ZHAW SML Nearshoring Index

The Center of European Business at the ZHAW School of Management and Law partnered with swissICT and ISSS to survey Swiss IT service firms about the determinants of nearshoring decisions. The survey yielded 56 high-quality responses: 82 percent of respondents (46 firms) are actively engaged in European nearshoring, 41 percent have more than 100 employees, and roughly one quarter report revenues above USD 50 million.

Respondents rated the importance of individual factors on a seven-point scale — 1 being not important at all and 7 very important — which allowed the identification of five major groups of determinants.

The index is built from five pillars: Economic, Labor, Institutional, Social, and Location factors. Each pillar combines multiple weighted variables corresponding to the survey questions, and the overall IT Nearshoring Index is a weighted average of the five pillars.

The index relies on Eurostat NUTS 1 regional data covering 115 European regions, with variables aligned to the survey questions — for instance, market potential is measured through GDP, and accessibility through the distance from Bern and airport passenger volumes.

Where data are missing at the NUTS 1 level, values are aggregated from NUTS 2 observations using population weights; country-level variables are applied uniformly across a country's regions. The primary source is the Eurostat regional database, supplemented by the specialized sources listed in the data documentation.

Missing observations are imputed within each pillar using a multi-equation random forest approach (Wulff & Ejlskov, 2017), which allows systematic handling of coverage gaps across roughly fifty variables while maintaining dataset integrity.

All variables are normalized to a 0–100 scale using min-max (and max-min) normalization, where 0 reflects the least attractive and 100 the most attractive location, with linear interpolation in between. For example, for hourly IT wages — ranging from roughly 4 EUR in Northern Bulgaria to roughly 40 EUR in Sweden — the lowest-wage region scores 100 and the highest-wage region scores 0, with all other regions distributed along the continuum.

Variable weights within each pillar are derived from the survey responses using the rank-sum method (Danielson & Ekenberg, 2017). When several variables map to a single survey question, the question's importance weight is distributed equally among them.

The pillar weights themselves come from expert interviews in which experts assessed each pillar's importance directly in percentage terms rather than as ranked preferences. A comparative analysis shows that rank-sum weighting of the pillars would substantially reduce the weight of the economic pillar and slightly reshuffle regional rankings.