better measure on that， we’ll have a much better measure on the real economy.
What are your biggest headaches when developing data on China？
I think it’s just that you have to cover so much. It’s such big numbers. Because it’s such a big country， the margins of error can be amplified greatly. I think that’s the real issue with China， the sheer size of it and the fact that it keeps changing. You can’t go back and assume the figures you had last year were correct. You have to pretty much start from scratch every time， so that you can revise upwards， well， usually upwards. It’s just hard work.
What’s your most significant discovery about China data from writing the book？
We all know China has grown big and fast. But I think the one thing that surprised me is the fragility of the economy， still. The fact that there are problems with corruption， a lack of a social security net， a lack of labor mobility， a lack of rule of law in certain aspects of the economy and also the amount of debt there is. In 2012， it was something like 205% of GDP. According to Standard Chartered， in the first quarter of this year total debt amounted to around 250% of GDP. So， it’s getting worse. It’s that fragility that frightens me， and something that continues to concern me.
There is a lot of skepticism among foreign observers of official Chinese data. It can’t all be unreliable， can it？
I don’t think any statistic is ever entirely accurate； it’s just the best estimate. I think one of the things about China is that it’s growing so fast that the expectation that the data could be accurate in such a situation is probably a false one. And I think the problem China presents is that common western expectation is that if the figures are published， that they therefore must be true. So yes there are problems in the data but it’s by order of degrees. As with anywhere， there will be slight inaccuracies in gathering data， and of course in China those get amplified as you go up the different levels towards the national picture.
Yes， there are problems with the GDP figures. But I think there are other common assumptions that I hear such as that the Chinese government knows what the real figures are and they don’t like to publish them. I think the Chinese government is as much as in the dark about what the real numbers as anybody else. If you think about it， the Chinese government needs an accurate picture as possible. They need to tax the economy so they can provide social services as a government. But the depth of that coverage remains very thin. It still needs a lot more money. The only way they can pay for that is tax from the economy， but without accurate figures it’s very difficult to do that.
How should we read economic data given regional disparities in reporting numbers？
Well， here are what we think is a common problem. We often look at a figure， and we assume that’s true. You read the figure， but you don’t investigate it. I think what you have to do is that you have investigate it， and not assume that figure says it what it is. One of the key things with a statistic is investigating how it was created it， why it was created， what the problems might be， what the definitions are. Before we agree on these numbers we have to know where they came from. Investment banks are constantly trying to forecast the country’s growth figures， and in so doing， have relied on an ever-changing array of proxy statistics such as rail freight volume. What are the most reliable proxy statistics for GDP now？
As the economy shifts away from export manufacturing， more towards domestic consumption， the range of proxies definitely have to shift. Before， statistics on cement， power generation had always been useful. Figures on construction and heavy industrial inputs are proxies for how industry is doing， and just looking at that now， its quite limiting. You need to understand what’s happening in consumer markets. You could look at retailers. I know some investment banks that are looking at the growth of KFC and certain retail companies. That could give you an idea of where the consumer market is going. Household spending indicators that come from the government - the rural and urban household consumptions surveys are quite interesting. These figures might be skewed but they give you a good indication of trends. Now， you’ve got big data in the form of the trends that are generated by online retail sales， which can be more time-sensitive and more current. These can be quite useful indicators and proxies as well.
Economists have complained that accurately measuring China’s services economy is close to impossible. How can we go about measuring it， or at the very least， improve methods of measuring it？
I know the National Bureau of Statistics is staring to try and measure services much more than in the past. It might take a while to really develop the figures. The way I did it with retail was to do it from the bottom up. And， in the end， I think that’s the only way you can do it. Split services into their component parts and research each one individually， and get the picture from the bottom. Services is different from retail because retail is volume of sales， whereas services is value-added， in terms of the value of human service， so for example you need to look at wages in the different services sectors to get an idea of growth and the relative strength of different services.
By clamping down on corruption at state-owned enterprises the government could be looking for a way to get better information on corporate revenues. Do you think there is some truth to this？
The government has social security net spending duties to fulfill and to be honest the money has not been enough to cover what it really needs to cover. The only way they’re going to get that [money] is if they can get true figures from companies， and particularly state-owned enterprises. A few months ago， the national oil corporation was caught for under-declaring its profits and not paying enough tax. If flagship state-owned enterprises are avoiding tax， what’s the picture underneath that？ Of course， it’s widespread. That’s a real problem， because this is money that should be going to the government coffers. But the money isn’t there. I think its part of clamping down on this rife tax avoidance. What are the mistakes that foreign companies make when using data to enter or expand in China and how can they try and avoid them？
One of the main mistakes that foreign companies make is reading a figure and believing it. If you really want to operate in China， you’ve really got to know it. It comes back to forensic due diligence into the figures and understanding the granular and the micro markets that you’re dealing with. Ask： What do consumers really want to buy？ Not just assuming what people are buying， when that might be because that’s all there is available to buy. It’s about getting that much better understanding and not assuming what you’re told is true.
If you could have any complete economic data set， what would it be？
Services are the one grey area that people really need to work on. What do people rent？ Who do they hire？ What do they spend doing rather than buying？ I think there’s a big shift in the consumer market more towards buying experiences rather than buying things. It’s about quality of life. We think this kind of thing is becoming much more important， it’s how you use your time， your life to your best advantage. It’s “I don’t want to spend hours doing the laundry， I want to have somebody else do it， or have somebody else collect it， wash it， and then send it back to me”. That’s the kind of thing people want. It’s that lifestyle efficiency if you like. Services are only going to grow and grow. If we have a better measure on that， we’ll have a much better measure on the real economy.
What are your biggest headaches when developing data on China？
I think it’s just that you have to cover so much. It’s such big numbers. Because it’s such a big country， the margins of error can be amplified greatly. I think that’s the real issue with China， the sheer size of it and the fact that it keeps changing. You can’t go back and assume the figures you had last year were correct. You have to pretty much start from scratch every time， so that you can revise upwards， well， usually upwards. It’s just hard work.
What’s your most significant discovery about China data from writing the book？
We all know China has grown big and fast. But I think the one thing that surprised me is the fragility of the economy， still. The fact that there are problems with corruption， a lack of a social security net， a lack of labor mobility， a lack of rule of law in certain aspects of the economy and also the amount of debt there is. In 2012， it was something like 205% of GDP. According to Standard Chartered， in the first quarter of this year total debt amounted to around 250% of GDP. So， it’s getting worse. It’s that fragility that frightens me， and something that continues to concern me.