This study employs a panel dataset on forest area of 22 countries and for the time period 1995–2016 in order to provide an up-todate assessment on determinants of forest area change. Our set of explanatory variables is drawn from the Environmental Kuznets curve hypothesis and well-established factors affecting forest area change. In this respect, macroeconomic stability, socioeconomic development, institutional efficiency and quality of governance are taken into account to assess the cumulative nature of changes in forest cover. Using static and dynamic specifications our analysis produce different results with an inverted U-shape curve only in the static case and an inverted N-shape relationship in the dynamic specifications. Specifically, using an appropriate fixed effects model validates the EKC hypothesis finding an inverted U-shaped curve with high within-the-sample turning points. The random coefficients method applied does not support the EKC hypothesis in all possible specifications while the orthogonal deviations GMM estimates indicate the presence of an inverted U-shape relationship in all cases with lower cases within-the-sample turning points. A slow adjustment rate is found in all dynamic specifications indicating the need of around four time periods for a country to adjust to the LR equilibrium level of forest recovery.

本研究采用了 1995-2016 年期间 22 个国家森林面积的面板数据集，以提供对森林面积变化决定因素的最新评估。我们的一组解释变量来自环境库兹涅茨曲线假设和影响森林面积变化的公认因素。在这方面，宏观经济稳定性、社会经济发展、制度效率和治理质量被考虑在内，以评估森林覆盖变化的累积性质。使用静态和动态规范，我们的分析产生了不同的结果，仅在静态情况下具有倒 U 形曲线，而在动态规范中具有倒 N 形关系。具体来说，使用适当的固定效应模型验证了 EKC 假设，发现具有高样本内转折点的倒 U 形曲线。所应用的随机系数方法在所有可能的规范中都不支持 EKC 假设，而正交偏差 GMM 估计表明在所有情况下都存在倒 U 形关系，样本转折点内的情况较低。在所有动态规范中都发现了缓慢的调整率，表明一个国家需要大约四个时间段来调整到森林恢复的 LR 平衡水平。所应用的随机系数方法在所有可能的规范中都不支持 EKC 假设，而正交偏差 GMM 估计表明在所有情况下都存在倒 U 形关系，样本转折点内的情况较低。在所有动态规范中都发现了缓慢的调整率，表明一个国家需要大约四个时间段来调整到森林恢复的 LR 平衡水平。所应用的随机系数方法在所有可能的规范中都不支持 EKC 假设，而正交偏差 GMM 估计表明在所有情况下都存在倒 U 形关系，样本转折点内的情况较低。在所有动态规范中都发现了缓慢的调整率，表明一个国家需要大约四个时间段来调整到森林恢复的 LR 平衡水平。