Anti-globalisation sentiment has risen significantly in recent decades. Focusing on French voting patterns since 1981, this column analyses the relationship between anti-globalisation votes and local exposure to exports and imports, and finds that anti-globalisation is driven by imports of final goods rather than intermediate goods. In fact, greater exposure of national employment to intermediate imports, which complement certain economic activities, reduces the share of votes against globalisation. Similarly, exports mitigate the rise in anti-globalisation votes through their positive effects on economic activity in the most affected regions.
Trade globalisation impacts labour markets in various ways. When globalisation enables domestic companies to export, it can stimulate activity in the affected sectors, improve earnings, and contribute to job creation. However, it can also lead to the economic and social downgrading of the workers most exposed to competition from imported goods that are already produced locally. When imports consist of goods that are not produced locally but are used by local firms – either for domestic sales or for exports – they can help stimulate local economic activity and improve wage and employment conditions. Thus, the relationship between exports and individuals’ economic opportunities is relatively straightforward, while the economic impact of imports should depend on whether they are intermediate or final goods. As a result, perceptions of globalisation may be ambivalent. Because of its diverse economic effects on the labour market, voting behaviour toward political platforms that support or oppose globalisation may depend on the concentration of employment in sectors most exposed to imports (whether intermediate or final) and exports at the local level.
Focusing on Germany between 1987 and 2009, Gold et al. (2016) show that an increase in exports per worker to low-wage countries decreases the electoral success of the far right, while a rise in import exposure to low-wage countries increases it. Barone and Kreuter (2021) find that an increase in imports from China raises the populist vote in Italy. For the US, Autor et al. (2020) find that counties with an initial white majority are more likely to vote Republican after a rise in exposure to import competition from China. Docquier et al. (2024) estimate the impact of imports and immigration on votes for populist parties across 55 countries over the period 1960–2018, and find that an increase in imports of unskilled labour-intensive goods boosts electoral support for right-wing populism, while an increase in imports of skilled labour-intensive goods reduces it.
In recent research focusing on French voting patterns since 1981 (Bouët et al. 2024), we analyse the relationship between anti-globalisation votes and local exposure to exports and imports, whether of intermediate or final goods, and show that:
- local labour market exposure to exports decreases anti-globalisation votes,
- local labour market exposure to imports of final goods increases anti-globalisation votes, and
- local labour market exposure to imports of intermediate goods decreases anti-globalisation votes.
One of the key contributions of this research is the distinction between final and intermediate goods. Additionally, we refine the measurement of anti-globalisation votes by conducting a textual analysis of candidates’ manifestos from the first round of eight presidential elections between 1981 and 2022.
Geographical concentration of exposure to international trade
Figure 1 shows local differences in the exposure of employment to variations in imports and exports between 1988 and 2021. A department’s exposure to exports or imports of goods is calculated by totalling the changes in trade per worker in each sector, weighted by the sectors’ shares in total employment. While France’s average exposure to imports rose by $56,500 per worker over the period, there are significant local disparities. Employment in some northern and eastern departments is concentrated in sectors that experienced a sharp rise in imports at the national level, but some southern and western departments show a different pattern over the same period. In contrast, regions such as Paris specialise in sectors where both imports and exports have increased significantly.
Figure 1 Change in employment exposure to trade between 1988 and 2021, by department (in dollars per worker)
Source: Authors’ calculations based on data from the 1982 census and the COMTRADE database
The rise of anti-globalisation voting
A textual analysis of the content of each candidate manifesto in the first round of presidential elections shows a rise in the opposition to globalisation. The proportion of sentences in the manifestos that oppose globalisation (from which pro-globalisation sentences have been subtracted) has risen from around 1% in 1981 to 4% in 2022.
By combining the textual content of candidates’ manifestos from the first round of presidential elections with their electoral success, we construct a measure of the local share of anti-globalisation votes. At the national level, this share increased from about 1% in 1981 to 6% in 2022. Notably, the anti-globalisation content in these manifestos is not exclusive to far-right parties, which have traditionally opposed openness and trade. Since 2007, for instance, candidates from the Gaullist right incorporated substantial anti-globalisation themes in their platforms. Consequently, the far-right’s share of the total number of anti-globalisation statements in all manifestos decreased from 84.2% in 1988 to 58.3% in 2022.
The proportion of anti-globalisation votes is unevenly distributed across the country (Figure 2). In 1988, these votes were concentrated around the Mediterranean and certain eastern departments, whereas by 2022, they became more prominent throughout the northeastern departments.
Figure 2 Share of anti-globalisation votes in the first round of the French presidential election in 1988 and 2022
Sources: Authors’ calculations based on candidates’ speeches and results of presidential elections.
Contrasting effects of French imports and exports on anti-globalisation voting patterns
Over the 1988–2022 period, we find that anti-globalisation votes react differently according to the nature of exposure to international trade: increased exposure to imports favours anti-globalisation votes, while an increase in exposure to exports reduces it (Figure 3). An increase of one standard deviation in our index of exposure to imports (approximately $18,000 per worker) within a department leads to a 0.17 percentage point rise in its anti-globalisation vote. This magnitude suggests that raising France’s average exposure to imports by $55,500 per worker between 1988 and 2022 increased the share of anti-globalisation votes by approximately 0.5 percentage points during this period.
In contrast, the increase in France’s average exposure to exports – by $46,200 per worker between 1988 and 2022 – would have contributed to a reduction in the share of anti-globalisation votes by 1.1 percentage points. The variation in the share of anti-globalisation votes resulting from an increase in exposure to exports is two to three times larger than that associated with higher exposure to imports.
Figure 3 Effects of exposure to international trade on the share of anti-globalisation votes
Notes: Diamonds represent the value of estimated coefficients; vertical lines represent 95% confidence intervals. When these lines do not intersect the x-axis at zero, the coefficient is statistically significant: exposure to international trade has an effect on the anti-globalisation vote. The explanatory variables in the model are standardised so that the estimated coefficient indicates the percentage point change in the anti-globalisation vote in response to a one standard deviation increase in these variables.
Source: Bouët et al. (2024).
Contrasting effects depending on the type of goods imported
If an increase in a department’s overall exposure to imports leads to a rise in anti-globalisation votes, what occurs when we distinguish between imported goods based on their use? We find that only increased exposure to imports of final goods contributes to the rise in anti-globalisation votes. In contrast, greater exposure to imports of intermediate goods reduces the share of anti-globalisation votes. Specifically, our estimates imply that an increase of $55,500 in imports of final goods per worker raises our anti-globalisation voting index by 0.65 percentage points, while a similar increase in imports of intermediate goods per worker results in a decrease of 0.55 percentage points in anti-globalisation voting.
This asymmetric impact arises because imported final goods may compete directly with domestically produced goods, while imports of intermediate goods may contribute to regional dynamism. More generally, it highlights the complexity of analysing the impact of international trade on electoral behaviour, which must encompass all its dimensions for a complete understanding. Indeed, anti-globalisation voting does not merely reflect a geography of discontent linked to increasing external competition from imports in certain sectors. Globalisation also stimulates economic activity and job creation through exports and imports of intermediate goods, which mitigates the intensity of anti-globalisation sentiment.
Source : VOXeu
