Towards a Win-Win Solution for Dietary Health and Carbon Reduction—Evidence from the Yangtze River Delta in China

Abstract

Considering the contradiction between the need to change the food consumption structure of Chinese residents and the constraints of resources and the environment, as well as the changes in the consumption structure of Chinese residents in the Yangtze River Delta, we explore the path to achieve environmental sustainability while maintaining residents’ dietary health. Based on 1995–2019 Yangtze River Delta food consumption data, this paper uses the two-stage Engel–QUAIDS model to conduct an empirical analysis of the food consumption and carbon emissions of urban and rural residents in the Yangtze River Delta and simulates the impact of income growth and food price changes on per capita food consumption carbon emissions and nutrient intake. The results show that the residents of the Yangtze River Delta consume too much meat and poultry, and the carbon emissions are high; the consumption of eggs and fruits is obviously insufficient, and the carbon emissions are low. With an increase in income, the increase in food carbon emissions among rural residents (0.406%) is greater than that among urban residents (0.247%); higher prices of meat, poultry, and aquatic products can significantly reduce food carbon emissions, and higher prices of fruits will promote food carbon emissions. The nutritional intake of residents can still be guaranteed under the low-carbon policy. It is worth mentioning that after the price adjustment simulation, residents’ fat intake will be significantly reduced within the recommended range, which is also beneficial to residents’ health. Therefore, appropriately regulating food prices and increasing people’s income would not only ensure nutritional health but also contribute to reducing carbon emissions and creating a sustainable agricultural food system.

1. Introduction

Since China’s reform and opening up, and the corresponding economic development and increase in national income, the Engel coefficient of urban and rural residents in China has declined significantly, and resident food consumption is gradually transitioning from subsistence consumption to development-oriented consumption [1]. A scientific research report on dietary guidelines for Chinese residents (2021) showed that people in China have sufficient intake of dietary energy and macronutrients, and the total food consumption has significantly increased and presented a trend of diversification. As the report noted, Chinese residents’ food consumption structure is changing from “simple food” to “high oils and high fats”. On the one hand, changing residents’ food consumption structure can lead to increasingly prominent dietary imbalances among residents, as well as nutrition-related chronic diseases [2]. On the other hand, this measure may increase food carbon emissions and affect the sustainability of resources and the environment [3]. How to maintain a healthy national diet while considering the sustainability of the environment has become a practical problem that must be solved urgently. Under this background, it is important to solve the above problems to explore a sustainable food consumption model that not only meets nutritional and health needs but also minimizes negative externalities on the social environment and adjusts the food consumption structure of residents [4,5]. The economic development level of the Yangtze River Delta holds a leading position in China; the contradiction between the improvement demands of the food consumption structure among residents in this region and the restrictions of resources and the environment is more prominent that in other regions. In addition, the Yangtze River Delta has entered a new phase of common prosperity, and the income gap between urban and rural residents in this region is smaller than that in other regions of China. Therefore, studying the differences in food consumption structures between urban and rural residents can provide a simulated effect that promotes the integrated development of urban and rural areas, thereby realizing common prosperity.

2. Literature Review and Theoretical Framework

2.1. Research on the Change of Food Consumption Structure of Chinese Residents

Over the past 30 years, the food consumption structure of China has generally manifested a relatively stable change in plant food consumption, while animal food consumption has increased year by year [6]. The increase in the consumption of meat, poultry and fruits will eventually crowd out vegetables, dairy products, and other food, which currently occupy a relatively low proportion of total food consumption and may aggravate the imbalance of the dietary structure of residents [7]. Although the food consumption level gap between urban and rural residents in China is narrowing, there are still significant differences. With the improvement of the quality of life, urban residents are more inclined to eat out and give more attention to dietary structure and personalized needs. Although the consumption of meat, poultry, and eggs by rural residents is increasing rapidly, it remains at the level of meeting basic needs [1]. Factors influencing food consumption mainly include income [8], price [7], urbanization [9], and family structure [10]. To study the impact of different factors on food consumption, scholars usually adopt the demand system model. At the same time, with the evolution of the demand system model, many scholars choose the two-stage demand analysis framework to study the change in food consumption among different income groups of residents [8,11].

2.2. Research on Health Effects of Food Consumption

Food consumption is closely related to people’s nutrition and health. The nutrition and health status of residents is an important indicator that reflects the economic and social development of a country and the health quality of the population [12]. Huang discussed the relationship between the determinants of consumer food choice and consumer nutrition availability [13]. At the end of the last century, food demand in China’s poor areas was very elastic, with an average value of about 0.74, while nutrition demand was relatively inelastic, with an average value of about 0.14. Therefore, in poor areas of China, the increase in food consumption did not bring the same improvement in nutritional status [14]. However, in recent years, with the deepening of livelihood projects, such as the construction of a healthy China and poverty alleviation through health, the nutritional and health status of residents has been significantly improved. Nevertheless, there are still major nutritional health problems, such as inadequate intake of some foods and increasing rates of overweight and obesity. For this reason, some scholars have noted that dietary structure adjustments based on plant foods should be adhered to in order to optimize the consumption structure of animal foods, mainly by reducing meat and poultry consumption and increasing the consumption of aquatic products to increase the intake of high-quality protein such as fish and provide a protein-rich diet [2].

2.3. Research on Environmental Effects of Food Consumption

Food production and consumption are necessary to maintain a healthy diet among urban and rural residents and represent an important factor for the rapid growth of agricultural carbon emissions. In recent years, domestic and foreign scholars have studied the carbon emissions and carbon footprint of the food consumption of residents. Salo et al. studied the relationship between household expenditures and carbon footprint in Finland [15]. Kanemoto et al. argued that red meat, processed meat, dairy products, rice, and other foods contributed significantly to food carbon emissions [16]. Wei et al. explored the future path of agricultural development in China toward 2060 under a dual carbon goal. The authors maintained that hastening the technological progress in agriculture is one of the most efficient ways to maintain domestic food security [17]. In addition, respondents on high-GHG diets had higher intake levels of saturated fats and sodium, as well as lower quality diets [18,19]. The per capita food carbon consumption of urban and rural residents showed a significant duality, while the food carbon consumption in rural areas converged with that in urban areas [20]. When calculating carbon emissions from food consumption, scholars usually adopt the carbon conversion coefficient method and input-output method for analysis [21,22].

Considering the dual goals of a balanced diet and low carbon life, it is particularly necessary to carry out research on low-carbon food consumption that can consider a balanced diet. However, within the existing literature, empirical research combining these two aspects is scarce. Therefore, this study starts with food consumption data of urban and rural residents in China’s Yangtze River Delta from 1995 to 2019, adopts the two-stage demand analysis (Engel–QUAIDS) model for empirical analysis, and then simulates the impacts of changes to residents’ income and prices on food consumption carbon emissions and nutrient intake.

3. Materials and Methods

3.1. Data Sources

3.1.1. Data on Food Consumption

The differences in food consumption between urban and rural residents are an ongoing problem, and many scholars are also studying this problem. These differences should be paid attention to by society [23,24]. In this paper, we use the consumption data of urban and rural residents in 7 categories of food (grains, meat and poultry, eggs, aquatic products, vegetables, oils, and fruits) in China’s Yangtze River Delta (Jiangsu, Zhejiang, Anhui Provinces and Shanghai Municipality) from 1995 to 2019 to study the above problems. The data on food consumption, consumer price index, and food expenditures were collected from the China Household Statistical Yearbook, China Statistical Yearbook, and provincial statistical yearbooks over the years. In this paper, the missing values for the consumption of some food types were treated using the linear interpolation method [25]. Price data used the CPI for each food category among urban and rural residents in various provinces as the price index, taking the price in 1995 as the fixed base for index processing [26]. For the data on food expenditure over the years, all varieties of food expenditure over the years were first subtracted from the CPI classification index, and then expenditures after subtraction were added together to obtain the total expenditures of the region in that year. Then, the expenditure share of all types of food was calculated.

This study adopted the two-stage demand analysis method (Engel-QUAIDS). In the Engel model, during the first stage, the explained variable is the logarithm of resident per capita food consumption expenditures, and the explained variable is resident per capita disposable income, which in our case was derived from the statistical yearbook of each province. In the second-stage QUAIDS model, in addition to explained variables and explanatory variables, control variables included regional, urban, and rural dummy variables. In addition, based on the methodology of Zhou and Li, we assume that the population dependency ratio affects the food consumption of residents; therefore, the elderly dependency ratio and child dependency ratio were added as demographic characteristic variables [27]. These data were obtained from the China Population and Employment Statistical Yearbook.

3.1.2. Data on Food Carbon Emissions and Nutrient Coefficients

This paper applied the carbon emissions coefficient in the food system carbon emissions database of Poore and Nemecek [28] and sorted the data (

Table 1. Carbon emissions coefficient of food.

Category	Name	Scope of Definition	Carbon Emissions Coefficient (kg CO2e E/kg)
1	Grains	Rice, flour, rice, wheat, etc.	2.60
2	Meat and Poultry	Pork, beef, mutton, poultry, etc.	16.27
3	Eggs	Eggs	4.67
4	Aquatic Products	Fish, shrimp, shelf, crabs, etc.	13.63
5	Vegetables	Vegetables	0.81
6	Oils	Animal oil, vegetable oil	5.29
7	Fruits	Melons and fruits	1.01

The average carbon emissions of grains are the total average of rice, grains, and other grains. Due to the large gap in the unit carbon emissions coefficients of pork, beef, mutton, and poultry, when unifying the carbon emissions coefficients of meat and poultry, we selected the average of the total carbon emissions of meat and poultry food consumption among urban and rural residents in the Yangtze River Delta in 2019 divided by the quotient of meat and poultry consumption as the unified coefficient.

).

The Chinese food nutrition list is widely used in food research and has a certain reference value [29]. On this basis, the nutrient coefficient was further calculated. The method of calculating the nutrition coefficient of each food group involves taking the average value of the nutrients contained in different subgroups of the same food group in the Chinese food composition list. Due to space limitations, this method is not expanded upon here.

3.2. Model Construction and Elastic Calculation

In this paper, the two-stage demand analysis method (Engel–QUAIDS) model was mainly used to analyse the elasticity of food demand. Then, based on calculating the elasticity of food demand, the carbon emissions and nutrient elasticity of the food system were estimated indirectly.

3.2.1. Engel Model

To analyse how the main food expenditures were affected by income, in the form of the model of the first stage, we adopted a simplified double logarithm Engel equation for analysis based on the methodologies of some other scholars [30,31]. The model is as follows:

$$\ln exp={\alpha }_0+{\alpha }_1\ln y+{\alpha }_2{(\ln y)}^2$$

(1)

where $\ln exp$ is the logarithm of resident annual per capita expenditures on major food; $\ln y$ is the logarithm of annual per capita disposable income; and ${\alpha }_0,{\alpha }_1,{\alpha }_2$ is the constant term to be estimated and parameter to be estimated. Then, the income elasticity is:

$$e_I={\alpha }_1+2*{\alpha }_2*\ln y$$

(2)

3.2.2. QUAIDS Model

Banks et al. added an expenditure quadratic term, namely, the QUAIDS model, on the basis of the traditional AIDS model according to the extended form of the PIGLOG preference [32]. This model can reflect the characteristics of nonlinear changes in the Engel curve. We selected the QUAIDS model for estimation, obtaining Formula (3) as follows:

$$w_{it}={\alpha }_{i0}+{\displaystyle \sum _{j=1}^n{\gamma }_{ij}\ln p_{jt}}+{\beta }_i\ln \frac{x_t}{a(P)}+\frac{{\mu }_i}{b(P)}{[\ln \frac{x_t}{a(P)}]}^2+{\displaystyle \sum _{j=1}^n{\eta }_{kjt}{Z}_{kjt}}+{\epsilon }_{it}$$

(3)

where $w_{it}$ represents the proportion of the consumption expenditures of food I in the consumption expenditures of the seventh food in year t; $x_t$ represents the total food expenditures of residents in year t, which is the sum of the food expenditures in 7 categories; $p_{jt}$ represents the price of the second food type in the first year; and $Z_k$ represents the demographic characteristic variable (k = 1, 2, 3 and 4 represent the elderly dependency ratio, child dependency ratio, regional dummy variable, and urban and rural dummy variable, respectively). Lastly, ${\alpha }_{i0},{\gamma }_{ij},{\beta }_i,{\eta }_{kjt},{\mu }_i$ is the parameter to be estimated, and ${\epsilon }_{it}$ is the error term.

$a(P)$ and $b(P)$ represent the price index, which is, respectively, defined as

$$\ln a(P)={\alpha }_0+{\displaystyle \sum _{\mathrm{i}=1}^n(}{\alpha }_{i0}+{\displaystyle \sum _j^n{\eta }_{kjt})\ln p_{it}}+\frac{1}{2}{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{\gamma }_{ij}\ln p_{it}\ln p_{jt}}}$$

(4)

$$b(P)={\prod }_{i=1}^n{p}_{it}{}^{{\beta }_i}$$

(5)

The QUAIDS model also satisfies the AIDS model constraints, where the summation constraint is ${\displaystyle \sum _i^{}{\alpha }_{i0}=1}$, ${\displaystyle \sum _i^{}{\gamma }_{ij}=0}$, ${\displaystyle \sum _i^{}{\beta }_i=0}$, and ${\displaystyle \sum _i^{}{\mu }_i=0}$; the homogeneity constraint is ${\displaystyle \sum _j^{}{\gamma }_{ij}=0}$; and the symmetry constraint is ${\gamma }_{ij}={\gamma }_{ji}$.

Then, the elasticity of income and expenditures and the price elasticity of Marshall demand for all types of food can be deduced (Equations (6) and (7)) as follows:

$$e_i=1+\frac{1}{w_{it}}[{\beta }_i+\frac{2{\mu }_i}{b(P)}\ln (\frac{x_i}{a(P)})]$$

(6)

$$e_{ij}=\frac{1}{w_{it}}\{{\gamma }_{ij}-[{\beta }_i+\frac{2{\mu }_i}{b(P)}\ln (\frac{x_i}{a(P)})]({\alpha }_{j0}+{\displaystyle \sum _{k=1}^n{\gamma }_{jk}\ln p_{kt})-\frac{{\mu }_i{\beta }_j}{b(P)}{(\ln \frac{x_i}{a(P)})}^2\}-{\vartheta }_{ij}}$$

(7)

In Formula (7), if $i=j$, then ${\vartheta }_{ij}=1$,$e_{ij}$ represents the self-price elasticity of food; if $i\ne j$, then ${\vartheta }_{ij}=0$,$e_{ij}$ represents the cross price elasticity of food. Here, income elasticity is expressed as ${\mu }_{i,I}=e_i*e_I$.

According to the endogeneity solution proposed by Blundell and Robin, we identified and addressed the endogeneity problem of total food expenditures in the QUAIDS model [33]. The specific steps included the fitting value of residual term $\widehat{v}$ obtained by estimating the reduced equation $lnx=M\tau +v$. Here, M is a matrix form ($M=[1,lnY,{(lnY)}^2,lnp,Z]$), which is composed of a series of explanatory variables, including resident income and the square term of income, logarithms of various food price indices, and a series of demographic characteristic variables (the same as the demographic characteristic variables in Formula (3)). Then, $\widehat{v}$ is substituted into Equation (3) for estimation. If the coefficient of $\widehat{v}$ significant, it indicates that the endogeneity of expenditure terms $\ln x$ and ${(\ln x)}^2$ in the QUAIDS model is solved.

3.3. Calculation of Food Carbon Emissions and Nutrient Elasticity

To calculate the price elasticity of food demand, the indirect estimation method of Huang and the method of Yang and Mu were used to calculate the price elasticity of the water footprint, and the price elasticity of carbon emissions from the food system was further calculated [13,34].

The food price elasticity of food system carbon emissions ${\delta }_j$ can represent the percentage of change in j food system carbon emissions caused by a 1% change in food prices. The formula is as follows:

$${\delta }_j=\frac{{\displaystyle \sum _i^n{e}_{ij}*n_i*q_i}}{{\displaystyle \sum _i^n{n}_i*q_i}}$$

(8)

where i and j represent the food categories that are used to calculate carbon emissions from the food system, $e_{ij}$ represents the price elasticity of unconditional Marshall demand for i food to j food (the same below), $q_i$ represents the per capita consumption of i food, and $n_i$ is the carbon emissions coefficient. Similarly, the income elasticity of food carbon emissions ${\theta }_{}$ can be calculated as follows:

$${\theta }_{}=\frac{{\displaystyle \sum _i^n{\mu }_{i,I}*n_i*q_i}}{{\displaystyle \sum _i^n{n}_i*q_i}}$$

(9)

The price elasticity and income elasticity of food nutrients can be expressed via Formulas (10) and (11), where ${\sigma }_{wj}$ represents the price elasticity of food nutrients, and ${\pi }_w$ represents the income elasticity of food nutrients:

$${\sigma }_{wj}=\frac{{\displaystyle \sum _i^n{e}_{ij}*{\rho }_{wi}*Q_i}}{{\displaystyle \sum _i^n{\rho }_{wi}*Q_i}}$$

(10)

$${\pi }_w=\frac{{\displaystyle \sum _i^n{\mu }_{i,I}*{\rho }_{wi}*Q_i}}{{\displaystyle \sum _i^n{\rho }_{wi}*Q_i}}$$

(11)

where $w$ (nutrients) refers to energy, protein, fats, and carbohydrates; $Q_i$ represents the total edible portion of food; and ${\rho }_{wi}$ is the unit quality of food contained in the $w$ nutrient content, namely the food nutrient coefficient.

3.4. Simulation of per Capita Food Carbon Emissions Change and Nutrient Change

After the food carbon emissions income and price elasticity are obtained, the change rate of per capita food carbon emissions in urban and rural areas of the Yangtze River Delta can be considered under the condition that the two change separately or simultaneously, which can be written in the following form:

$$\%\Delta Q_1=\theta \%\Delta income+{\displaystyle \sum _{j=1}^n{\delta }_j}\%\Delta p_j$$

(12)

where $\%\Delta Q_1=\Delta Q_1*100\%/Q_1$ represents the percentage change in the annual per capita food consumption carbon emissions of residents; $\%\Delta income=\Delta income*100\%/income$ represents the percentage change in per capita annual disposable income of residents; and $\%\Delta p_j=\Delta p_j*100\%/p_j$ represents the percentage change in the price of food j.

The change rate of food nutrients per capita is shown in Equation (13), and the change percentage of food nutrients per capita per year is expressed as $\%\Delta Q_2=\Delta Q_2*100\%/Q_2$:

$$\%\Delta Q_2={\pi }_w\%\Delta income+{\displaystyle \sum _{j=1}^n{\sigma }_{wj}}\%\Delta p_j$$

(13)

4. Results and Discussion

4.1. General Situation of Resident Food Consumption in the Yangtze River Delta

On the whole, the total share of grains and meat and poultry expenditures is more than 50%, and with the improvement of urban and rural residents’ living standards in the Yangtze River Delta, the proportion of resident grains consumption decreased sharply. The proportion of eggs and oils expenditures is small, and the consumption trend is relatively stable. In terms of demographic characteristic variables, the mean elderly dependency ratio and child dependency ratio in the Yangtze River Delta were both lower than the national average in 2020 (23.8% for children and 17.8% for elderly people).

4.2. Resident Food Consumption Structure and Carbon Emissions

With the gradual improvement of urban and rural residents’ living standards in the Yangtze River Delta, the food consumption structure is also changing. Figure 1 and Figure 2 show the trend among food consumption of urban and rural residents in the Yangtze River Delta. With the improvements of living standards among urban and rural residents, the changes in grains consumption fluctuated greatly, showing a trend of decreasing first and then increasing, Additionally, the decrease in grains consumption among rural residents was found to be larger, which is consistent with previous research by other scholars [23,35]. The consumption of meat, poultry and fruits by urban and rural residents showed a rising trend, and urban residents consume more than rural residents. The main reason for this result is that with the progress of urbanization, the living standard of urban residents has become higher than that of rural residents. The consumption of vegetables largely remains stable at approximately 100 kg per person per year. Beyond the food categories mentioned above, the consumption of other foods by urban and rural residents was also found to be relatively stable.

As seen in

Table 2. Per capita food consumption and carbon emissions of residents in the Yangtze River Delta, 2019.

	Grains	Meat and Poultry	Eggs	Aquatic Products	Vegetables	Oils	Fruits
Food consumption	134.64	40.80	11.35	21.61	99.50	9.89	52.57
Urban	112.88	41.70	11.45	23.40	102.53	8.93	61.60
Rural	156.40	39.90	11.25	19.83	96.48	10.85	43.53
Recommended in guidelines	91~146	15~27	15~18	15~27	110~183	9~11	73~128
Food carbon emissions	350.06	663.82	53.00	294.58	80.60	52.30	53.09
Urban	293.48	678.46	53.47	318.94	83.05	47.21	62.22
Rural	406.64	649.17	52.54	270.21	78.14	57.40	43.97
Guidelines for carbon emissions	236.6~379.6	244.05~439.29	70.05~84.06	204.45~368.01	89.1~148.23	47.61~58.19	73.73~129.28

The food consumption unit is kg; the food carbon emissions unit is kg CO₂e.

, the meat and poultry consumption of residents in the Yangtze River Delta in 2019 exceeded the recommended amount in the 2016 Dietary Guidelines for Chinese Residents. Although the consumption of eggs and fruits also increased, the consumption of both still fall short of the recommended minimum per capita annual intake. Overall, the per capita food carbon emissions of residents in the Yangtze River Delta are similar to those of the food consumption, showing a trend of decreasing first and then increasing. However, in 2019, the total per capita food consumption carbon emissions of residents in the Yangtze River Delta totalled 1547.45 kg, which is higher than the sum of 1286.13 kg from the average food carbon emissions after conversion of the guidelines. Therefore, it is important to explore mitigation measures for food consumption. Meat and poultry account for a large proportion of food carbon emissions, and urban residents engage in more meat and poultry consumption carbon emissions than rural residents. Due to the low annual per capita consumption of oils and fats, overall the per capita carbon emissions of oils and fats were found to be equal to or even lower than those of eggs and fruits. Taking the food consumption of residents in the Yangtze River Delta in 2019 as an example, based on the food consumption and carbon emissions of residents in the region, meat and poultry consumption carbon emissions were placed on the list (more than the standard carbon emissions for this kind of food), and the annual per capita consumption was deemed excessive; therefore, they are classified as high-carbon foods. The consumption carbon emissions of eggs and fruits ranked low (with a gap from the normal standard), and the annual per capita consumption was found to be insufficient; therefore, these types of food are classified as low-carbon foods. The benchmark model in this paper is based on the recommended food intake in the dietary guidelines for Chinese residents and the recommended food consumption carbon emissions standard after conversion according to the carbon emissions coefficient, followed by the application of different policy simulations to meet the requirements of reducing food consumption carbon emissions.

4.3. Resident Food Consumption: Energy and Nutrient Intake

As shown in Figure 3, the nutrient intake of residents in the Yangtze River Delta showed a trend of decreasing first and then increasing, but the nutrient intake of residents fluctuated within the standard range. Possibly influenced by the gradual decrease in grains consumption, nutrition intake decreased first and then increased due to the improvement of residents’ living standards and the trend of diversified food consumption.

Table 3. Daily nutrient intake of residents in the Yangtze River Delta, 2019.

	Energy/kcal	Protein/g	Fats/g	Carbohydrates/g
Totality	2127.88	76.98	64.70	310.35
Urban	1964.55	73.66	62.44	277.59
Rural	2291.21	80.30	66.97	343.10
Recommended daily intake for residents	1800~2250	55~65	40~75	225~365

The recommended daily intake of residents is calculated by referring to the recommended intake for adults aged 18–49 in the guidelines and the conversion coefficients of energy and three macronutrients [36].

shows the daily nutrient intake of residents in the Yangtze River Delta in 2019. The per capita energy intake of urban residents was lower than that of rural residents. Except for the high intake of protein among urban and rural residents, the intake of other nutrients here is within the recommended amount, indicating that the current dietary structure of residents in the Yangtze River Delta is reasonable. The possible reason for the high protein intake is that residents in the Yangtze River Delta consume more aquatic products, which are rich in protein, reflecting the high quality of life among residents in the area.

4.4. Analysis of Food Consumption, Carbon Emissions, and Nutrient Elasticity

4.4.1. Analysis of Food Consumption Elasticity

Using Formulas (6) and (7), the Marshall own-price elasticity of food consumption of residents in the Yangtze River Delta can be obtained, as shown in

Table 4. Marshall own-price elasticity and income elasticity of food consumption in the Yangtze River Delta.

	Grains	Meat and Poultry	Eggs	Aquatic Products	Vegetables	Oils	Fruits
Marshall’s own-price elasticity
Totality	−1.132 ***	−1.396 ***	−1.724 ***	−1.671 ***	−1.490 ***	−1.177 ***	−1.506 ***
	(0.139)	(0.098)	(0.127)	(0.130)	(0.059)	(0.087)	(0.133)
Urban	−1.111 ***	−1.352 ***	−1.627 ***	−1.509 ***	−1.465 ***	−1.166 ***	−1.431 ***
	(0.245)	(0.088)	(0.118)	(0.099)	(0.057)	(0.083)	(0.113)
Rural	−1.161 ***	−1.484 ***	−1.829 ***	−2.034 ***	−1.501 ***	−1.195 ***	−1.603 ***
	(0.100)	(0.105)	(0.148)	(0.180)	(0.059)	(0.098)	(0.152)
Income elasticity
Totality	0.368 ***	0.260 ***	0.350 ***	0.364 ***	0.373 ***	0.370 ***	0.348 ***
	(0.025)	(0.019)	(0.025)	(0.024)	(0.020)	(0.029)	(0.030)
Urban	0.284 ***	0.200 ***	0.270 ***	0.281 ***	0.287 ***	0.285 ***	0.268 ***
	(0.027)	(0.021)	(0.026)	(0.027)	(0.026)	(0.030)	(0.030)
Rural	0.452 ***	0.319 ***	0.429 ***	0.446 ***	0.457 ***	0.453 ***	0.427 ***
	(0.030)	(0.024)	(0.031)	(0.030)	(0.026)	(0.037)	(0.037)

The elasticity and standard error in the table were obtained with a bootstrap condition of 1000 times, where *** indicates significance at a statistical level less than 1%, and the standard error is in parentheses.

. The Marshall own-price elasticity of all types of food is negative and significant at a 1% level. The absolute value of egg own-price elasticity is the largest, indicating that fluctuations in egg prices can have the greatest impact on the food consumption of residents, and the absolute value of the egg own-price elasticity of aquatic products is also large. Eggs are livestock products with high cost performance. Thus, an uncertain price will affect the daily consumption of eggs. With the improvement of residents’ living standards and the increase in awareness regarding the importance of eating a healthy diet, people have acquired a deeper understanding and appreciation of the nutritional value of eggs [37]. Grains and oils are daily necessities for residents, and their absolute value of own-price elasticity is small. From the perspective of urban and rural areas, the absolute values of the seven food categories for urban residents are smaller than those of rural residents, indicating that there is still a gap between the food consumption level of rural residents and urban residents. From the perspective of resident sensitivity to food prices, urban residents are more sensitive to the price of eggs, while rural residents are more sensitive to the price of aquatic products. Additionally, the consumption of basic food (grains and oils) by urban and rural residents is relatively stable. With the improvement of living standards and health awareness of residents in the Yangtze River Delta region, urban residents have gradually realized the importance of eggs consumption, so the price changes of eggs have a great impact on urban residents. Rural residents sometimes consume aquatic products, so the price changes of aquatic products have a great impact on the food consumption of rural residents. Basic food (grains and oils) is necessary for residents’ daily food consumption, so residents’ consumption of such food was found to be relatively stable.

From the perspective of the income elasticity of food consumption, the income elasticity of all types of food was found to be significant at a 1% level. In general, the income elasticity of all food is less than 0.5, corresponding to the normal commodity category. The income elasticity of vegetables is the highest, that of meat and poultry is the lowest, and that of other food is generally between 0.3 and 0.4. In addition, the income elasticity of urban residents is lower than that of rural residents, indicating that when the income of urban and rural residents increases, rural residents consume more food than urban residents, This result is also consistent with the study by Yin et al. [1].

4.4.2. Elasticity Analysis of Carbon Emissions from Food Consumption

The elasticity of food carbon emissions can be deduced from Formulas (8) and (9), as shown in

Table 5. Price elasticity and income elasticity of food consumption carbon emissions of urban and rural residents in the Yangtze River Delta.

	Price Elasticity							Income Elasticity
	Grains	Meat and Poultry	Eggs	Aquatic Products	Vegetables	Oils	Fruits
Totality	−0.241	−0.423	−0.038	−0.232	−0.076	0.003	0.029	0.326
Urban	−0.144	−0.465	−0.050	−0.260	−0.091	−0.008	0.010	0.247
Rural	−0.332	−0.385	−0.025	−0.218	−0.061	0.016	0.047	0.406

. If the price elasticity of food carbon emissions is negative and the absolute value is larger, then after the price increases by 1%, resident consumption of this kind of food will further reduce the carbon emissions of the whole food system; if the price elasticity is positive, the result will be the opposite. A greater income elasticity of food carbon emissions means that an increase in income will lead to an increase in food consumption carbon emissions. In general, the price elasticity of food carbon emissions is negative except for that of fruits and oils. Additionally, the price elasticity of meat and poultry, grains, and aquatic products is relatively large, indicating that reducing the consumption of these types of food can significantly reduce the carbon emissions within the food system. Notably, the price elasticity of carbon emissions is positive in the table, possibly because there are many substitutes for these two types of food. An increase in price will drive an increase in the consumption of other foods but will increase the per capita food carbon emissions. The price elasticity of food carbon emissions among rural residents is consistent with the whole, while the price elasticity of oils carbon emissions among urban residents is negative. From the perspective of income elasticity, the overall food carbon emissions income elasticity of residents is 0.326, and the food carbon emissions income elasticity of rural residents is higher than that of urban residents, indicating that the increase in food carbon emissions among rural residents is greater than that of urban residents after the increase in income.

4.4.3. Elasticity Analysis of Nutrition from Food Consumption

Food nutrient elasticity can be derived from Formulas (10) and (11) (

Table 6. Price elasticity and income elasticity of food consumption nutrients of urban and rural residents in the Yangtze River Delta.

	Price Elasticity							Income Elasticity
	Grains	Meat and Poultry	Eggs	Aquatic Products	Vegetables	Oils	Fruits
				Energy
Totality	−0.714	0.105	0.016	0.041	−0.155	−0.164	−0.201	0.358
Urban	−0.538	0.144	0.021	0.014	−0.180	−0.204	−0.275	0.273
Rural	−0.856	0.115	0.017	0.065	−0.132	−0.132	−0.139	0.443
Protein
Totality	−0.547	−0.090	−0.087	−0.113	−0.195	0.018	−0.022	0.345
Urban	−0.375	−0.109	−0.110	−0.161	−0.226	0.020	−0.049	0.261
Rural	−0.691	−0.050	−0.064	−0.076	−0.165	0.021	0.005	0.430
Fats
Totality	−0.252	−0.222	0.083	0.231	0.124	−0.521	−0.479	0.345
Urban	−0.165	−0.211	0.069	0.220	0.131	−0.526	−0.496	0.264
Rural	−0.365	−0.216	0.102	0.245	0.114	−0.513	−0.450	0.427
Carbohydrates
Totality	−0.942	0.278	0.011	−0.003	−0.217	−0.070	−0.150	0.365
Urban	−0.797	0.410	0.028	0.059	−0.284	−0.095	−0.245	0.280
Rural	−1.043	0.247	0.007	0.036	−0.172	−0.050	−0.083	0.450

). Overall, the energy price elasticity of meat, poultry, eggs, and aquatic products is positive, indicating that an increase in these food prices will increase the overall energy intake of residents. Except for oils, the protein price elasticity of other foods is negative. Except for eggs, aquatic products, and vegetables, the elasticity of the price of fats and other foods is negative. The price elasticity of carbohydrates for meat, poultry, and eggs is positive. Since the quantity of grains consumed is larger than that of other foods, the absolute value for the nutrient price elasticity of grains is larger than that of other foods, which also indicates that an increase in grains price will significantly reduce the intake of nutrients by residents, especially carbohydrate intake. There is a large difference in the price elasticity of grains nutrients between urban and rural residents. Additionally, the absolute value of elasticity in rural areas is larger than that in urban areas, indicating that an increase in grains prices will lead to a significant decrease in the intake of nutrients by rural residents, which may be mainly because grains consumption is an important source of maintaining the energy intake of rural residents.

The increase in the price of meat has a more significant impact on the fat intake of rural residents and the carbohydrate intake of urban residents. An increase in the price of oils and fruits will significantly reduce the fat intake of urban and rural residents, with a larger decrease for urban residents. While all income elasticity is similar, the income elasticity of carbohydrates is slightly greater, indicating that an increase in resident income would lead to greater carbohydrate intake than other nutrients. The elasticity of nutrient income in rural areas is greater than that in urban areas.

4.5. Simulation Analysis of Income and Policy Changes

Equations (12) and (13) can be used to simulate the impacts of income and food price changes on the per capita food consumption carbon emissions and nutrition intake of residents in the Yangtze River Delta, where food price changes are reflected by different tax and subsidy policies [38]. Based on the above analysis, and referring to the methodologies of some scholars [7,34], five plans were designed in this paper. The specific practices of these five plans are briefly explained below, and the policy simulation results are shown in

Table 7. Simulation results of income and policy changes.

	Plan 1	Plan 2	Plan 3	Plan 4	Plan 5
Carbon emissions change ratio/%
Totality	2.94 ↑	4.23 ↓	0.09 ↓	4.31 ↓	1.37 ↓
Urban	2.12 ↑	4.65 ↓	0.40 ↓	5.05 ↓	2.93 ↓
Rural	4.02 ↑	3.85 ↓	0.22 ↑	3.63 ↓	0.39 ↑
Energy change ratio/%
Totality	3.22 ↑	1.05 ↑	1.85 ↓	0.80 ↓	2.42 ↑
Urban	2.35 ↑	1.44 ↑	2.54 ↓	1.10 ↓	1.25 ↑
Rural	4.39 ↑	1.15 ↑	1.22 ↓	0.07 ↓	4.32 ↑
Protein change ratio/%
Totality	3.11 ↑	0.90 ↓	1.09 ↓	1.99 ↓	1.12 ↑
Urban	2.25 ↑	1.09 ↓	1.59 ↓	2.68 ↓	0.43 ↓
Rural	4.26 ↑	0.50 ↓	0.59 ↓	1.09 ↓	3.17 ↑
Fats change ratio/%
Totality	3.11 ↑	2.22 ↓	3.96 ↓	6.18 ↓	3.07 ↓
Urban	2.27 ↑	2.11 ↓	4.27 ↓	6.38 ↓	4.11 ↓
Rural	4.23 ↑	2.16 ↓	3.48 ↓	5.64 ↓	1.41 ↓
Carbohydrates change ratio/%
Totality	3.29 ↑	2.78 ↑	1.39 ↓	1.39 ↑	4.68 ↑
Urban	2.41 ↑	4.10 ↑	2.17 ↓	1.93 ↑	4.34 ↑
Rural	4.46 ↑	2.47 ↑	0.76 ↓	1.71 ↑	6.17 ↑

This “↑” represents the increase and “↓” represents the decrease.

.

Plan 1: Referring to the growth rate of the per capita disposable income of residents in the Yangtze River Delta in 2019, it is assumed that the overall per capita disposable income of residents in the Yangtze River Delta increases by 9%, urban per capita disposable income increases by 8.6%, rural per capita disposable income increases by 9.9%, and all food prices remain unchanged.

Plan 2: To discourage the consumption of meat and poultry (high-carbon foods), meat and poultry should be taxed. For example, the prices of meat and poultry could be raised by 10% while keeping income unchanged.

Plan 3: To promote the consumption of eggs and fruits (low-carbon foods), the consumption of these types of food could be subsidized; specifically, the prices of these types of food could be decreased by 10% while keeping income unchanged.

Plan 4: Combine Plans 2 and 3 to impose taxes on high-carbon foods and subsidize low-carbon foods with the same income.

Plan 5: Combine Plans 1 and 4 and consider the impact of income and price changes on the carbon emissions and nutritional intake of residents.

From the perspective of carbon emissions changes, the results of Plan 1 show that the overall per capita food consumption carbon emissions of residents in the Yangtze River Delta increased by 2.94% after an increase in resident income, and the per capita food consumption carbon emissions of rural residents increased nearly twice as much as those of urban residents. Tiboldo et al. noted that a carbon tax on food purchases could reduce greenhouse gas emissions from the agriculture and food sectors by 1.9% to 4.8% [39]. Plan 2 shows that by taxing high-carbon foods, the per capita food carbon emissions of residents will be reduced by 4.23%. It can be seen that the price regulation of meat and poultry has a very significant effect and is most effective in reducing the carbon emissions of food consumption among urban residents. Overall, Plan 3 can reduce the food carbon emissions of residents by 0.09%, but urban and rural residents show the opposite effects under the simulation of Plan 3, in which the reduction in urban resident carbon emissions is greater than the increase in rural resident carbon emissions. The results of Plan 4 show that under the combined effects of taxes and subsidies, residents in the Yangtze River Delta can reduce their food carbon emissions by 4.31%, and a tax on high-carbon food has a more significant effect on reducing carbon emissions. A combination of taxes and subsidies is more effective than taxes and subsidies alone [40]. Plan 5 shows that income and prices change at the same time. Under this plan, the Yangtze River Delta carbon emissions per capita for food is reduced by 1.37%, where urban residents reduce carbon emissions by 2.93% for food. Rural residents, based on a 0.39% increase in carbon emissions and their own perspectives, indicated that food price controls could counteract the effects of income increases. Resident food carbon emissions will grow slowly and even have a negative growth effect on food carbon emissions.

From the perspective of nutrient intake changes, under the price control of taxes and subsidy coexistence, the intake of other nutrients except for carbohydrates showed a downwards trend [39]. Here, the change ratio of fats was the most significant, providing values as high as 6.18%. In addition, under the combined effects of income and price, resident intake of energy and carbohydrates will significantly increase, while the intake of fats will significantly decrease. Under the current trend of the “high oils and fats” diet of residents, the implementation of carbon reduction measures could reduce the intake of fats within this group. Compared to the recommended intake of nutrients per capita shown in Table 3, the intake of energy, fats, and carbohydrates was within the recommended range before and after the policy, except that the intake of protein was higher than the recommended value both before and after the policy. These results indicate that the implementation of carbon reduction measures has not only maintained the current normal nutritional intake of residents in the Yangtze River Delta but also corresponds to sustainable environmental development criteria [41,42].

5. Conclusions

Based on the food consumption data of residents in the Yangtze River Delta from 1995 to 2019, combined with the two-stage Engel–QUAIDS model, this paper analysed the food consumption and carbon emissions of urban and rural residents in the Yangtze River Delta, simulated the impact of income growth and food price regulation on the per capita food consumption carbon emissions and nutrition intake of residents, and drew the following conclusions.

First, due to the fluctuation of grains consumption, the total food consumption of residents in the Yangtze River Delta presented a U-shaped pattern. Second, with an increase in rural residents’ income, the growth rate of their carbon emissions from food consumption will be greater than that of urban residents. Third, the nutritional intake of residents can still be guaranteed under a low-carbon policy. Notably, before and after the simulation, residents’ fat intake was significantly reduced within the recommended range, which would also be beneficial to residents’ health. Fourth, food price regulations will effectively restrain the growing effects of income increases on carbon emissions from food consumption. Therefore, the government needs to make reasonable use of price regulation policies to reduce carbon emissions from food consumption.

To better realize the “Healthy China 2030” strategy and support the “2060” dual carbon target while maintaining the basic nutrient intake of residents, we should focus on exploring the best measures to reduce the carbon consumption of food. Based on the above research results, in the future, we can focus on the low-carbon food consumption patterns of other targeted areas (especially poor areas) or targeted populations (e.g., the elderly, women, and children), precisely implement dietary nutrition interventions, and improve nutritional health levels.