Does digitalisation affect the adoption of electric vehicles? New regional-level evidence from Google Trends data

ABSTRACT

Digitalisation is an important dimension that contributes to fostering the adoption of electric mobility. We investigate this unexplored topic by focusing on the regional level of analysis, and presenting new data and evidence for a large number of regions in Europe, Canada and the United States. The empirical analysis makes use of Google Trends data. It constructs new indicators of digitalisation and the adoption of electric vehicles, as measured by Google search queries. The new dataset contains indicators for 182 regions in 15 countries for the period 2010–23. We use this dataset to carry out a time-series analysis (vector error correction (VEC) model) of the relations between digitalisation and electric vehicles in each region. The results show that digitalisation is an important factor that has fostered the adoption of electric vehicles in the last decade. The analysis, though, also points out that there is considerable heterogeneity in the time-series results among regions in our sample. Digitalisation has a more visible effect on electric mobility for regions that have higher gross domestic product per capita, better internet infrastructures, a young and well-educated workforce, and higher population density.

KEYWORDS:

• digitalisation
• electric vehicles
• Google Trends data
• regions
• vector error correction (VEC) model

JEL:

• C22
• O18
• O33

ACKNOWLEDGEMENTS

The paper was presented at the GEOINNO Conference in Manchester, UK, January 2024. We thank the journal editor and three anonymous reviewers for helpful comments and suggestions on a previous draft of the paper. We also thank Burim Orlishta for research assistance and data-gathering support.

DATA AVAILABILITY STATEMENT

No data are available for this article.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. The Global EV Outlook is available on the IEA website for 2013 and then for every year between 2016 and 2023; https://www.iea.org/search?q=Global%20EV%20Outlook/.

2. We decided not to translate into national languages most of the keywords related to smart manufacturing, smart city and smart appliances. The reason is that these keywords, as they derive originally from the English language, they are commonly used and discussed in English, regardless of the mother tongue spoken in regions. This is the case either because these keywords refer to technical terms or because they are used with reference to their English anacronyms. There is also the case of some services, platforms or concepts whose name is of universal use, and thus there is no meaningful translation into other languages. The translation was performed by individuals who master the respective national languages. In this translation exercise, the objective was not to find a literal translation, but a translation that is meaningful in the national language and that represents an expression that a national individual refers to when he/she searches for a specific concept or idea.

3. When the English language and the mother tongue of the country are coincident (e.g., the case of the United States), we count English language twice. The same happens for keywords that we do not translate from English to other languages: their score in languages other than English is equal to their score in English. In the case of Belgium, although the most spoken language in the country is Dutch, for this analysis, in Belgian regions we only collect data regarding keywords in English and French. For simplicity, for Luxembourg, we also only collected data regarding keywords in English and French.

4. The model specified in equation (5) estimates, therefore, a bivariate time-series relationship between each digitalisation variable and electric mobility. We did not estimate a multivariate version of the model that includes simultaneously all five digitalisation variables because these are highly correlated with each other, and such a multivariate model would therefore present severe multicollinearity issues. Note also that we do not have the possibility to include additional control variables in the model in equation (5), because we do not have the availability of control variables for monthly data at the regional level. However, we control for potentially confounding factors in the meta-regression analysis in which we pool all time-series results for all regions in our sample.

5. When more than half of the observations are zero, the estimation procedure is often fallible due to collinearity problems: the variables vary little over time, and the co-occurrence of many zeros between pairs of variables or between different lags of the same variable is verified.

6. As noted previously, indicators obtained through Google Trends measure relative frequencies of key words vis-à-vis the maximum frequency registered during the entire time span considered in the analysis. As such, these indicators cannot be used to carry out comparisons of these variables among different regions. Hence, the scatterplot in Figure 2 is only meant to provide a general illustration of the relationship between digitalisation and electric mobility in the whole sample, and it should not be interpreted as an indication of a causal relation in a cross-regional setting. To analyse cross-regional heterogeneity econometrically, we will instead use the results of the time-series analysis in a meta-regression analysis, as explained further below.

 

Additional information

Funding

This work was supported by the Research Council of Norway, Intransit Centre [grant number 295021].