This thesis examines how different methodological choices can affect the measurement of energy poverty and its representation at the regional level. The analysis uses microdata from the Italian Household Budget Survey and focuses, for the main empirical application, on Italian regions in 2023. Household energy-expenditure shares are modelled using a Beta distribution and are used to identify a candidate endogenous threshold. More specifically, the threshold corresponds to the rightmost admissible solution of f′′′(s)=0, representing a stationary-curvature point of the estimated density. The conditions α>1 and β>3 are used as sufficient conditions for the geometric configuration considered in the analysis. The endogenous threshold is then compared with the fixed 10% threshold through four specifications combining fixed or endogenous thresholds with a national or regional economic reference. The results show that the estimated endogenous thresholds range from approximately 11.1% to 19.1% in 2023 and are close to 15% on average. Moving from the endogenous specification with national conditioning to the one with regional conditioning produces only a small change in average incidence, from 2.69% to 2.49%, while concealing larger changes across individual regions. The spatial analysis also shows that the statistical structure and the associations with regional characteristics vary according to the indicator considered. Overall, the results show that the measurement of energy poverty is sensitive to the choice of threshold and geographical reference. The endogenous threshold is therefore treated as an experimental tool that complements existing measures rather than as a universally valid threshold. The results of the spatial analysis are also interpreted as associations rather than causal relationships.
La tesi analizza come diverse scelte metodologiche possano influenzare la misurazione della povertà energetica e la sua rappresentazione a livello regionale. L’analisi utilizza i microdati dell’Indagine sui Bilanci delle Famiglie e si concentra, per l’applicazione empirica principale, sulle regioni italiane nel 2023. La quota di spesa energetica delle famiglie viene modellata attraverso una distribuzione Beta e viene utilizzata per individuare una soglia endogena candidata. In particolare, la soglia corrisponde alla soluzione ammissibile più a destra della condizione f′′′(s)=0, cioè a un punto di curvatura stazionaria della densità stimata. Le condizioni α>1 e β>3 sono utilizzate come condizioni sufficienti per la configurazione geometrica considerata. La soglia endogena viene poi confrontata con la soglia fissa del 10% all’interno di quattro specificazioni, costruite combinando soglie fisse o endogene con un riferimento economico nazionale o regionale. I risultati mostrano che nel 2023 le soglie endogene stimate variano tra circa l’11,1% e il 19,1% e si collocano mediamente intorno al 15%. Il passaggio dalla specificazione endogena con riferimento nazionale a quella con riferimento regionale modifica solo leggermente l’incidenza media, dal 2,69% al 2,49%, ma nasconde variazioni più consistenti tra le singole regioni. L’analisi spaziale mostra inoltre che la struttura statistica e le associazioni con le caratteristiche regionali cambiano a seconda dell’indicatore considerato. Nel complesso, i risultati evidenziano come la misurazione della povertà energetica sia sensibile alla scelta della soglia e del riferimento geografico. La soglia endogena viene quindi considerata come uno strumento sperimentale e complementare alle misure esistenti, e non come una soglia universalmente valida. I risultati dell’analisi spaziale sono inoltre interpretati come associazioni e non come relazioni causali.
Energy Poverty Scenarios with Endogenous Thresholds: A Class of Distributional Indicators Based on the Beta Distribution
ROCCHEGIANI, MIRCO
2025/2026
Abstract
This thesis examines how different methodological choices can affect the measurement of energy poverty and its representation at the regional level. The analysis uses microdata from the Italian Household Budget Survey and focuses, for the main empirical application, on Italian regions in 2023. Household energy-expenditure shares are modelled using a Beta distribution and are used to identify a candidate endogenous threshold. More specifically, the threshold corresponds to the rightmost admissible solution of f′′′(s)=0, representing a stationary-curvature point of the estimated density. The conditions α>1 and β>3 are used as sufficient conditions for the geometric configuration considered in the analysis. The endogenous threshold is then compared with the fixed 10% threshold through four specifications combining fixed or endogenous thresholds with a national or regional economic reference. The results show that the estimated endogenous thresholds range from approximately 11.1% to 19.1% in 2023 and are close to 15% on average. Moving from the endogenous specification with national conditioning to the one with regional conditioning produces only a small change in average incidence, from 2.69% to 2.49%, while concealing larger changes across individual regions. The spatial analysis also shows that the statistical structure and the associations with regional characteristics vary according to the indicator considered. Overall, the results show that the measurement of energy poverty is sensitive to the choice of threshold and geographical reference. The endogenous threshold is therefore treated as an experimental tool that complements existing measures rather than as a universally valid threshold. The results of the spatial analysis are also interpreted as associations rather than causal relationships.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi_Magistrale_Rocchegiani_Mirco_FINALE_PDFA.pdf
accesso aperto
Dimensione
1.02 MB
Formato
Adobe PDF
|
1.02 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/28432