*4.2. Empirical Models*

In this paper, the DID method will be adopted and the implementation of the BTH coordinated development strategy will be put forward as a quasi-natural experiment. Therefore, the following DID models are constructed:

$$\text{INIDEN}\_{it} = \beta \text{BTH}\_{i} \cdot \text{Post}\_{it} + \gamma X\_{it} + \alpha\_{i} + \psi\_{t} + \varepsilon\_{it} \tag{1}$$

$$INDDNA\_{it} = \beta BTF\_i \cdot Post\_{it} + \gamma X\_{it} + \alpha\_i + \psi\_t + \varepsilon\_{it} \tag{2}$$

*INDSDit* = β*BTHi* · *Postit* + γ*Xit* + α*<sup>i</sup>* + ψ*<sup>t</sup>* + ε*it* (3)

$$INDST\_{it} = \beta BTH\_i \cdot Post\_{it} + \gamma X\_{it} + \alpha\_i + \psi\_t + \varepsilon\_{it} \tag{4}$$

Formulas (1)–(4) are the DID estimation models that take time and city as fixed effects into account. In all models, *i* stands for the city from 1 to 30 and *t* stands for the year from 2007 to 2017. The dependent variables *INDENit*, *INDWAit*, *INDSDit*, and *INDSTit* respectively denote the industrial energy intensity, industrial wastewater emission intensity, industrial sulfur dioxide emission intensity, and industrial dust emission intensity. *Postit* is the dummy variable for the processing time effect of the BTH coordinated development strategy. For *Postit*, the years after 2014 are set as 1 and the previous years are set as 0. *BTHi* is the dummy variable for processing the treatment group, indicating whether the city is located in the BTH region. If it is a city in the BTH region, it is set as 1; otherwise, it is 0. *BTHi*·*Postit* is the interaction term between *BTHi* and *Postit*, which is the core variable concerning by the DID method. Besides, *Xit* is a series of control variables that may cause changes in industrial energy intensity and related pollution emission intensities, including urbanization rate, industrial structure, per capita GDP, FDI, and R&D. ψ*<sup>t</sup>* is the year fixed effect, α*<sup>i</sup> is* the city fixed effects, and ε*it* is the random disturbance term.

The DID method focuses on the coefficient β of the variable *BTHi*·*Postit*, for which the economic implication can be explained by the impact of the BTH coordinated development strategy on industrial energy and pollution intensities. If the coefficients of β are positive and statistically significant, the BTH coordinated development strategy tends to promote industrial energy intensity and related pollution emission intensities. On the contrary, if they are negative and statistically significant, the BTH coordinated development strategy tends to restrain industrial energy intensity and related pollution emission intensities.
