Data from: Re-assessing causality between energy consumption and economic growth

The energy consumption-growth nexus has been widely studied in the
empirical literature, though results have been inconclusive regarding the
direction, or even the existence, of causality. These inconsistent results
can be explained by two important limitations of the literature. First,
the use of bivariate models, which fail to detect more complex causal
relations, or the ad hoc approach to selecting variables in a multivariate
framework; and, second, the use of linear causal models, which are unable
to capture more complex nonlinear causal relationships. In this paper, we
aim to overcome both limitations by analysing the energy
consumption-growth nexus using a Flexible Fourier form due to [1] The
analysis focuses on the US over the period 1949 to 2014. From our results
we can conclude that, where the linear methodology supports the neutrality
hypothesis (no causality between energy consumption and growth), the
Flexible Fourier form points to the existence of causality from energy
consumption to growth. This is contrary to the linear analysis, suggesting
that lowering energy consumption would adversely affect US economic
growth. Thus, by employing the Flexible Fourier form we find the
conclusions can be quite different.

能源消费与增长的关联关系（energy consumption-growth nexus）已在实证研究文献中得到广泛探讨，但关于因果关系的方向乃至因果关系本身是否存在，学界始终未能得出一致结论。这类不一致的研究结果，可归因于现有实证文献存在的两处关键局限：其一，要么采用双变量模型（无法识别更为复杂的因果关联），要么在多变量框架下采用特设变量选择法；其二，使用线性因果模型，这类模型无法捕捉更为复杂的非线性因果关系。本文采用源自文献[1]的柔性傅里叶形式（Flexible Fourier form）分析能源消费与增长的关联关系，以期克服上述两处局限。本次分析的研究对象为1949年至2014年的美国经济。基于本文的研究结果可得出如下结论：当线性方法支持中性假说（即能源消费与经济增长间不存在因果关系）时，柔性傅里叶形式方法则表明存在从能源消费到经济增长的因果关系。这一结论与线性分析的结果相悖，意味着降低能源消费将对美国经济增长产生不利影响。由此可见，采用柔性傅里叶形式方法后，所得研究结论可能会存在显著差异。