Healthy Life Expectancy at Birth in the Italian Regions

 

It has grown in all regions with the exception of Basilicata

Istat-BES calculates the value of life expectancy in good health. Healthy life expectancy at birth "expresses the average number of years that a child born in a given calendar year can expect to live in good health, using the prevalence of individuals who respond positively (" well " or "very well") to the question about perceived health". This indicator therefore reflects a set of elements that refer to both environmental, food and wine and individual aspects that can have a significant impact in determining the degree of health of the population. In addition, there is a specific economic interest in healthy life expectancy at birth. In fact, a healthier population costs less in terms of the health system and therefore appears to be more sustainable in the long term for the state and the community.

Ranking of the Italian regions by value of healthy life expectancy at birth in 2019. Trentino-Alto Adige is in first place for value of healthy life expectancy with an amount equal to 65.8 years, followed by Friuli-Venezia Giulia with an amount equal to 61.8 years, followed by Valle d'Aosta with an amount of 60.7 years. In the middle of the table there are Liguria with an amount equal to 58.9 years, followed by Piedmont with an amount equal to 58.9 years and the Marche with an amount equal to 58.8 years. Sardinia closes the ranking with an amount equal to 54.4 years, followed by Basilicata with an amount equal to 54.3 years and Calabria with an amount equal to 49.7 years. On average in the Italian regions, life expectancy in good health at birth was equal to a value of 58.32 years.

Ranking of Italian regions by percentage change in life expectancy at birth between 2009 and 2019. Umbria ranks first in terms of the percentage change in healthy life expectancy between 2009 and 2019 with an amount of 8.11% equal to 4.50 years, followed by Valle d'Aosta with an amount equal to 7.05% equal to an amount of 4.00 years, and by Emilia Romagna with an amount of 6.05% equal to an amount of 3.40 years. In the middle of the table there are Veneto with an amount equal to 4.15% equal to an amount of 2.40 years, followed by Lazio with a value of 3.90% equal to an amount of 2.20 years, and Piedmont with a variation of 3.88% equal to an amount of 2.20 years. Liguria closes the ranking with an amount equal to 0.51% equal to an amount of 0.30 years, Molise with a percentage change equal to 0.00% and Basilicata with a change equal to -0.37% equal to to an amount of -0.20 years. On average between 2009 and 2019, the value of the percentage change in life expectancy at birth in the Italian regions increased by an amount equal to 3.57% equal to an amount of 2.02 years.

Italian macro-regions. The healthy life expectancy at birth in the North has grown from an amount of 57.4 years in 2009 to 60.1 years in 2019 or a change equal to an amount of 2.70 years equal to an amount of 4 , 70%. Between 2015 and 2019, the value of life expectancy in good health at birth in the North grew by 0.50 years, equal to an amount of 0.84%. Between 2009 and 2019, life expectancy in good health at birth grew by an amount equal to 0.27 years per year, while between 2005 and 2019 this value grew by 0.14 years per year. In Central Italy, the value of life expectancy in good health at birth increased from an amount of 56.9 years in 2009 to a value of 59.1 years in 2019 or equal to an amount of 2.20 years, equal to at 3.87% and an average growth of 0.22 years. Between 2015 and 2019 the value of life expectancy in good health at birth increased from an amount of 58.8 years to a value of 59.1 years or equal to a variation of 0.30 years equal to a value of 0.51%. However, between 2015 and 2019 the value of the average absolute changes in life expectancy in good health at birth decreased by an amount equal to -0.04 units. In the South of Italy between 2009 and 2019 the value of life expectancy in good health at birth increased from 54.5 years to a value of 56.2 years or a variation of 1.70 years equal to an amount of 3, 12%. Between 2015 and 2019, the value of life expectancy at birth in good health increased from an amount of 56 years to a value of 56.2 years or equal to a variation of 0.20 years equal to a value of 0.36%. On average in Italy between 2009 and 2019 the value of life expectancy in good health at birth increased from an amount of 56.4 years up to a value of 58.6 years in 2019 or a variation of 2.20 years equal to an amount of 3.90%. Between 2015 and 2019, the value of life expectancy in good health at birth increased from an amount of 58.3 years to a value of 58.6 years in 2019 or a variation of 0.30 years equal to a variation of 0.51%. There are therefore two phenomena to consider in the Italian macro-regions: on the one hand there is a growth in life expectancy in good health at birth in all Italian geographic areas, on the other hand there is a growing gap between the North and Center-South. In fact, considering the difference between the healthy life expectancy at birth in the North and in the South, it appears that: in 2009 the southerners had a healthy life expectancy at birth of 2.9 years lower than in the North, a value which worsened to 3.6 years in 2015 and to 3.9 years in 2019. It therefore follows that the healthy life expectancy of the southerners improved in historical series even if it worsened in comparison with Northern Italy.

Clusterization. An analysis of clustering is proposed below using the k-Means algorithm used to identify the presence of groups among the Italian regions. Since the k-Means algorithm is an unsupervised algorithm, it is necessary to identify a criterion to choose the optimal number of clusters. In the specific case, the Silhouette coefficient was used. The Silhouette coefficient ranges from -1 to 1. The optimal number of clusters corresponds to the highest Silhouette coefficient. In the case analyzed, the higher Silhouette coefficient occurred in the presence of several clusters equal to 3. The following clustering therefore derives:

•         Cluster 1: Aosta Valley, Lombardy, Veneto, Friuli-Venezia Giulia, Tuscany, Emilia-Romagna, Liguria, Piedmont, Marche, Umbria, Lazio, Abruzzo, Molise;

•         Cluster 2: Trentino-Alto Adige;

•         Cluster 3: Sardinia, Basilicata, Calabria, Sicily, Puglia.

In terms of the median value of the clusters it is therefore possible to identify the following order, i.e. C2 with a median value of life expectancy in good health at birth equal to an amount of 65.8, followed by C1 with a value of 59.6 and from cluster 3 with an amount of 55.15 units. Therefore, the following ordering derives from it, namely C2> C1> C3. Analyzing the geographical map of clustering, the central-northern regions have a significantly better value of life expectancy in good health at birth than the southern regions. Among all the regions, Trentino-Alto Adige stands out as it has healthy life expectancy levels at birth that far exceed those of the southern regions. It therefore follows that the levels of human and social capital probably also have a significant impact in determining the quality-of-life levels of the population from the health point of view. In fact, Trentino-Alto Adige is not only one of the regions with the highest per capita income in Italy but also one of the regions with the greatest culture of cooperation and social cohesion.

Network Analysis. A network analysis was carried out below. The analysis is carried out through the use of the Manhattan distance. The following elements are detected:

• •         Number of nodes equal to 20;
• •         Number of edges 24;
• •         Average degree 2.4;
• •         Density 0.1263.

Specifically, there are two significant links within the analysis carried out, namely:

• •         Piedmont-Veneto with a value of 2.9;
• •         Campania-Sicily with a value of 1.6.

These links indicate a significant correlation between the indicated regions. These relationships may indicate a convergence within the economic, social, institutional and behavioral dynamics that have an impact in determining healthy life expectancy at birth.

Machine Learning and Predictions. Below eight different machine learning algorithms are used to predict the future value of healthy life expectancy at birth in the Italian regions. The algorithms were trained with 70% of the available dataset while the remaining 30% of the dataset was used for prediction. The performance analysis was carried out considering the algorithms' ability to maximize the R-square and minimize the main statistical errors, namely "Mean Absolute Error", "Mean Squared Error", "Root Mean Squared Error". The following ordering of the algorithms is therefore derived, that is:

• •         Random Forest Regression with a payoff value of 4;
• •         Tree Ensemble Regression with a payoff value of 9;
• •         Artificial Neural Network-ANN with a payoff value of 14;
• •         PNN-Probabilistic Neural Network with a payoff value of 15;
• •         Simple Regression Tree with a payoff value of 18;
• •         Linear Regression with a payoff value of 26;
• •         Gradient Boosted Tree with a payoff value of 26;
• •         Polynomial Regression with a payoff value of 32.

Therefore, using the best performer algorithm or the Random Forest Regression it was possible to make the following predictions, namely:

• •         Liguria with an increase from an amount of 59.20 units up to a value of 50.43 units or equal to a variation of 1.23 units equal to 2.08%;
• •         Veneto with a decrease from an amount of 60.30 units up to a value of 59.57 units or equal to a variation of -0.73 units equal to a variation of -1.21%;
• •         Emilia Romagna with an increase from an amount of 59.60 units up to a value of 60.15 units or equal to a variation of 0.55 units equal to an amount of 0.93%;
• •         Abruzzo with an increase from an amount of 57.30 units up to a value of 57.72 units or equal to a variation of 0.42 units equal to an amount of 0.73%;
• •         Calabria with an increase from an amount of 49.70 units up to a value of 55.16 units or equal to a change of 5.46 units equal to a change of 10.99%;
• •         Sicily with an increase from an amount of 55.90 units up to a value of 57.09 units or equal to a change of 1.19 units equal to a value of 2.12%;

On average, the value of life expectancy in good health at birth is predicted to grow in the regions considered from an average value of 57.00 units up to a value of 58.35 units or equal to a value of 1.35 units equal at a value of 2.37%.

Conclusions. In summary, it is possible to verify that the value of life expectancy in good health at birth increased in the period between 2009 and 2019 in all Italian regions with the exception of Basilicata. However, it should be borne in mind that the growth rate of life expectancy in good health at birth in the regions of Northern Italy was equal to a value of 4.7% while in central Italy it was equal to an amount of 3.8% and in the South equal to a value of 3.11% between 2009 and 2019. It follows that on average in 2019 a citizen of the North had a healthy life expectancy at birth of about 3 years higher than a citizen of the South. In particular, the most virtuous regions in terms of life expectancy in good health are the regions of the North-East with Trentino Alto Adige in the lead. Since these regions are also the richest in terms of per capita income and human and social capital it follows that inevitably there is a positive connection between the value of healthy life expectancy at birth and the capacity to produce added value and public and social goods. Furthermore, it must be considered that increasing even marginally the life expectancy in good health at birth means saving a lot in terms of regional public spending and therefore also having more sustainable regional health systems from a financial point of view. Economic policies to improve life expectancy at birth are therefore complex and require a series of composite interventions ranging from the growth of added value to the improvement of the production capacity of public and social goods through the appreciation of human capital. Finally, it must be considered that these data are pre-covid. Since Covid has significantly affected Northern Italy, especially Lombardy, it is likely to have led to a significant reduction in life expectancy in good health, especially in the northern regions.

 

References

Laureti, L., Costantiello, A., Leogrande, A. Satisfaction with the Environmental Condition in the Italian Regions between 2004 and 2020. University Library of Munich, Germany, 2022.

Leogrande, A., Costantiello, A., Laureti, L., & Leogrande, D. (2021). The Determinants of Landscape and Cultural Heritage among Italian Regions in the Period 2004-2019. Available at SSRN 3971174.

Leogrande, A., A. Saponaro, A. Massaro, and A. M. Galiano. A GIS based estimation of quality of life in Italian regions.

Leogrande, A., Birardi, G., Massaro, A., & Galiano, A. M. (2019). Italian Universities: Institutional Mandate and Communitarian Engagement. European Journal of Educational Management, 2(2), 85-110.

Magazzino, C., & Leogrande, A. (2021). Subjective Well-Being In Italian Regions: A Panel Data Approach. Applied Econometrics and International Development, 21(1), 1-18.