Your Understanding of GDP May Be Wrong - Huxiu#
#Omnivore
This article discusses some common misconceptions about GDP, as well as the statistical methods of GDP and issues related to special projects.
• 📊 GDP measures the value added from domestic production activities, not total output.
• 💡 Different economies have different central growth rates for GDP, which need to consider what stage the economy is in.
• 🌍 Analyzing the total amount of GDP tells us what major phase the economy is currently in, while short-term economic cycles have limited impact on GDP.
This time, I want to talk about data, mainly because it is very important for our daily lives, whether we are reading financial news or analyzing economic issues. Discussing economic issues cannot be separated from data, but economic data, from the statistical stage, is something that requires a lot of subjective judgment.
It is not like physical data, where measuring the length of a track means 100 meters is 100 meters; you wouldn't say that this length needs seasonal adjustments or that there are issues with statistical standards. The 100 meters now is not the same as the 100 meters from 20 years ago. This is a characteristic of most natural sciences, where there is not much controversy in the measurement stage, although it does exist. In social sciences, from the measurement stage, there is a large amount of subjective judgment involved.
Simply believing whatever data others give you can easily lead to deception, which is why I titled this "The Trap of Data." If there is something about economic data that truly needs attention, it is definitely not about how to analyze the data, but rather whether there are traps in the data and whether it has been distorted by human factors.
Therefore, I hope to start a series to discuss the issues surrounding economic data itself. Of course, I will also talk about how to understand and analyze data, but after I finish, you may find that knowing how the data is collected is much more important than how to analyze it.
Additionally, discussing data is also meant to convey a methodology: When discussing economic issues, it is essential to talk quantitatively, not qualitatively based on feelings.
The most common thing we say is that the economy is not good this year, or it will not be good in the future, but this "good" or "bad" is a subjective statement. What exactly does good or bad mean? Is a 5% GDP growth bad, or is a 3% growth bad? Now everyone is pessimistic, believing that the Chinese economy will stagnate like Japan for a long time, but if you look back at the financial news from 2003 and 2004, during that bear market, many people were already saying that the Chinese economy would stagnate for a long time.
During every economic recession, people feel that things are not good, that the economy is failing, but one person's feeling cannot measure the state of the entire economy. Therefore, we need to establish this methodological concept: economic issues cannot be discussed without data; they cannot be discussed qualitatively; they must be discussed quantitatively.
- What is GDP?
So, at the beginning, we must talk about GDP. The importance of GDP goes without saying, but at the same time, the misunderstandings about GDP in the market are as important as GDP itself; there are too many misconceptions that need to be corrected.
First, there is the issue of naming. GDP stands for "Gross Domestic Product," which is translated as "Gross Domestic Product" in Chinese. The translations of "Domestic" and "Production" are fine, but the translation of "Gross" as "Total" is somewhat problematic.
The word "Gross" in English does not strictly mean "total" in a data sense; the mathematical term for total in English is "Sum." So what does "Gross" generally refer to in the field of economics? It actually corresponds to the Chinese concept of "毛" (gross), while "Gross" corresponds to "Net," which means "净" (net). For example, "毛利润" (gross profit) corresponds to "Gross profit" in English, while "净利润" (net profit) corresponds to "Net profit." Therefore, it is more accurate to say that "Gross" in English refers to the concept of "gross profit" rather than "total."
So what is gross profit? Generally speaking, it is the income you receive from selling something after deducting the cost of the raw materials used to make that thing. This is the closest to gross profit. Of course, when we say "closest," it does not mean "exactly equal," but the number for gross profit is closest to the value of a product after deducting the cost of raw materials.
This actually highlights that the concept of Gross is value added. For example, if you process 5 yuan worth of raw materials and sell it for 15 yuan, then your value added is 10 yuan. The 5 yuan is called intermediate products, which is also what most textbooks will tell you; GDP accounting means that intermediate products must be deducted. So a more precise translation of GDP should actually be "the value added from domestic production activities."
This distinction is important because in economic data statistics, we often see a term called "total output." When we first established the country, we did not calculate GDP; we calculated total output. You can see that many places write data on the economic history of the republic, using total output during the planned economy era, not GDP. Moreover, you will find that many state-owned enterprises still calculate total output. Total output does not deduct the value of intermediate products. If you produce 10,000 yuan worth of refrigerators, then the total output is 10,000 yuan, regardless of how much you spent on raw materials, even if you used 20,000 yuan worth of steel; if you finally produced a refrigerator worth 10,000 yuan, your total output is still 10,000 yuan.
Therefore, compared to total output, it is clear that measuring value added, which is GDP, is more reasonable because total output has a lot of double counting issues and is a very rough measure of production activities. But here we can also see that such a rough measurement method was a key data point for a country to make economic decisions just a few decades ago, so think about how crude economic management was at that time; this highlights the importance of measurement.
Of course, even internationally, the popularization of GDP statistics only occurred after World War II.
- Where is the mistake in the phrase "Increase the proportion of consumption in GDP"?
I have just explained the difference between the concepts of total output and value added, and it is clear that the GDP statistics now are much more reasonable than the total output figures from many years ago. The second issue I want to discuss regarding GDP, which I have mentioned in previous podcasts and many occasions, is that GDP statistics refer to production activities, not the buying and selling activities in the market. This issue mainly points to the common error in the phrase "increase the proportion of consumption in GDP."
Let me briefly repeat this logic: GDP measures production, and the so-called division between consumption and investment refers to how many consumer goods and investment goods are produced. If you want to increase the production of consumer goods, you must simultaneously build factories and production lines, which must be included in the investment statistics. Therefore, there is no situation in the world where you can separately increase the proportion of consumer goods because increasing the production of consumer goods will necessarily increase the production of investment goods. The output of consumer goods can increase, but the proportion of consumer goods in GDP cannot increase. This is because increasing the absolute amount of consumer goods will necessarily increase the absolute amount of investment goods, so the only way for the proportion of consumer goods to increase is if the reduction in investment goods production leads to a passive increase in the proportion of consumer goods. The high proportion of consumption in developed countries is actually due to this reason.
So here we need to realize that the so-called increase in the proportion of consumption in GDP has nothing to do with the increase in the consumption capacity of ordinary people. The increased consumption of ordinary people is reflected in the retail data, which is the social retail consumption data. Moreover, if you look at that data, you will find that when our retail data is high, it coincides with a high proportion of investment in GDP. Therefore, this erroneous statement that has circulated online for many years is still not understood by many so-called professionals, who have not carefully examined the correlation between retail data and the proportion of consumption in GDP, and have blindly followed the crowd.
- How is GDP calculated?
Next, let's discuss the third issue: how is GDP calculated? Many people may be curious about where this data comes from and how all the production activities of such a large country can be accounted for. In fact, it is impossible to account for everything; a lot of data is estimated, which we will discuss later.
First, let's talk about the original data sources. One source is the survey data from the statistical bureau, which conducts surveys across various industries, many of which are sample surveys. The second source is the data managed by various ministries of the State Council, such as the Ministry of Transport managing the transportation industry, and the central bank or financial regulatory authority managing the banking industry. They certainly collect a lot of data from their respective industries. The third source is fiscal data, which goes without saying; data on fiscal revenue and expenditure is also important original data.
We know that there are three methods for calculating GDP: production method, income method, and expenditure method. In practice, we mainly use the income method and expenditure method for statistics; only one industry currently uses the production method, which is agriculture. This is mainly because the production method is too cumbersome in practice, as it is too complicated to account for intermediate inputs. In agriculture, there is basically no intermediate input; it is all labor-intensive, so it is relatively easier to measure.
All other industries, except agriculture, use the income method for statistics. The income method is particularly simple; it just requires enterprises to fill out a form, and you can even adjust the original financial statements of the enterprises without requiring them to fill anything out. You mainly need to find the figures for labor compensation, production taxes, depreciation, and net profit from the financial statements, and summing these gives you the income method statistics. Therefore, we say that the income method is the simplest and most convenient statistical method in practice.
In the income method, people may wonder how to account for those individual businesses, street vendors, fried rice sellers, and scallion pancake sellers, who do not have financial statements. How do we account for them? Right, we cannot directly account for that part; it is all estimated.
How to estimate depends on the characteristics of each industry. For example, in the construction industry, there are many unqualified enterprises, even individual businesses, contracting teams, and some organized crime groups. So what to do? The statistical bureau may conduct a sample survey, randomly select a few areas, and see the proportion of value added from qualified enterprises in the entire industry. The workload of sample surveys is not that large, and they can also survey the situation of small vendors, then calculate a proportion, which will be used to estimate the GDP of the entire industry, including those small vendors and individual businesses.
Of course, in reality, many times the proportions are obtained directly from economic census data, without the statistical bureau needing to conduct special sample surveys. Our five-year economic census data is meant to supplement the difficulties of daily statistics.
So how much of this indirect estimation accounts for GDP? We currently have data from 2013, where the indirect estimation data accounted for about 45% of total GDP, meaning that half of the GDP is actually unaccounted for and is estimated. Here I want to point out that the more a country's economy is composed of a few large enterprises, the higher the accuracy of economic data. If a country is made up of small vendors and individual businesses, you have to rely on estimates to determine GDP, and the less developed the economy, the less accurate the data itself will be. This is actually a very interesting point.
Next, let's discuss the expenditure method. The expenditure method is what we often refer to in terms of how much consumption and investment account for GDP. The data on household consumption is entirely based on survey data, and it is a sample survey, so there is a significant component of estimation in this data. The investment and government consumption data are relatively more accurate; the government data goes without saying, while investment can mainly be obtained from the financial statements of enterprises, so investment is relatively accurate, while consumption may be less accurate.
Theoretically, the GDP figures obtained from the three methods should be the same, but in reality, there will definitely be some discrepancies. What is the margin of error? I calculated that in 1980, the discrepancy between the GDP figures from the income method and expenditure method in our country was around 1% to 2%. Since 1995, the GDP statistical error has stabilized at within 0.5% to 1%. Doesn't this align with our conclusion that the more developed the economy, the more accurate the GDP statistics?
Moreover, friends who have listened to my podcast on local debts may remember that I mentioned that 1995 was a very important year because from that year onward, our country's economy was considered to have finally transitioned from a supply shortage state to a demand shortage state, as a survey found that the queuing phenomenon for most retail goods had disappeared.
- Does the sum of regional GDP equal national GDP?
Next, the fourth issue, which is a topic of ongoing debate, is that the sum of regional GDP in China does not equal national GDP. Each province used to calculate its own GDP, while the National Bureau of Statistics was responsible for calculating national GDP. Both sides calculated their own figures, and the total figures reported each year were different, which has been a significant reason for many people to question the authenticity of China's statistics. For most of the past time, the sum of provincial GDP was higher than national GDP. How big can this discrepancy be? At its peak, between 2013 and 2014, the difference was 4 trillion yuan, accounting for 6% to 7% of that year's GDP, which is quite significant.
It goes without saying that there are definitely issues with local governments exaggerating their GDP for political performance. Tianjin is a typical example; the district where I was born has always claimed to have the highest per capita GDP in China, but I have never felt that way. Later, after adjusting for inflated GDP, it turned out that nearly 40% was removed.
This issue was basically resolved in 2019. In 2019, GDP statistical reforms were implemented, and local GDP is no longer self-reported; it is all calculated by the National Bureau of Statistics, with cooperation from provincial statistical bureaus. Therefore, after 2019, it can be seen that this discrepancy has become very small, now possibly less than 0.5%.
Interestingly, after the 2019 statistical reform, we found that not all local GDP figures were inflated. In economically developed areas, such as Shanghai, Fujian, and Guangdong, GDP was adjusted upwards; these wealthy areas had reported lower GDP figures intentionally, while poorer areas had reported inflated figures, generally overstating GDP.
Why does this happen? I personally feel it is still a political logic. The leaders in economically developed provinces are mostly those who are already certain to be promoted in the future, so their role in developed provinces is more about "polishing their resumes." Therefore, they do not need to improve economic development to have a bright future. However, in provinces with poor economies, if you can improve the economy even a little, it can change the future of those who originally had no bright prospects, opening up upward mobility.
We have talked about HNA; the two team leaders took on a huge mess that everyone knew was difficult to handle, but in the end, they succeeded and became the youngest vice governors. Those who can tackle tough challenges are the ones who get political opportunities; this goes without saying.
- Special project statistics in GDP
The fifth issue is the statistics of some special projects in GDP. This is a bit more complex, but I think it is also very important, so I will discuss it. First, there are intangible assets, such as research and development activities. Are these included in GDP? In 2017, China began to include R&D expenditures in GDP, which were not counted before, reflecting our emphasis on R&D. Therefore, there was a significant adjustment in GDP in 2017, and after the adjustment, the figures increased.
Then, we know that if you work in Beijing and rent a house, the rent must be included in GDP. However, if you live in a house you bought and do not pay rent, does that mean it is not counted in GDP? Actually, no. At this time, we will assume that you pay rent for your own house, which is also included in GDP. This is a concept that often appears in statistics, called imputed income.
Because GDP essentially measures all production and service activities in a country, even if you do not spend money renting your own house, you are essentially providing yourself with a rental service. The rent you save is the compensation for the service you provide. Therefore, even if you do not spend money on the surface, you still create a service value through your own house, so it must also be included in GDP.
Finally, there is a currently highly debated issue: Do household activities performed by family members count as GDP? We currently include services provided by housekeeping companies in GDP, which goes without saying. So, according to similar logic to rent, should household chores done by family members also be included in GDP?
This is currently a major controversy. Most countries do not include this, but including myself, those with more radical views believe that this should be included in GDP, even if the amount is low, but it should not be completely excluded. Because this reflects our society's view of what constitutes valuable activities, and many modern social issues may stem from this. Have you ever thought that perhaps we could solve the low birth rate issue through a statistical adjustment? Of course, this is more of a mixed issue of sociology and economics, so I won't elaborate further.
- How to analyze the GDP figure itself?
I believe I have covered some of the GDP statistical issues that people are relatively concerned about. Finally, in the sixth issue, let's discuss how to analyze the GDP figure itself.
In fact, regarding the total analysis of GDP, I think the key is to first find out what the current economic growth rate central tendency of our entire economy is. Simply comparing GDP figures is meaningless. Our current economic growth rate is at 5%, and many people are shouting about a major recession, but if you look at Japan, which also had a growth rate of 5% last year to this year, it is considered to have emerged from economic recession. So you see, looking at an absolute growth rate alone is not useful; the first point in GDP analysis is to understand that different economies are at different stages, so their economic growth rate central tendencies are different.
Combining historical data from various countries since the Industrial Revolution, especially after the widespread adoption of GDP statistics post-World War II, we can see that the modernization process of an economy generally goes through three growth rate central tendency stages. The first stage generally has a growth rate central tendency of above 8%. We can see that countries like the United States and Japan have experienced continuous growth rates above 8% for over ten years. Of course, the early data from the United States, as an old capitalist country, is certainly estimated. However, we note that if an economy can maintain a GDP growth rate above 8% for ten consecutive years, it can be considered that the modernization process has begun.
Here I want to mention certain late-developing countries that are often thought to be on the verge of rising, such as India. In the past two years, India's GDP growth rate has approached 8%, and there have been constant discussions about India possibly being the next China. However, if you look at India's historical data in full, you will find that from the 1960s to now, its growth rate peak has been around 8%, while its low points are generally around 3% to 4%, sometimes even 1%. So you can see that its growth rate central tendency has never exceeded 8%, remaining around 5%, with significant fluctuations every few years.
This growth rate characteristic is clearly not a state of modernization initiation. With such a low base, India's per capita GDP is still only one-fifth of China's, and it remains in a state of being an agricultural country or a rural-urban fringe. In such a state, if the growth rate central tendency is only around 5%, it is fundamentally impossible for the economy to take off. But why do people keep saying that India's economy is about to take off every few years? It is because of a kind of superficial data comparison, taking India's cyclical high point of 8% growth as a talking point, but a few years later, it drops to 3% or 1%. So comparing in this way is completely meaningless.
Based on historical experience, a growth rate central tendency of over 8% indicates the initiation of modernization or the opportunity to enter the ranks of developed countries. For China, this growth rate central tendency was around 9% from 1978 to 2012, maintained for a very long time, which is indeed impressive.
The second growth rate central tendency stage generally ranges from 8% to 5%. By this time, basic modernization has been completed, and the supply shortage issue has been resolved; the economic problems are more about demand deficiency. We can see that Japan entered the second growth rate central tendency stage after the oil crisis in the 1970s.
I want to say that while many analyses discuss why China's economic growth rate has fallen below 8%, I believe the most important issue is natural growth; you cannot maintain such a high speed forever. From a quantitative comparison perspective, we can note that China has actually entered the second stage with a growth rate around 5%. To accurately correspond to Japan, it should be aligned with Japan's growth rate central tendency after the oil crisis in the 1970s. However, many people now tend to correlate China's economic recession with Japan's post-1990 bubble burst, which is a problem of emotional cognition.
Because if you start from an emotional perspective, every time you encounter an economic downturn, you will feel that the economy will not be good in the future. Since 2003, there have been people saying that the Chinese economy is failing, and at that time, Japan had not yet been in recession for 30 years. If Japan had already been in recession for 30 years at that time, perhaps in 2003, people would have started to think that China would become like Japan. This is all due to discussing economics without quantitative analysis. Therefore, you should see that even if you superficially compare China and Japan, you should compare China with Japan's economic recession in the 1970s, not the recession in the 1990s.
The third growth rate central tendency stage generally has a GDP growth rate of around 2% to 3%. This is typically represented by the United States, and China has not yet reached that stage, so analyzing it now is not particularly useful.
After discussing this part, the principles of total analysis of GDP are mostly covered. I believe that the analysis of GDP growth rates serves one purpose: to tell us what major phase the economy is currently in. During this major phase, we will experience short-term fluctuations around this central tendency; it may be lower this year and higher next year, forming what is known as short-term economic cycles. However, based on experience, short-term economic downturns generally last for 2 to 4 quarters, which may not even constitute a complete GDP statistical year. Therefore, we ultimately find that GDP annual data in various countries tends to decline continuously, which is why I believe GDP has little significance for short-term economic analysis; GDP helps us locate what stage of economic development we are in long-term.
This concludes the total analysis of GDP. What about the structural analysis of GDP? This involves analyzing the proportions of consumption, investment, and net exports, as well as various industry analyses. This part is certainly important, but it is more about analyzing specific data, which is not the content of this podcast episode.