PSI - Issue 78

Cristina Cantagallo et al. / Procedia Structural Integrity 78 (2026) 1482–1489

1486

v. Economic Indicators : Income levels, the distribution of income recipients, and estimates of value added and local Gross Domestic Product (GDP) were collected from provincial, regional, and national sources. vi. Employment Structure : The distribution of employment across economic sectors was obtained from the annual ISTAT reports. This information supports the identification of the relative weight of tourism and related activities in local labor markets. 4.2. Construction of the Tourism Dependency Index A Tourism Dependency Index (TDI) was developed to quantify the degree to which local economies depend on tourism-related activities and accommodation infrastructure. This synthetic indicator was designed to provide a standardized metric for comparing municipalities with different demographic and structural characteristics. The index construction was based on a multivariate statistical approach combining Principal Component Analysi s (PCA) and Rotated Factor Analysis (Kaiser 1958). The model considered five core variables, selected for their relevance in capturing the structural and functional dimensions of tourism dependency: 1) Tourist Presences per Capita , defined as the annual number of overnight stays in relation to the resident population; 2) Tourist Density per km² , representing the concentration of tourist presences per unit of municipal surface area; 3) Accommodation Capacity per Capita , measuring the ratio between the number of available beds and the resident population; 4) Density of Beds in Private Housing , reflecting the incidence of short term rentals and holiday homes relative to total dwellings; 5) Share of Tourist Presences in Private Housing , indicating the proportion of total overnight stays registered in non-hotel facilities. The TDI was calculated by first standardizing each variable with respect to the maximum observed value in the reference dataset (min – max normalization to a 0 – 1 range). Each standardized variable was then weighed using the coefficients derived from the rotated factor loadings corresponding to the principal components retained in the analysis. The resulting composite Raw Score was obtained by summing the normalized input variables multiplied by their respective weights, as shown in Eq. (1):

5

1 = =  i

(

)

(1)

Raw Score

X×w

i

i

where X i represents the standardized value of variable i , and w i is the associated factor loading. The Raw Score was subsequently rescaled to a 0 – 100 scale using Eq. (2):

Raw Score Minimum Observed Score Maximum Observed Score Minimum Observed Score − =  −

100

TDI

(2)

In Eq. (2), Raw Score denotes the unscaled composite score obtained through the weighted aggregation of standardized variables, Minimum Observed Score indicates the lowest raw score recorded among all municipalities included in the analysis and Maximum Observed Score corresponds to the highest raw score in the same dataset. This transformation ensures that all input indicators contributed proportionately to the composite index, regardless of their original units of measurement. Higher scores indicate a stronger dependence on tourism as an economic driver. A score exceeding the national reference mean (13.33) suggests above-average exposure to tourism-related fluctuations and potentially higher vulnerability to external shocks. Conversely, scores below this threshold are associated with relatively diversified local economies. The reference mean was derived from a nationwide dataset of Italian municipalities compiled for benchmarking purposes. 4.3. Scenario Modelling of Seismic Impact The scenario analysis was developed to estimate the potential socio-economic impact of a major seismic event that could affect Popoli Terme (PE). For this purpose, a benchmark including municipalities within the seismic crater of the Amatrice – Norcia – Visso earthquake series was selected. For each benchmark municipality, data were compiled on