2. Explanatory variables

Industrial structure (IS). With the total ban on commercial logging in state-owned forest areas, forest areas that once had tree harvesting and wood product processing as their main economic development model must adjust their industrial structure in order to steadily develop their economy while ensuring sustainable forest development. At present, different forestry bureaus in forest areas have different dominant industries, and therefore their main mode of operation is not the same. The primary forestry industry's main mode of operation is the above-mentioned planting and breeding of forest products; the secondary forestry industry's main mode of operation is the processing of wood products; and the tertiary forestry industry's main mode of operation is the vigorous development of forest tourism and the service industry in the process. In the process of development and operation of these industries, there is bound to be spatial spillover and spatial benefit of forestry economic development, therefore the industrial structure is also one of the influencing factors of FECO. To make this study dynamic, this paper uses the proportion of forest industry output value to total forestry output value to express [44].

Economic development (PGDP). This paper uses GDP per capita to reflect the sum of the value of products in a country or region in that year, which to a certain extent shows the degree of economic development of a region. Economic growth can promote scientific and technological progress, and a high level of economic development will lead to a higher demand for technology by the inhabitants of that place, which will lead to a higher eco-efficiency of local production through a demand-induced effect [45]. In economically developed areas, urbanization is bound to develop rapidly, which leads to the rapid development of the real estate and decoration markets, which provide good opportunities for the development of the forest products market. Meanwhile, people's demand for a good ecological environment is increasing. Urban gardening, greening, forest tourism, tourism forestry and other industries will develop rapidly. In summary, a good economic environment can promote forestry industry development as well as the optimization and upgrading of the forestry structure.

Foreign direct investment (FDI). Based on the analysis of FDI technology spillover channels, from the perspective of the competition effect, foreign enterprises usually enter China's forestry field by virtue of their capital, technology, scale and other advantages. On the one hand, they will introduce advanced technology, which is conducive to improving forest enterprises' ability to introduce, absorb and apply new technologies and promote technological progress; these can be understood as improvements brought by competition, which are positive spillover effects. On the other hand, forestry FDI generally does not choose to invest in greenfield sites, but is more involved in competition for existing forestry resources and markets, leading to a reduction in the market share of local enterprises, and coupled with China's preferential policies for foreign investors, it is easy to squeeze out domestic capital, thus inhibiting the improvement of TFP, which is a negative spillover effect. Whether the positive or negative effect is larger or smaller has not been determined, but to some extent, it can explain the negative effect of FDI on total factor productivity in Chinese forestry [46]. In this paper, the ratio of forestry FDI to total forestry output is used to measure FDI intensity.

Investment in scientific research (TI). Forestry is one of the most special industries in the national economic system, with very strong benefit spillovers. In addition to providing a large number of products and services for people's lives and social production, there are also generally sizable ecological benefits. In addition, forestry has an important role in ensuring the basic livelihood of forest farmers in forest areas and in rural revitalization. The most prominent feature of forestry is the long cycle, which determines a longer scientific research cycle, so the adequacy and stability of scientific research funding is particularly important [47]. In the paper, the ratio of research funding to GDP is chosen to represent the intensity of research funding.

Market-based environmental regulation (ER). Environmental regulation includes command-and-control and market-based. The former mainly includes setting environmental standards, pollutant emission standards and technical standards; the latter mainly includes establishing an emission charging or taxation system and an emission rights trading system [48]. In this paper, we use the proportion of the total emission fee levied by the forestry industry to the total regional forestry output value.

Due to the constraints of the eco-efficiency index calculation formula, the eco-efficiency values measured by the DEA method range between 0 and 2. In this case, if the traditional ordinary least squares (OLS) method is used to analyze the actual effect of each influencing factor on eco-efficiency, the results will be biased and inconsistent. In order to avoid the bias, the Tobit model was selected to analyze the factors influencing forestry eco-efficiency. Based on the above variable selection, the Tobit model was constructed as follows:

$$\text{FECO}\_{it} = \mathcal{c} + \mathcal{B}\_1 IS\_{it} + \mathcal{B}\_2 PGDP\_{it} + \mathcal{B}\_3 FDI\_{it} + \mathcal{B}\_4 TI\_{it} + \mathcal{B}\_5 ER\_{it} + \varepsilon\_{it\prime} \tag{10}$$

where *β* is the elasticity coefficient of variables; FECO*it* is the explanatory variable of FECO; *ISit*, *PGDPit*, *FDIit*, *T Iit* and *ERit* are the industrial structure, economic development, foreign direct investment, investment in scientific research and environmental regulation, respectively. The data are mainly obtained from the statistical yearbooks of each province.

#### **4. Results**

#### *4.1. Results of Regional FECO Measurement in China*

This paper is based on the input-output panel data from 2008 to 2021 (Tibet is not included in the analysis because of incomplete data), and the FECO of each province is measured based on EMS software. Considering the existence of regional heterogeneity, this paper divides the 30 provinces (regions and municipalities) into three major regions: east, central and west, with 11 provinces (municipalities) in the east, 8 provinces in the central and 11 provinces in the west. The measurement results are shown in Table 2.

**Table 2.** Results of forestry eco-efficiency by province in China from 2008 to 2021.


From the above table, it can be seen that from 2008–2021, the integrated efficiency of Shandong and Shanghai exceeds or approaches 1.0, and the integrated efficiency of Jiangsu, Jiangxi and Fujian approaches 0.9. Shandong, Shanghai and Jiangsu show an upward trend, indicating that the above regions have been effective in adjusting the balance between forestry production output and environmental pollution. The comprehensive efficiency of provinces such as Inner Mongolia, Heilongjiang, Jilin and Liaoning is below 0.3. Although the forestry resource stock and forestry output values of these provinces are high, the input consumption in forestry production is too large, which makes their comprehensive efficiency hover at a low level. The low ecological overall efficiency of forestry in Beijing and Tianjin is due to the innate condition of their limited forestry resource stocks. The comprehensive ecological efficiency of forestry in western provinces such as Xinjiang, Gansu and Qinghai is less than 0.2. These regions are constrained by topography, resources and other factors that do not release the scale benefit; the desertification of land is serious, which affects forestry production and leads to low comprehensive efficiency.

The average FECO values of the three major economic regions in the country in the 14-year period are, in descending order, the eastern, central and western regions. The mean value FECO in the eastern region is significantly higher than the national average and the rest of the provinces and cities, except Hebei, have relatively high efficiency. Hebei province is a special case and in the preferred position to undertake the transfer of high pollution industries from Beijing and Tianjin, resulting in its low FECO. Beijing, because of its special location and position as a political and economic center, has a low level of FECO due to its inherent condition of limited forestry resource stock; generally speaking, the forestry input and output of the eastern provinces and cities are more reasonable. The FECO in the central region is basically above 0.5, except for Jilin and Heilongjiang. Shanxi Province, with a forest cover of 20.5% in 2016, has relatively few forestry resources and mainly focuses on coal energy production, and pays less attention to the development and utilization of forestry resources, resulting in a backward production technology level; this leads to a low FECO. Hunan Province, with abundant forestry resources, has a low FECO level in the early stage due to the model of exchanging resources for economic development, and then in the process of undertaking industrial transformation, actively adjusts the strategic structure for industry. In the process of undertaking industrial transformation, Hunan Province actively adjusted its strategic structure for industrial upgrading and transformation, and FECO reached above 0.5 after 2015. The overall FECO value in the western region is low, but Chongqing's FECO is much higher than other western regions due to its special geographical location and economic level in the west.
