In recent years,Consumption downgradeBut what exactly is “consumption downgrade”? How can we see “consumption downgrade” from the data? It is not easy to explain this.

▍What is considered “consumption downgrade”?

The most commonly used argument for "consumption downgrade" is the data released by the Statistics Bureau.Total retail sales of consumer goodsThe following table lists the consumer retail situation in Shanghai in the first half of 2024. All categories have negative growth compared with the same period last year.

Beijing's total retail sales of consumer goods also saw a similar year-on-year decline.

Can such a decline in the total retail sales of consumer goods be considered a form of "consumption downgrade"?

uncertainAs part of the young working population has returned from first-tier cities to third- and fourth-tier cities in recent years, they have also taken their consumption capacity with them.While consumption in Beijing and Shanghai declined, consumption in central and western provinces increased, and consumption at the national level remained in a state of positive growth.。Therefore, the decline in consumption in Beijing and Shanghai may be seen as a kind of "Consumption Shift"——People are still consuming, they just shifted their consumption elsewhere. The total amount nationwide may have even increased, so it is not considered a downgrade.

It's not just macroeconomic data, even in some seemingly small observations, "consumption shifts" often make it difficult to identify the real situation.

For example, if a store used to be packed with customers but now has few customers and its turnover has dropped, is this because of "consumption downgrade"? Not necessarily. It may be that customers who used to frequent this store have now gone to other places. This is just aConsumption Shift。

For example, a regular customer of a store used to spend 50 yuan each time he came, but now spends 30 yuan each time he comes. Is this "consumption downgrade"? It is more likely, but there are exceptions. For example, a store selling similar products opened nearby, and this old customer spent the other 20 yuan in the other store after comparison. This is also "consumption downgrade".Consumption Shift。

So if a regular customer of a restaurant comes to eat lunch on time every day at noon, and previously spent 15 yuan each time, but now spends 12 yuan, is this "consumption downgrade"? Of course, the possibility is greater, but there may also be "consumption transfer". For example, he may have originally had a cup of 3 yuan iced black tea when dining in this restaurant, but now chooses to bring his own iced black tea - even if the iced black tea is bought from the convenience store next door. This is stillConsumption Shift。

Even for micro-consumption data, as long as there is room for division and transfer of consumer demand, "consumption transfer" may occur, which may be misunderstood as "consumption downgrade". This is caused by the complex characteristics of consumer behavior.

But what if the following situation occurs?

Situation A: An old customer of a noodle restaurant comes to the restaurant on time for lunch every day at noon. He used to spend 15 yuan each time, of which 10 yuan was a bowl of noodles and 5 yuan was a chicken leg; but now he spends 12 yuan each time, of which 10 yuan is still a bowl of noodles and the other 2 yuan is for a braised egg.

Situation B: A regular customer of a coffee shop comes to the shop every afternoon on weekdays to buy a cup of coffee. Previously, he would buy a cup of latte for 18 yuan each time, but now he buys a cup of Americano for 13 yuan.(Latte requires milk, while Americano does not, and is generally cheaper)

When consumption data is accurate to the level of consumers, stores, and products, “consumption transfer” is not so easy to occur:You still need to add toppings to your noodles, but the toppings have become cheaper.People still drink coffee, but it is cheaper. The energy requirement of a lunch and the caffeine requirement of a cup of coffee are difficult to divide and transfer. It is hard to imagine that the customer who spent less money on coffee used the 5 yuan saved from switching from latte to Americano to buy another cup of coffee at another store. The customer who changed the chicken leg to a braised egg could not have used the 3 yuan saved to buy a takeout topping at another restaurant. The 5 yuan and 3 yuan reduction here are real expenditure reductions, which is a real "consumption downgrade."

So we can give the definition of "consumption downgrade" -When people are willing to pay a lower total price to meet specific needs that are indivisible and non-transferable, this is considered "consumption downgrade"。

▍How to calculate the degree of "consumption downgrade"?

However, such a strictly defined "consumption downgrade" index requires very high data precision. Is it really possible to find such data? Yes.

In this article, we refer to data from "Shou Qian Ba" and other sources to study China's consumption issues.

Collect money, yesChina's leading comprehensive digital store service providerIts ordering and payment systems cover a large number of restaurants and retail stores in all cities across the country.

The macroeconomic indicators such as social retail and consumption released periodically by the statistical department often focus onLarger scaleof enterprises, and Shouqianba covers a larger proportion ofSmall and micro business entitiesOn the basis of trend consistency with macro data, it is closer to people's daily life scenes and is a useful supplement to macro data.

The analysis example is as follows(All examples below do not contain consumer personal information)：

The table above shows a customer(Personal information has been anonymized)Consumption in a certain store. This customer has made purchases in this store more than 300 times. The above table only shows the most recent consumption and some earlier consumption.

It can be seen that between 11:30 and 12 o'clock every day, this customer will have a lunch in this store. At the beginning of 2023, this customer would spend 10 to 12 yuan per meal, including a bowl of pickled mustard and shredded pork noodles, and sometimes a bottle of Master Kong iced black tea. In the last few days of July 2024, this customer only spent 7 to 9 yuan a day. On July 24, he consumed fried egg noodles. From July 25 to July 28, he consumed green vegetable noodles for only 7 yuan. Obviously, this customer has experienced a significant "consumption downgrade" in this restaurant.(Interestingly, the menu prices of this store have not changed in the past few years)

Next, we calculate the consumption upgrade/downgrade index of this customer in this example through these steps:

1，Keep this customer's lunch consumption between 11:00 and 12:00 every day to ensure that each day's consumption is to meet the same comparable demand.

2，Remove drinks, etc. and only keep staple food consumption - as we mentioned earlier, consumption of retail bottled drinks, etc. may be transferred to other stores. Removing them can more accurately calculate consumption upgrades/downgrades.

3. Calculate the average price of the customer's lunch staple food in this store every day. Divide the customer's daily staple food consumption price by the average price of the customer's consumption in this store to getThe degree to which this customer's daily consumption is higher or lower than the average, that is, the customer's "consumption upgrade/downgrade" index in this store。

The customer's dining consumption upgrade/downgrade at this restaurant from December 2022 to July 2024 is shown in the following figure:

In the above figure, values ​​above 0 mean that the customer's staple food consumption during this period is higher than the average, which is an upgrade; values ​​below 0 mean that the food is cheaper, which is a downgrade.

It is not difficult to see that this customer's dining habits have fluctuated to a certain extent, but before April 2024, except for the Spring Festival, this customer's consumption is becoming more and more "upgraded". In April 2024, his consumption upgrade has reached 0.12 - that is, this month's consumption is 12% higher than the customer's average consumption in this store.

However, starting from May 2024, this customer's consumption began to decline significantly. From May to July 2024, this customer's consumption was lower than his average consumption. In May, it was about 25% lower. In June, it rebounded slightly, only 5% lower than the average. In July, it fell again, 12% lower than the average.

Through this method, we can see each person's consumption upgrade/downgrade more accurately.

Using the same method, we calculated the consumption upgrade/downgrade index of all consumers in the sample in all stores, and then took a weighted average according to the number of consumption orders to obtain a "consumption upgrade/downgrade index" for a larger range. The national index and its year-on-year changes are shown below:

It can be seen that the national consumption upgrade/downgrade situation(Blue line)It reached its highest value in January 2023, and then fluctuated downward. It has been below 0 for four consecutive months, which means that in the past four months, the average price people paid when dining at a certain restaurant was lower than the average price they paid at this restaurant in the past three years.

In addition, the index has a clear seasonal trend and is always higher during the Spring Festival. If we calculate the year-on-year change of the index(Orange Line), observe its changing trend, then we can see that fromOctober 2023Start toJuly 2024，The year-on-year difference in the national consumption upgrade/downgrade index has been below zero for 10 consecutive months.。

This means that consumers' consumption in the catering and retail sectors is indeed declining, and the degree of decline is deepening. More importantly, since February this year, the decline has been getting bigger every month.Not only are consumers spending less, they are also spending more。

The results by city are shown in the figure above. Among the 32 cities we selected from across the country, only seven cities saw an upgrade in consumption compared with the first half of last year, including Hefei, Chongqing, Wenzhou, Jinan, Zhaoqing, Nanning and Nanjing. The other 25 cities all experienced "consumption downgrades" of varying degrees, including first-tier cities such as Beijing, Shanghai, Guangzhou and Shenzhen, as well as a large number of second- and third-tier cities.

Judging from the data, "consumption downgrade" is no longer a vague generalization of a certain phenomenon, but a fact that is actually happening.

▍What is “consumption downgrade” related to?

What factors are related to "consumption downgrade"? The most obvious factor is naturallyincomeAs the saying goes, "live within your means." When income is low, people's consumption will decrease accordingly.

Use data groupPrevious articlesBased on the recruitment data mentioned in the previous article, we can calculate the average salary of recruitment in different cities in different months to measure the average income in different cities at different times. So, what is the impact of income on "consumption downgrade"?

The following figure shows the impact of wage changes in different lagged months on the local consumption upgrade/downgrade index. Its meaning is as follows

——The horizontal axis represents the number of lagged months. A horizontal axis of 0 means that the regression is based on the month-on-month change in the average salary of each city.Current monthThe consumption upgrade/downgrade index, the meaning of this coefficient is the change in wages andCurrent monthIn the right half of the picture, the part with a horizontal axis greater than 0, such as 1 for the lag month, indicates that the wage change andnext monthThe grey area indicates the 95% confidence interval.

As can be seen from the figure above, the only place where this coefficient curve is significantly greater than zero is when the lag period is 0 months, and it is also close to being significant when the lag period is 1 to 2.Wage changes have a significant positive correlation with consumption upgrades and downgrades in the current month, and a nearly significant negative correlation with consumption upgrades and downgrades with a lag of 1 to 2 months.。

Therefore, wages are indeed related to consumption upgrades and downgrades, but this correlation is relative.short termYes, if wages increase this month, we will only observe an upgrade in consumption in that month, but it will not last and will be discounted by the nearly significant reverse effect in the next two months.

This is also a normal psychological state. When income increases in a month, many people will choose to spend more money. However, if this income increase is not sustainable, then The guilt of "why do I have to spend all the money as soon as I get it" will actually cause the person to spend less in the next two months.

So, in addition to income, what other factors may affect consumption? We tried two important indexes:Housing Prices、share price。

Let’s look at housing prices first. Using the monthly housing price data of 70 large and medium-sized cities released by the National Bureau of Statistics, and regressing the “consumption downgrade” of each city, we can get the following chart:

It can be seen that compared with the correlation between wages and consumption,The correlation between housing prices and consumption is more persistentWhen house prices rise this month, not only will consumption this month be upgraded, but consumption in the next month and the month after that will also be significantly upgraded.

Then the reverse effect is easy to understand.If housing prices in a city fall this month, there is a high probability that this city will experience a "consumption downgrade" for 2-3 months. When housing prices fall by 1% month-on-month, people's consumption will decrease by 5.93% in the next three months.

A similar effect also appears in the stock index. We use the daily rise and fall of the Shanghai Composite Index to regress the daily consumption upgrade/downgrade index, and we can get the following figure:

It can be seen that there is a significant correlation between the rise and fall of the stock index and consumption upgrade/downgrade, and its significance will last for three days - when the stock price rises, people's consumption will be upgraded on the day of the rise, the second day and the third day. And this effect is also true in reverse.When the stock index falls, the average consumption level of the whole nation will drop significantly on the first, second and third day of the decline.。When the stock index falls by 1%, the national consumption level will drop by about 0.57% in the next three days.

▍Faith is the most precious

In the past three years, the trend of housing prices and stock indexes is self-evident, and the growth rate of income has also slowed down. From the recruitment data we have captured on the entire network, we can see that the average salary of the market recruitment positions peaked in July 2023 and has been in a state of slow decline in the past year.

Which of the "consumption downgrade" effects produced by different channels is bigger or smaller? We multiply the correlation coefficients between consumption upgrade/downgrade and income/housing price/stock price calculated in the previous article by the real changes in the data of these three dimensions in recent years, and we can roughly estimate the size of the "consumption downgrade" effect that may be produced by different channels. Putting the three together, we can get the following figure:

As can be seen from the above figure, house prices have fallen.Stock indexes fallThe "consumption downgrade" effect brought about by the decline in recruitment income is dozens of times greater than the "consumption downgrade" effect brought about by the decline in recruitment income.Falling asset prices may be one of the main causes of “consumption downgrade”。

This reminds me of the discussion about the relationship between "housing prices" and "consumption" in the past decade: for those who have not yet bought a house, the view that "rising housing prices will squeeze out consumption" is widely accepted, and even a conclusion is drawn from this - "only when housing prices fall will people consume." In fact, the correlation between the two observed in the data is just the opposite - there is no "crowding out effect" on consumption when housing prices rise, but consumption is significantly downgraded when housing prices fall.

Why does this phenomenon occur? One explanation in academia is the “wealth effect”, which means that when a family’s main asset(In China, it is most likely a house)When prices rise, the paper wealth generated expands, making people more willing to spend money. It is not easy to prove the causal relationship between "paper wealth" and consumption academically. We prefer to interpret housing prices and stock indexes as the external manifestations of another indicator, which is correlated with consumption.

——Confidence in the future。

In theory, house prices and stock indices are both discounted future returns - house prices are the returns from rent, while stock indices are the returns from dividends.

Only when people believe that "tomorrow will be better" and that there will be higher returns in the future can they discount the future to a higher value, and house prices and stock indexes will rise.

Only when people believe that future income will have greater sustainability and room for stable growth can consumption return to the upgrading range. As we have seen before, the impact of short-term income increases on consumption is very short-lived. Only long-term stable income growth expectations can produce a sustained impact on consumption.

Only when people regain confidence in the future can vegetable noodles become shredded pork noodles again, American coffee can become latte again, and what was downgraded to a topping for braised eggs can be upgraded back to chicken legs.

Confidence is the most valuable thing. How can we make people willing to believe in the future? Perhaps we should first "correctly understand the present."

What is the last thing you need at this moment? Perhaps it is to find a seemingly "reasonable" excuse for the objective facts that already exist. For example, "The reason why this person replaced the chicken drumstick topping with a braised egg is not because of "consumption downgrade", nor because he has no money, nor because he has no confidence in a stable income in the future, but because - he wants to lose weight."