Research on Industrial Economic Innovation Development Strategy Based on Ordered Logit Modeling

With the continuous improvement of China’s socialist market economic system, China has also been able to gradually improve the development of industrial economy, which occupies an indispensable position in the process of China’s national economic development and is an important part of China’s socialist market economy. Therefore, the industrial structure and industrial economic system still have an important impact on promoting the perfection of China’s socialist market economy, improving the division of labor and collaboration within the society, and promoting the rational allocation of resources [1-2].

The types of industrial economic development include low-carbon economy industry, high-tech industry, e-commerce industry and so on. Low-carbon economy is of positive significance for the promotion of sustainable economic stability, social harmony and healthy construction, enterprises continue to adjust and transform the traditional high-energy-consuming industries, which greatly promotes the optimization of industrial structure and the soundness of the industrial chain [3-5]. The development of high-tech industry with the continuous improvement of the industrial economic chain, many enterprises in the use of technological innovation on the basis of the premise of realizing the development of their own information technology [6-8]. E-commerce industry development with the continuous development and soundness of high-tech industry, the traditional real economy has also suffered a certain impact, in the support of the sales platform, e-commerce and e-logistics industry gradually developed, and gradually formed a more sound industrial chain, for the industrial economic innovation has played an important role in promoting [9-10]. Although China has already got rid of the blindly expanding industrial development mode under the planned economy system in the process of economic development, and takes the promotion of high-tech industry and the optimization of industrial structure as the main development direction. Therefore, under the guidance of the new industrial development model, China’s financial system is constantly sound, the level of industrial development is gradually improved, has gradually abandoned the planned economic system to the high energy consumption, high pollution industry development disadvantages, industrial economic development presents a good situation.

Logit model, which can also be called rating model and categorical rating model, is an ordered polynomial regression model designed for solving classification problems and is widely used [11-14]. It is a deformed model of the traditional logit model, which uses binary transformations for categorical variables, while the ordered logit model introduces acceptance metrics, thus better modeling the natural structure in the data [15-16]. In addition the ordered logit model can also be used to analyze the causes of traffic accidents and the severity of the accidents in order to effectively improve the body functions and ensure driving safety [17-18]. Overall, the ordered logit model has many advantages, which can realize more effective recommender system and targeted delivery to provide better service for industrial economy.

In order to explore the influencing factors of industrial economic innovation development and propose corresponding industrial development strategies, this paper analyzes the factors affecting industrial economic innovation development based on the ordered Logit model. Ten specific variables are subdivided according to the level of economic development, urbanization level, and other first-level indicators, and the hypotheses concerning the role of these variables in industrial economic development are established accordingly. The different variables are divided into five levels and assigned values, and then the variables are logarithmized and tested for covariance. The variables were included in the ordered Logit model, and the independent variables were screened by the backward method to get the statistically significant independent variables on the industrial economic innovation and development, and the goodness of fit of the ordered Logit model was compared with that of the multinomial Logit model. Based on the ordered Logit regression results, causal inference of industrial innovation development is carried out, and several strategies to enhance industrial innovation development are given.

2

A preliminary study of the factors influencing the development of innovation in the industrial economy

The degree of innovation and development of industrial economy marks the level of economic development of a country, and also becomes an important symbol to measure the degree of development of modern economy, so accurately grasping the factors affecting the development of industrial economy is of great significance to enhance the level of national economic development. The factors affecting the development of the industrial economy are diverse. This paper initially categorizes the factors that affect the innovative development of the industrial economy into the following points.

2.1

Level of economic development

A country’s economic development level directly determines the people’s living standard and consumption level, driving the development of industry, the country’s economic development plays an important role in the development of industrial economy, this paper takes GDP and per capita GDP (PCGDP) as the indicators of the country’s economic development level.

2.2

Level of urbanization

China’s urbanization rate has been accelerating and has become an important factor in the development of the industrial economy. Gradually improved medical services, education, financial services and the Internet (service) industry, which has been increasingly rising in the urbanization process in recent years, play an important role in the rapid development of the industrial economy. This paper uses the proportion of the urban population (UR) and the utilization rate of urban land (TLUR) as a measure of China’s urbanization level.

2.3

Standard of living of the population

The living standard of the residents is an important element to consider for the development and growth of the national industrial economy, the living standard of the residents has been improved, so that there will be a transformation of the residents’ consumption from basic clothing, food, housing and transportation to a variety of enjoyment, enhancement and convenience as the main purpose of the service-oriented consumption, which to a large extent contributes to the development and growth of the industrial economy, this paper uses the per capita disposable income of the residents (PCDI) and the cost of living index (CLI) as the indicators to measure the living standard of Chinese residents.

2.4

Labor

Labor force is the core of industrial economic development, and the proportion of employed people in the industrial economy and the quality of the labor force are two important factors of labor force factors. The number of employed persons in industrial economy can reflect the development of industrial economy to a certain extent, the quality of labor force largely determines the development space and process of industrial economy, and high-quality labor force is an important factor for the high-quality development of industrial economy. In this paper, we would like to explore the influence of two variables of labor force quantity and labor force quality on the development of industrial economy at the same time, and adopt the proportion of the number of employees in the industrial economy to the total number of employees (PIE) and the proportion of the number of college graduates (PCG) as the measurement indexes for the quantity of labor force in the industrial economy and the overall quality of labor force, respectively.

2.5

Role of Government

Under China’s current socialist market economic system, government policies have a significant impact on the development of industries or sectors. The government’s role in the development of industrial economy is mainly reflected in the government’s fiscal expenditure, and this paper mainly takes the government’s total fiscal expenditure (TGE) and government efficiency (GE) as the main indicators of the government’s role.

Meanwhile, the share of industrial economy in GDP (IESGDP) is used as an indicator of industrial economic development.

3

Theory of Ordered Logit Models

3.1

Overview of Logit Models

To date, the Logit model is the earliest and most widely used non-setting model. The theoretical basis of the non-aggregate model is the hypothesis that consumers seek to maximize “utility” when making choices. In economics, utility is defined in various ways. Simply put, utility is the pleasure consumers derive from their consumption choices, or the satisfaction of their needs [19].

3.2

Assumptions for Logit model estimation

The following five basic assumptions must be met when estimating the parameters of a Logit model:

First, the data must come from a random sample.

Second, the dependent variable y_i is a function of k the independent variable x_ki(k = 1, 2, …, k). And dependent variable y_i is a dichotomous variable; this variable can only take 0 or 1.

Third, the relationship between the dependent variable and each of the independent variables in the Logit model is nonlinear.

Fourth, the Logit model has no assumptions about the distribution of the independent variables; the respective variables can be continuous, discrete, or dummy variables. And there is no need to assume the existence of multivariate normal distribution among them. However, the existence of multivariate normal distribution relationships between independent variables will increase the efficacy of the model and the stability of the solution.

Fifth, Logit model is sensitive to multivariate covariance. The existence of multivariate covariance between independent variables can lead to the inflation of standard errors.

3.3

Logit transformations

First, the ratio of the probability of a certain outcome occurring to the probability of it not occurring, or odds, is calculated: (1) $o d d s = \frac{P}{1 - P}$ $$odds = \frac{P}{{1 - P}}$$

Second, odds is converted to lnodds: (2) $o d d s = \ln \frac{P}{1 - P}$ $$odds = \ln \frac{P}{{1 - P}}$$

where ln odds is called LogitP:

The above two steps are called Logit transformations, and after the Logit transformations, the general linear regression model can be used to build a multivariate analysis model Logit model between the explanatory variables and the explanatory variables [20]: (3) $L o g i t P = β_{0} + \sum_{i = 1}^{k} β_{i} x_{i}$ $$LogitP = {\beta _0} + \sum\limits_{i = 1}^k {{\beta _i}} {x_i}$$

4

Ordered logit regression analysis of the innovative development of the industrial economy

4.1

Research hypotheses and variable selection

4.1.1

Research hypotheses

Drawing on existing research results, this paper outlines the influencing factors of industrial economic innovation and development as: economic development level variables, urbanization level variables, residents’ living standard variables, labor force variables and government role variables. 1)

Economic development level variables (GDP, GDP per capita). According to the survey, countries with a high level of economic development will have a higher willingness of enterprises to engage in economic innovation than countries with a low level of economic development. The more stagnant the economic development, the more the enterprises tend to be satisfied with the status quo, unwilling to update the concept and production technology, the enterprises will be less willing to implement the innovation strategy.

2)

Urbanization level variables (urban population share, urban land use rate). Societies with a high proportion of urban population are more likely to increase the level of innovation in the industrial economy, because the higher degree of population concentration in towns and cities, the relatively advanced production relations and production organization, compared with the rural areas are more likely to implement innovative strategies. The urban land utilization rate, on the other hand, indirectly reflects the degree of utilization of factors of production and the pressure on industrial development. Higher urban land utilization may reflect a higher degree of industrial intensification, thus providing the basic conditions to stimulate innovative development.

3)

Variables of residents’ living standard (per capita disposable income and cost of living index). Residents’ living standards may also affect the innovative development strategy of industrial economy. The higher the living standards of residents, the higher the requirements for food quality, quality of living goods and living service level will be correspondingly increased, thus forcing the primary, secondary and tertiary industries to promote industrial innovation.

4)

Labor force variables (the number of employees in the industrial economy, the proportion of college graduates). Industrial innovation ultimately depends on innovative talents. Adequate number of labor force can provide human resources for the promotion of industrial innovation, while high-quality labor force is the foundation and guarantee for the realization of industrial innovation and transformation and upgrading.

5)

Government role variables (total government financial expenditure, government efficiency). The stronger the government’s financial support is, the more sufficient the innovation power of each industry is, and the stronger the willingness to push forward the development of new-quality productivity and the implementation of high-quality development. The greater the technical support and financial support provided by the government, the lower the capital cost and technical pressure, and the higher the willingness of enterprises to implement innovative behavior.

4.1.2

Multivariate Ordered Logit Models

In this paper, a multivariate ordered Logit model is used to study the influencing factors of industrial economic innovation and development [21]. In it, the different variables are divided into five levels from low to high, and are assigned to five different degrees such as 1, 2, 3, 4, 5, etc. with sequential order.

Its basic model is: (4) $y_{i}^{*} = x_{i}^{'} β + u_{i}^{*}$ $$y_i^* = x_i^{\prime} \beta + u_i^*$$

In equation (4), $y_{i}^{*}$ $$y_i^*$$ is the latent variable, x_i is the influential factors that may affect the development of industrial economic innovation, which is the value of the independent variable, β is the regression coefficient of the independent variable, y_i is the dependent variable, and in this paper, the value of y_i is in the range of {1, 2, 3, 4, 5}, and y = 1 stands for the sample’s level is very low, y = 2 stands for the low level, y = 3 stands for the general, y = 4 stands for the high level, and y = 5 stands for the very high level. In general, the relationship between y_i and $y_{i}^{*}$ $$y_i^*$$ is shown below: (5) $y_{i} = {\begin{array}{l} 0 & y_{i}^{*} \leq c_{1} \\ 1 & c_{1} < y_{i}^{*} \leq c_{2} \\ 2 & c_{2} < y_{i}^{*} \leq c_{3} \\ ⋮ & ⋮ \\ M & c_{M} < y_{i}^{*} \end{array}$ $${y_i} = \left\{ {\begin{array}{*{20}{l}} 0&{\quad y_i^* \leq {c_1}} \\ 1&{\quad {c_1} < y_i^* \leq {c_2}} \\ 2&{\quad {c_2} < y_i^* \leq {c_3}} \\ \vdots &{\quad \vdots } \\ M&{\quad {c_M} < y_i^*} \end{array}} \right.$$

c_i is the critical value, and assuming that the distribution function for $u_{i}^{*}$ $$u_i^*$$ is F(x), the following probability distribution can be obtained: (6) ${\begin{array}{l} p (y_{i} = 0) = F (c_{1} - x_{i}^{'} β) \\ p (y_{i} = 1) = F (c_{2} - x_{i}^{'} β) - F (c_{1} - x_{i}^{'} β) \\ p (y_{i} = 2) = F (c_{3} - x_{i}^{'} β) - F (c_{2} - x_{i}^{'} β) \\ ⋮ \\ p (y_{i} = M) = 1 - F (c_{M} - x_{i}^{'} β) \end{array}$ $$\left\{ {\begin{array}{*{20}{l}} {p({y_i} = 0) = F({c_1} - x_i^\prime \beta )} \\ {p({y_i} = 1) = F({c_2} - x_i^\prime \beta ) - F({c_1} - x_i^\prime \beta )} \\ {p({y_i} = 2) = F({c_3} - x_i^\prime \beta ) - F({c_2} - x_i^\prime \beta )} \\ \vdots \\ {p({y_i} = M) = 1 - F({c_M} - x_i^\prime \beta )} \end{array}} \right.$$

Eq. (6) where p(y_i)(i = 1, 2, ⋯, M) is the degree probability of the level of the variable and F(·) is the logistic distribution function, integrating the above formulas leads to the following equation: (7) $\log i t {\begin{matrix} p (y_{i} \leq k) \end{matrix}} = c_{k} - x_{i}^{'} β$ $$\log it\left\{ {\begin{array}{*{20}{c}} {p({y_i} \leq k)} \end{array}} \right\} = {c_k} - x_i^{\prime} \beta$$

The above model was estimated using the econometric analysis software Stata to find out the significant influencing factors affecting the level of innovation development in the industrial economy.

4.1.3

Variable selection

The influencing factors of industrial economic development can be summarized into five categories and 10 variables, as shown in Table 1.

Table 1.

Model variables and related features

Primary variable	Secondary variable	Variable meaning
Economic development level (Dollar)	GDP	<500billion=1, 500~999billion=2, 1000~2999 billion=3, 3000~9999billion=4, ≥10000billion=5
Economic development level (Dollar)	Per capita GDP(PCGDP)	<1000=1, 1000~2999=2, 3000~9999=3, 10000~29999=4, ≥30000=5
Level of urbanization	Urban population(UR)	<30%=1, 30%~49%=2, 50%~69%=3, 70%~84%=4, ≥85%=5
Level of urbanization	Town land use ratio(TLUR)	<20%=1, 20%~39%=2, 40%~59%=3, 60%~79%=4, ≥80%=5
Living standard (Dollar)	Per capita disposable income(PCDI)	<3000=1, 3000~5999=2, 6000~14999=3, 15000~29999=4, ≥30000=5
Living standard (Dollar)	Cost of living index(CLI)	<50=1, 50~74=2, 75~99=3, 100~149=4, ≥150=5
Labor force	Personnel in the industry economy(PIE)	<20%=1, 20%~39%=2, 40%~59%=3, 60%~79%=4, ≥80%=5
Labor force	Proportion of college graduates(PCG)	<1%=1, 1%~2.9%=2, 3%~5.9%=3, 6%~9.9%=4, ≥10%=5
Government action	Total government expenditure(TGE)	<100 billion=1, 100 ~499 billion =2, 500~1999billion =3, 2000 ~9999 billion =4, ≥10000billiono=5
Government action	Government efficiency(GE)	Extremely slow=1, slow=2, average=3, high=4, extremely high=5

4.2

Descriptive statistics and covariance tests

All variables were logarithmized to take into account the potential adverse effects of heteroscedasticity on the measurement regression. The descriptive statistics of the final variables are shown in Table 2. Among all variables, the percentage of college graduates and GDP have higher values of 4.06 and 3.64, respectively.

Table 2.

Descriptive statistics of basic variables

Variable	N	Mean	SD	Min	Max
GDP	20	3.64	1.09	1	5
PCGDP	20	2.06	0.39	1	5
UR	20	2.24	0.94	1	5
TLUR	20	2.72	0.89	1	5
PCDI	20	2.78	1.09	1	5
CLI	20	2.28	0.07	1	5
PIE	20	3.33	0.58	1	5
PCG	20	4.06	1.02	1	5
TGE	20	1.13	1.14	1	5
GE	20	1.37	0.57	1	5

Table 3 is a description of the correlation of the variables, where the bolded data are the results of the variables that show significance at the 5% level or less. From the composition of the correlation coefficients between the various potential influencing factors and the various aspects of industrial economic performance, it can be found that the impact of the number of employees in the industrial economy shows a negative correlation, although in terms of significance, its impact on the economic performance of the industry itself is not significant. Other aspects of the role of factors, most of which have passed the significance test and are positive. The largest correlation coefficients between these factors do not exceed the level of 0.8, and in general, it can be considered that the impact of the covariance problem may not be very prominent.

Table 3.

The correlation description of each variable

	GDP	PCGDP	UR	TLUR	PCDI	CLI	PIE	PCG	TGE	GE
GDP	1.00
PCGDP	0.18	1.00
UR	0.74	0.15	1.00
TLUR	0.63	0.13	0.58	1.00
PCDI	0.28	-0.38	0.08	0.24	1.00
CLI	0.56	0.37	0.58	0.43	-0.18	1.00
PIE	-0.66	0.18	0.79	0.56	-0.14	0.72	1.00
PCG	0.76	0.35	0.77	0.67	-0.21	0.52	0.55	1.00
TGE	0.71	0.01	0.57	0.49	0.03	0.47	0.46	0.71	1.00
GE	0.58	-0.04	0.62	0.48	0.33	0.36	0.36	-0.06	-0.14	1.00

4.3

Ordered Logit Industry Economic Innovation Development Analysis Model

4.3.1

Regression analysis of industrial and economic development

All 10 influencing factors were included in the ordered logit model, and the backward method was used to screen the independent variables, and the variables with significance p>0.05 were analyzed and eliminated one by one in the regression process, and the last ones to enter the model were the significant variables with p≤0.05. The ordered logit regression results are shown in Table 4.

Table 4.

The regression parameter statistics of ordered logit model

Variable	Parametric estimation	SD	z	p
GDP	0.429	0.267	3.088	0.00
PCGDP	0.433	0.090	-4.248	0.00
UR	0.247	0.107	-2.730	0.009
TLUR	0.242	0.087	-1.519	0.115
PCDI	0.701	0.228	-3.165	0.00
CLI	0.154	0.106	1.448	0.137
PIE	0.460	0.113	1.797	0.116
PCG	0.703	0.166	-4.583	0.027
TGE	0.285	0.097	-2.889	0.005
GE	0.398	0.155	4.049	0.00
_Cons	1.821	0.138

Taking the industrial economic innovation development as a reference, the regression process is specified as follows:

The 1st time: analyzing the regression results, it is found that the variables that do not have a significant impact on the degree of innovation and development of the industrial economy are as follows: urban land utilization (TLUR) (p=0.094>0.05), the number of people employed in the industrial economy (PIE) (p=0.682>0.05), the cost of living index (CLI) (p=0.312>0.05) and other three variables. After comparing, the variable number of people employed in the industrial economy (PIE) was excluded from the next round of regression.

2nd: The variables that do not satisfy the test of significance are: town land utilization rate (TLUR) (p=0.085>0.05), cost of living index (CLI) (p=0.308>0.05), then the variable cost of living index (CLI) is excluded from the next regression.

After a total of three regressions, we obtained seven variables that have a significant impact on industrial economic innovation: per capita GDP, the proportion of urban population, per capita disposable income of residents, the percentage of the number of college graduates, the total amount of government financial expenditures, and the efficiency of the government.

The regression parameter statistics of the ordered Logit model are shown in Table 4. It can be concluded: 1)

The regression coefficients of the level of economic development (GDP, PCGDP) are 0.429 and 0.433, showing a significance at the level of 0.05 (p=0.00<0.05), which indicates that the level of economic development exerts a significant positive influence on the development of industrial economic innovation. When the level of economic development (GDP, PCGDP) increases by one unit, the level of industrial economic innovation development increases by 34.58% and 28.38%.

2)

The regression coefficients of the proportion of urban population, per capita disposable income, and the number of college graduates as a percentage of the population are 0.247, 0.701, and 0.703, which show a significance at the 0.05 level (p=0.01<0.05), which indicates that the proportion of urban population, the per capita disposable income of the residents, and the number of college graduates have a significant positive impact on the industrial economic innovation development. When the proportion of urban population, per capita disposable income, and the number of college graduates increase by one unit respectively, the level of industrial economic innovation and development increases by about 24.35%, 26.63%, and 36.32% respectively.

3)

The regression coefficients of governmental role (total governmental financial expenditure, governmental efficiency) are 0.285 and 0.398 respectively, showing significance at the 0.05 level (p=0.02<0.05), which indicates that the governmental role exerts a significant positive influence relationship on the level of economic innovation and development. When the government’s role increases by one unit, the level of economic innovation development increases by 42.33% on average.

The regression results obtained need to be tested to increase the persuasiveness of the model, perform the model fit goodness-of-fit test and compare and analyze. In the fit test of the model, the multinomial Logit model is mainly based on the statistical likelihood function value (LR), likelihood ratio index McFadden’s R2, information criterion (AIC, BIC), and so on. Ordered Logit model fitting test statistics are likelihood function value LR, likelihood ratio index Pseudo R2, information criterion (AIC, BIC).

It can be seen from the regression process, as the number of independent variables decreases, the log-likelihood function value LR, AIC and BIC values in the overall decreasing, then it shows that the model fitting effect is improving. Where the likelihood ratio index McFadden’s R2 (Pseudo R2) does not change significantly throughout the regression process, probably due to the relationship between the independent variables in the sample data used to have a good effect on the dependent variable alone.

Comparative analysis of the statistics of the two models revealed that the log-likelihood function values and likelihood ratio indices of the multinomial Logit model were smaller than those corresponding to the ordered Logit model, indicating that the fit of the ordered Logit model was better than that of the multinomial Logit model. The AIC value of the ordered Logit is smaller than the AIC value of the multinomial Logit model, and the BIC value of the ordered Logit is larger than the BIC value of the multinomial Logit model, indicating that the fitting effect of the multinomial Logit model is better than that of the ordered Logit model.

In comprehensive comparative analysis, the fitted effect of the ordered Logit model is better than that of the multinomial Logit model, which is more applicable to the analysis of factors influencing industrial economic innovation. The goodness-of-fit test statistics for the Multinomial Logit and Ordered Logit models are shown in Table 5.

Table 5.

Goodness of fit test statistics of multinomial Logit and ordered Logit

Model	Test statistic	Model regression(N=20)
Model	Test statistic	A-1	A-2	A-3	A-4	A-5
Multinomial Logit	Log likelihood	-3060.918	-3060.951	-3061.478	-3063.398	-3065.066
	McFadden’s R²	0.126	0.064	0.100	0.068	0.060
	AIC	6173.791	6171.873	6170.879	6170.759	6172.084
	BIC	6331.527	6323.608	6316.502	6304.205	6299.503
Ordered Logit	Log likelihood	-3023.209	-3025.392	-3025.508	-3027.422
	Pseudo R²	0.065	0.097	0.122	0.091
	AIC	6146.486	6146.759	6142.931	6142.883
	BIC	6449.841	6438.002	6422.006	6409.790

4.3.2

Empirical analysis of the level of innovation and development of the industrial economy

The industry for innovation development is further subdivided into two indicators: the level of development and the degree of development, and regression analysis is carried out using the ordered Logit model. Table 6 exhibits an empirical analysis of the level of industrial economic innovation and development.

Table 6.

Ordered logit regression of the development level of industrial innovation

Variable		Model 1	Model 2	Model 3	Model 4	Model 5
Economic development level	GDP	0.582**(0.027)
Economic development level	PCGDP	1.738***(0.000)
Level of urbanization	UR		0.793**(0.038)
Level of urbanization	TLUR		0.353(0.212)
Living standard	PCDI			0.724**(0.027)
Living standard	CLI			-0.131(0.822)
Labor force	PIE				0.453(0.31)
Labor force	PCG				0.731**(0.013)
Government action	TGE					0.925**(0.023)
Government action	GE					1.284**(0.008)

Model 1: In terms of the level of economic development, the coefficients of GDP and PCGDP on the level of development are 0.582 and 1.738, indicating that the higher the level of economic development, the higher the level of innovation and development of the industrial economy, which may be due to the fact that the industrial transformation and upgrading itself will also bring about economic development.

Model 2: For the urbanization level, the coefficients of urban population share and urban land utilization rate on the development level are 0.793 and 0.353 respectively, but the regression result of urban land utilization rate is not significant. The reason for this may be that the urban land utilization rate only reflects the proportion of developed land and fails to further clarify the different nature of land, such as industrial and ecological land. And the higher proportion of urban population indirectly indicates the larger city scale and the more complete industrial innovation elements.

Model 3: In terms of the living standard of residents, only the per capita disposable income of residents shows statistical significance, and the coefficient on the development level is 0.724, indicating that the higher the per capita disposable income of residents, the better the level of industrial economic innovation and development. It is hypothesized that this is because the increase in disposable income pushes consumption, forcing industries to carry out supply-side reforms and innovations to cater to higher consumption requirements.

Model 4: In terms of labor force, the coefficients of the number of employees and the ratio of the number of college graduates on the level of development are 0.453 and 0.731, respectively, since the increase in the number of employees may only reflect the labor-intensive characteristics of the current industry, but does not necessarily indicate that technological innovation has been carried out in the industry. While the percentage of the number of efficient graduates reflects the quality level of the labor force, which is the necessary talent base for the development of industrial and economic innovation, the percentage of the number of college graduates shows statistical significance, while the number of employees variable does not pass the statistical test.

Model 5: Regarding the role of the government, the coefficients of total government financial expenditure and government efficiency on the level of industrial economic innovation and development are 0.952 and 1.248. This is probably due to the implementation of the socialist market economy in China, the government plays a relatively positive role in correcting the market deficiencies and promoting the development of industrial innovation, so the government plays its due role in having a positive impact on the development of industrial economic innovation.

4.3.3

Empirical analysis of the degree of innovation and development of the industrial economy

The development level of industrial economic innovation is empirically analyzed, and Table 7 shows the results of the analysis. It can be found that the influence of each variable on the degree of development of industrial economic innovation is about the same as that on the level of development of industrial economic innovation, which further verifies the role of each variable on the development of industrial economic innovation. Therefore, we can use the following strategies to promote the development of industrial economic innovation.

Table 7.

Ordered logit regression of the development degree of industrial innovation

Variable		Model 1	Model 2	Model 3	Model 4	Model 5
Economic development level	GDP	0.525**(0.05)
Economic development level	PCGDP	0.766***(0.008)
Level of urbanization	UR		0.741**(0.033)
Level of urbanization	TLUR		0.334(0.26)
Living standard	PCDI			0.706***(0.076)
Living standard	CLI			0.34(0.29)
Labor force	PIE				0.55(0.145)
Labor force	PCG				0.919**(0.082)
Government action	TGE					1.355***(0.000)
Government action	GE					1.254***(0.002)

(1) First of all, we should continue to promote the economy, especially the construction of the real economy, to provide a real solid soil for industrial transformation and upgrading. (2) The urbanization process should be reasonably promoted to give full play to the advantages of intra-city and inter-city industrial clusters, so as to provide a complete range of supporting industries for innovative development. (3) Hiding wealth from the people is one of the keys to innovative development, and it is necessary to enhance the consumption ability and confidence of the residents to provide a strong impetus for industrial innovation and development. (4) Talent resources are the first resource. High-quality talents are indispensable for the development of new productive forces and the realization of high-quality development, so it is necessary to increase the attraction of talents and provide facilities for them. (5) Although the market plays a decisive role in the allocation of resources, the government should continue to support new industries, improve government efficiency, and encourage and facilitate the innovative development of the industrial economy by streamlining procedures, providing market information and practical policies.

5

Conclusion

In this paper, urbanization level, residents’ living standard, labor force, and government’s role are selected as influential variables for ordered logit regression analysis respectively. The correlation analysis demonstrates that the correlation coefficients among the 10 secondary variables are below 0.8, and there is no issue of multiple covariance, fulfilling the requirement of ordered Logit regression. Excluding variables with significance p>0.05 one by one, all variables have significant effects on the level and degree of industrial economic innovation development, except for three variables that are not statistically significant, such as urban land utilization rate (TLUR), number of people employed in industrial economy (PIE), and cost of living index (CLI). In addition, the fitted effect of the ordered Logit model is better than that of the multinomial Logit model, which verifies the rationality of the method of this paper. Finally, it can be inferred that the strategy for industrial economic innovation development should be formulated from the aspects of economic construction, industrial clusters, boosting consumption, attracting talent, and government support.

Sprache:: Englisch

Zeitrahmen der Veröffentlichung:: 1 Hefte pro Jahr
Fachgebiete der Zeitschrift:: Biologie, Biologie, andere, Mathematik, Angewandte Mathematik, Mathematik, Allgemeines, Physik, Physik, andere

Zeitschrift RSS Feed

Research on Industrial Economic Innovation Development Strategy Based on Ordered Logit Modeling

Hui Wang

Online veröffentlicht: 21. März 2025

Eingereicht: 27. Okt. 2024

Akzeptiert: 22. Feb. 2025

DOI: https://doi.org/10.2478/amns-2025-0578

SchlüsselwörterOrdered logit model, Industrial economy, Innovative development, Regression analysis

© 2025 Hui Wang, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Schlüsselwörter
Ordered logit model, Industrial economy, Innovative development, Regression analysis