Do charging infrastructure and electric vehicle adoption drive each other? The limits of time-series evidence for policy in an emerging electromobility market

摘要

The development of electromobility has become an integral part of the modern automotive industry worldwide. However, when implementing a new technology, decision-makers face a chicken-and-egg dilemma. It is argued that electric vehicles are needed to drive development or charging infrastructure drives adoption. Answers to these questions vary depending on the parameters chosen. This paper seeks to determine whether they drive each other. It relies on cointegration evidence using data from an emerging, rather than a mature market in electromobility. To empirically verify these relationships, a VECM with bootstrap rank tests and Toda-Yamamoto causality tests in 216 variants were applied. A total of 73 monthly observations from December 2019 to December 2025 were analysed. The main control variable was the price of petrol, while reference rates were included in the robustness analysis. The estimated VECM reveals one long-run relation, conditional on petrol prices, from which charging infrastructure can be excluded. In the short term, charging infrastructure does not Granger-cause BEV adoption, but the test lacks power against published effect sizes. The causality detected from EV adoption to charging infrastructure is not robust. In the main specification, the BEV fleet response to a sustained 1% rise in charging points is 0.53% after two years (95% interval 0.16–1.13). Furthermore, the co-evolution in a four-variable model stems from collinear infrastructure measures and asymptotic tests. For public policy, these results imply that national time series cannot unambiguously tell which should come first and that such series alone justify neither push nor pull policies.