Analysis of agricultural economic development and optimisation measures under the strategy of rural revitalisation

The cointegration test commonly used Engle-Granger cointegration test or Johansen cointegration test. The EG cointegration test is generally suitable for cointegration between two variables [26], and the Johansen cointegration test is generally used for cointegration between multiple variables. This paper is a test between multiple groups of two variables; so, the EG cointegration test is used [27]. This paper establishes a cointegration regression model to test the long-term equilibrium relationship between agricultural industrial structure optimisation and rural revitalisation [28]. This long-term equilibrium relationship is maintained by continuously adjusting short-term fluctuations; so, an error correction model (ECM) needs to be introduced [29] to be additional to that. (1) Lnyt=b0+b1lnxt+εt {\rm{Ln}}{{\rm{y}}_{\rm{t}}} = {{\rm{b}}_0} + {{\rm{b}}_1}\ln {{\rm{x}}_{\rm{t}}} + {\varepsilon _{\rm{t}}}

Among them is the explanatory variable lnx_t, where x_t represents the selected indicators related to the measurement of agricultural industrial structure optimisation, and Lny_t is the explained variable, where y_t represents the selected indicators related to rural revitalisation, and ɛ_t is the amount of random error.

The ECM is as follows: (2) ΔLnyt=β0+β1Δlnxt−λecm+εt \Delta {\rm{Ln}}{{\rm{y}}_{\rm{t}}} = {\beta _0} + {\beta _1}\Delta \ln {{\rm{x}}_{\rm{t}}} - {\lambda _{{\rm{ecm}}}} + {\varepsilon _{\rm{t}}} where ecm represents the error correction term, β₀ is a constant term and, β₁, λ is a parameter term, which ɛ₁ is the random error amount.

The data are mainly from the 2001–2017 ‘A Statistical Yearbook’ and ‘A Government Work Report,’ and the official data are released by the city's bureau of statistics. This article selects reviews with 7.0 as the analysis tool for EG cointegration analysis, regression analysis and Granger causality test [30]. For the convenience of inspection, the data of each indicator are expressed as follows: the annual growth rate of agricultural output value (ZC), the comprehensive grain production capacity (LS), the proportion of rural employees engaged in non-agricultural industries (JS) and the agricultural labour productivity (LD), the growth rate of agricultural added value (XJ), the income ratio of urban and rural residents (CJ), the per capita net operating income of rural households (NJ), the Engel coefficient (NG) of rural residents and the proportion of rural residents’ expenditure on education, culture and entertainment (JY). According to the corresponding indicators in Figure 2, this paper will analyse the annual growth rate of agricultural output value and comprehensive grain production capacity (denoted as L4-1), the proportion of non-agricultural labour force and agricultural labour productivity (L4) among rural employees, the ratio of the growth rate of agricultural added value to the income of urban and rural residents (L4-3), the per capita net operating income of rural households and the Engel's coefficient of rural residents (L4-4) and the ratio of per capita net operating income of rural households to rural residents. The proportion of education, culture and entertainment expenditure (L4-5) is analysed in five groups of time series data. In order to ensure the accuracy and scientificity of the research and prevent the variables and numerical values from changing drastically, the logarithm of each group of data is used as a variable for research. LNZC, LNLS, LNJS, LNLD, LNXJ, LNCJ, LNNJ, LNNG, LNJY (logarithmic processing will not affect the nature of the data) are recorded, and the ADF test is performed. It can be seen that the labour force engaged in non-agricultural industries among the rural employees has gradually declined, and the number was the smallest in 2016, and the change in the labour generation rate after 2003 was small. The per capita net operating income of rural households and the Engel's coefficient of rural residents gradually decreased and reached their peaks in 2010. From Figure 4(a), it can be seen that the per capita operating net income of rural households has gradually increased, and the proportion of labour force has decreased. It shows that the living standard of rural residents has been improved.

Fig. 2

Fig. 3

Fig. 4

The principle of the ADF test method is that the null hypothesis is that the sequence has a unit root, that is, non-stationary. When testing time series data, under the conditions of a given confidence level (1%, 5%, 10%), when the obtained statistic is significantly less than the critical statistical value of confidence, it means that the null hypothesis is rejected, that is, it is stationary; otherwise, it is non-stationary. The ADF test model is as follows: (3) Model (1) Δyt=δyt−1+∑j=1pλjΔyt−j+μt {\rm{Model}}{\kern 1pt} {(1)}\quad \Delta {y_t} = \delta {y_{t - 1}} + \sum\limits_{j = 1}^p {\lambda _j}\Delta {y_{t - j}} + {\mu _t} (4) Model (2) Δyt=α+δyt−1+∑j=1pλjΔyt−j+μt {\rm{Model}}{\kern 1pt} {(2)}\quad \Delta {y_t} = \alpha + \delta {y_{t - 1}} + \sum\limits_{j = 1}^p {\lambda _j}\Delta {y_{t - j}} + {\mu _t} (5) Model (3) Δyt=α+βt+δyt−1+∑j=1pλjΔyt−j+μt {\rm{Model}}{\kern 1pt} {(3)}\quad \Delta {y_t} = \alpha + \beta t + \delta {y_{t - 1}} + \sum\limits_{j = 1}^p {\lambda _j}\Delta {y_{t - j}} + {\mu _t} t is a time variable, indicating a certain trend of the series over time. First start the test from model 3, then model 2 and finally model 1. The test can be stopped until the test result rejects the null hypothesis. Otherwise, the test will continue until model 1. The time series is considered stationary if any of the model detection results reject the null hypothesis.

The results of the data test for each group of variables:

Fig. 5

Taking LNZC as an example, when the sequence LNZC has a constant term, a time trend term and a lag order, the ADF value is greater than the critical value at the 5% significance level, and the P-value is >0.05; the null hypothesis cannot be rejected, that is, the sequence is not stationary; while the ADF value of the first-order difference series DLNZC of LNZC is less than the lower critical value of the 5% significance level, the P-value is <0.05. Reject the null hypothesis that the series is stationary. From Figure 5, it can be concluded that the other groups of variables are all first-order single integrations; it meets the requirements of the cointegration test.

The cointegration test results show the annual growth rate of the total agricultural output value and the comprehensive grain production capacity of a city over the years, the proportion of rural employees engaged in non-agricultural industries and agricultural labour productivity, the growth rate of agricultural added value and the ratio of urban and rural residents’ income. There is a long-term equilibrium relationship between the per capita operating net income and the Engel coefficient of rural residents and the proportion of rural residents’ expenditure on education, culture and entertainment, which means that there is an ECM. The result is as follows: (7) L4−1: DLy=0.0079DLx−0.1531DLy(−1)−0.0026ecm(−1)L4−2: DLy=−0.1329DLx−0.0228Dly(−1)−0.0892ecm(−1)L4−3: DLy=0.0621DLx+0.0316DLy(−1)−0.0763ecm(−1)L4−4: DLy=0.3876DDx−0.0719DLy(−1)−0.5873ecm(−1)L4−5: DLy=−1.4612DLx+0.1084DLy(−1)−0.8992ecm(−1) \matrix{{L4 - 1:\quad DLy = 0.0079DLx - 0.1531DLy(- 1) - 0.0026ecm(- 1)} \hfill \cr {L4 - 2:\quad DLy = - 0.1329DLx - 0.0228Dly(- 1) - 0.0892ecm(- 1)} \hfill \cr {L4 - 3:\quad DLy = 0.0621DLx + 0.0316DLy(- 1) - 0.0763ecm(- 1)} \hfill \cr {L4 - 4:\quad DLy = 0.3876DDx - 0.0719DLy(- 1) - 0.5873ecm(- 1)} \hfill \cr {L4 - 5:\quad DLy = - 1.4612DLx + 0.1084DLy(- 1) - 0.8992ecm(- 1)} \hfill \cr} (where ECM(−1) is the one lag of the residual term in the association.)

The signs of the correction coefficients of ECM(−1) in each group are all negative, which conforms to the reverse correction mechanism. From the above test results, it can be seen that the residual ɛt of each group is a stationary time series, and there is a cointegration relationship, which can reflect the existence of a long-term equilibrium relationship between the variables in each group. The constructed model is reasonable and has economic significance.

Due to the inconsistency of the measurement units of various indicators, it is impossible to compare them accurately and objectively. Therefore, before calculating specific indicators, it is necessary to convert the absolute value of the indicator into a relative value and let (13) xij=|xij| {x_{ij}} = \left| {{x_{ij}}} \right| in order to solve the problem that the indicators of different units cannot be compared. Since the values of positive and negative indicators have different meanings, it is necessary to perform dimensionless data processing on them respectively. The relevant algorithm formula is as follows:

Positive indicators: (14) Xij′=Xij−min(X1j,X2j……Xnj)max(X1j,X2j……Xnj)−min(X1,,X2j…Xnj) X_{ij}^{'} = {{{X_{ij}} - \min \left({{X_{1j}},{X_{2j}} \ldots \ldots {X_{nj}}} \right)} \over {\max \left({{X_{1j}},{X_{2j}} \ldots \ldots {X_{nj}}} \right) - \min \left({{X_{1,}},{X_{2j}} \ldots {X_{nj}}} \right)}}

Negative indicators: (15) Xij′=max(X1,,X2j⋯…Xnj)−Xijmax(X1j,X2j……Xnj)−min(X1,,X2j⋯…Xnj) X_{ij}^{'} = {{\max \left({{X_{1,}},{X_{2j}} \cdots \ldots {X_{nj}}} \right) - {X_{ij}}} \over {\max \left({{X_{1j}},{X_{2j}} \ldots \ldots {X_{{\rm{n}}j}}} \right) - \min \left({{X_{1,}},{X_{2j}} \cdots \ldots {X_{{\rm{nj}}}}} \right)}} where x_ij is the value j of the first indicator of the i year. (i = 1,2…, n; j = 1,2,…, m)

(16) pij=Xij∑i=1nXij, (i=1,2…,n; j=1,2,…,m) {p_{ij}} = {{{X_{ij}}} \over {\sum\limits_{i = 1}^n {X_{ij}}}},\quad ({\rm{i}} = 1,2 \ldots,{\rm{n}};\;{\rm{j}} = 1,2, \ldots,{\rm{m)}}

(17) ej=−k∑i=1npijln(pij)lnk>0, k=1/ln(n),ej≥0 \matrix{\hfill {{e_j} = - k\sum\limits_{i = 1}^n {p_{ij}}\ln \left({{p_{ij}}} \right)} \cr \hfill {\ln k > 0,\;k = 1/\ln (n),{e_j} \ge 0} \cr}

For the first index j, X_ij is the smaller the difference and e_j is the greater the difference. When X_ij all are equal (e_j = e_max = 1(k = 1/lnn)), the index has no effect on the comparison X_ij of the evaluated objects. When X_ij is the larger difference and e_j is the smaller distance, the index has a greater effect on the comparison of the evaluated objects. Therefore, g_j = 1 − e_j, g_j the greater the definition difference coefficient, the more attention should be paid to the role of this indicator.

(18) wj=gj∑j=1mgj (1≤j≤m) {w_j} = {{{g_j}} \over {\sum\limits_{j = 1}^m {g_j}}}\quad (1 \le j \le m)

Among them, w_j is the normalised weight coefficient

(19) si=∑j=1mwj⋅pij (i=1,2,…n) {s_i} = \sum\limits_{j = 1}^m {w_j} \cdot {p_{ij}}\quad (i = 1,2, \ldots n)

The development of agricultural economy needs a strong theory as a support, which requires the establishment of a corresponding management mechanism. In the new era, driven by the rural revitalisation strategy, a systematic and institutionalised economic situation has gradually emerged. Relevant departments have also stepped up their efforts to continuously adjust and optimise the management mechanism based on the actual situation, promote industrial upgrading and then ensure the management system. It can match the reality of rural economic development and meet the diversified needs of agricultural economic development. In the process of optimising the management mechanism, the transformation of farmers’ production concepts is also an important part of agricultural economic development. Relevant departments should continuously disseminate advanced production concepts among farmers through training courses and on-the-spot explanations by experts to assist in optimising the management mechanism. In addition, it is necessary to continuously update agricultural production equipment. Farmers themselves and relevant departments must do their homework in this regard. Advanced equipment can improve the efficiency of agricultural production and create more production capacity within a certain period of time. The agricultural management department should exchange feedback with farmers to form a good connection so that the excellent management mechanism can truly take root and germinate in the fields, so as to better serve the development of the agricultural economy.

Science and technology are the boosters of all productive forces, and the continuous improvement of the level of science and technology is an effective means for the continuous development of my country's agricultural economy. The agricultural management department must learn advanced technical means before farmers, strengthen the work of agricultural economics and do a good job in the overall management of the standardised development of agriculture through the use of high-tech means. On this basis, agricultural management departments should also make scientific predictions on agricultural development, and based on this, they must formulate relevant policies to adapt to the general trend of agricultural economic development. It is necessary to give full play to the advantages of scientific and technological means, make scientific judgements through reasonable calculation, comparison and data analysis, change the traditional way of relying on experience and old farmers, establish and improve a networked supervision system and better solve problems in agricultural development, which is an outstanding issue.

The implementation of agricultural economic development work needs to rely on talents; talents are the carrier of management technology, and the quality of talents is directly related to the quality of agricultural economic development work. The management department should also regularly organise management personnel to carry out systematic work training to ensure the work quality and professional ethics of all personnel and always maintain the advanced level of management concepts and methods. The actual and future development directions of agricultural development in different regions are different. Managers in each region should have an in-depth understanding of the local agricultural situation so as to ensure the effectiveness of their own management. In addition, it is also necessary to combine the local agricultural development with national policies, formulate clear development plans based on the actual situation, give full play to advantages and avoid disadvantages, allow farmers to participate more actively in agricultural management and promote the better development of regional agricultural economy. The construction of rural infrastructure and informatization network construction is inseparable from the investment of funds. Only when the funds are in place, can a unique information network in the area be formed and provide farmers with more targeted agricultural information. Therefore, whether the capital investment in the preparatory stage is sufficient is an important factor for the smooth development of rural agricultural economic development.