The Determinants of Landscape and Cultural Heritage Among Italian Regions in the Period 2004-2019

We estimate the Landscape and Cultural Heritage among Italian regions in the period 2004-2019 using data from ISTAT-BES. We use Panel Data with Fixed Effects, Panel Data with Random Effects, Pooled OLS, WLS, Dynamic Panel. We found that the Landscape and Cultural Heritage is negatively associated with “ Dissatisfaction with the landscape of the place of life ”, “ Illegal building ”, “ Density and relevance of the museum heritage ”, “ Internal material consumption ”, “ Erosion of the rural space due to abandonment ”, “ Availability of urban green ”, and positively associated with “ Pressure from mining activities ”, “ Erosion of the rural space by urban dispersion ”, “ Concern about the deterioration of the landscape”, “ Diffusion of agritourism farms ”, “ Current expenditure of the Municipalities for culture”. Secondly, we have realized a cluster analysis with the k-Means algorithm optimized with the Silhouette Coefficient and we found two clusters in the sense of “ Concern about the deterioration of the landscape ”. Finally, we use eight different machine learning algorithms to predict the level of “ Concern about the deterioration of the landscape ” and we found that the Tree Ensemble Regression is the best predictor.


Introduction
In this article we have analyzed the determinant of Landscape and Cultural Heritage among Italian regions in the period 2004-2019. We use data from ISTAT-BES. The role of Landscape and Cultural Heritage has acquired a growing interest among population and policy makers as a result either of the actions against climate change either of a re-discovery of ethno-linguistics and monumental traditions also at a local level. (Della Spina, 2017) afford the question of the relevance of a multi-methodological and multistakeholder approach in the context of Historical Urban Landscape with an active participation of communities and local experts in preserving either landscape either historical heritage. (Gravagnuolo & Girard, 2017) propose a series of metrics to evaluate the level of Historical Urban Landscape. (Bulian, 2021) considers the role of multi-locality and the interdependence of cultural heritage and landscape in Japan. (Rouhi, 2017) affords the question of the anthropological meaning of cultural heritage in the context of universal values. (Vallerani & Visentin, 2018) show the relevance of waterway as a cultural and socio-economic tool for civilization and landscape valuation. (Cicinelli, Salerno, & Caneva, 2018) apply a multidisciplinary approach to the preservation of Medieval Benedictine Monastery of San Vincenzo al Volturno considering natural, archeological, and agricultural elements. (Ravankhah, Schmidt, & Will, 2021) develop a methodology to evaluate the environmental risk for cultural heritage size such as in the case of earthquakes. The authors propose the "Cultural Heritage Risk Index" with an application for the World Heritage site of Bam in Iran. (Antonson, Buckland, & Nyqvist, 2021) analyze the question of the relationship among cultural heritage, climate change and active governmental policies in Sweden. (Assandri, Bogliani, & Pedrini, 2018) show the relevance of agriculture in creating the conditions to preserve landscape and biodiversity with an application to wine production in Italian region Trentino-Alto Adige. (Cai, Fang, Zhang, & Chen, 2012) consider the application of a connection between digital technologies and the cultural heritage protection in the case of Mount Lushan in China. To preserve Mount Lushan in China from the negative externalities of massive tourism the authors have promoted a model of virtual tourism based on 3D laser scanning, oblique aerial photography and 360 degrees panorama technology. (Foster, 2020) suggest new methodologies based on circular economy to cultural heritage buildings in the context of environmental sustainability. (Shirvani Dastgerdi, Sargolini, Broussard Allred, Chatrchyan, & De Luca, 2020) consider the role of climate change in worsening the condition of central Italy regions in terms of rainfall patterns suggesting a deeper coordination between the European Landscape Convention and local and territorial planning to pursue the objective of conservation. (Guzman, Fatorić, & Ishizawa, 2020) propose some methodology to mitigate the risk of climate change for World Heritage-WH sites at a global level suggesting a multidisciplinary and multistakeholder approach to promote resilience. (Li, Krishnamurthy, Roders, & Van Wesemael, 2020) consider the role of community participation in heritage management considering a peoplecentered approach with an application in China. The article continues as follows: the second paragraph presents the econometric model, the third paragraph present the cluster analysis, the fourth paragraph contains the machine learning and prediction algorithms, the fifth paragraph concludes.

The Econometric Model
We estimate the following econometric model: We found that the "Landscape And Cultural Heritage" is positively associate with:  Current Expenditure of the Municipalities for Culture: is defined as "payments in accountability for the protection and enhancement of cultural assets and activities, in euros per capita". There is a positive relationship between the Landscape and Cultural Heritage and the level of the "Current Expenditure of the Municipalities for Culture" i.e. the greater the local expenditures in cultural event the greater the level of Landscape and Cultural Heritage. This means that there is an effective role that policy makers, even at the local level, can play in promoting a culture more oriented toward landscape preservation.  Diffusion of Agritourism Farms: is defined as the number of farms per 100 km2. There is a positive relationship between "Diffusion of Agritourism Farms" and the level of "Landscape and Cultural Heritage". The presence of agritourism can improve the culture of respect for environmental public goods and public artistic goods among the population creating the conditions even for collective actions in defence of green and cultural commons.  Concern About the Deterioration of the Landscape: is defined as the percentage of people aged 14 and over who indicate the ruin of the landscape caused by excessive building construction among the five problems environmental issues more worrying than the total number of people aged 14 and over. There is a positive relationship between "Concern About the Deterioration of the Landscape" and "Landscape and Cultural Heritage" suggesting that if the population is emotionally engaged in the deterioration of the landscape than the condition of environmental and artistic goods can improve.  Erosion of the Rural Space by Urban Dispersion: is defined as "Percentage incidence of the agricultural regions concerned from the phenomenon on the total of the regional surface". There is a positive relationship between Landscape and Cultural Heritage and the Erosion of the Rural Space by Urban Dispersion.  Pressure from Mining Activities: is defined as the Volume of resources non-energy minerals extracted (cubic meters) per km2. There is a positive relationship between "Pressure from Mining Activities" and "Landscape and Cultural Heritage". We also find that the level of "Landscape And Cultural Heritage" is negatively associated with:  Availability of Urban Green: is defined as "Square meters of urban green space per inhabitant". The greater the "Availability of Urban Green" the lower the level of "Landscape and Cultural Heritage".  Erosion of the Rural Space due to Abandonment is defined as "Incidence percentage of agricultural regions concerned from the phenomenon on the total of the regional surface". There is a negative relationship between the Erosion of the Rural Space due to Abandonment and the level of Landscape and Cultural Heritage. The greater the level of the "Erosion of the Rural Space due to Abandonment" the lower the level of "Landscape and Cultural Heritage".  Internal Material Consumption: is defined as the "Quantity of materials processed into emissions, waste or new stocks of the system anthropic (in millions of tons)". There is a negative relationship between "Internal Material Consumption" and the level of "Landscape and Cultural Heritage". The greater the level of Internal Material Consumption the lower the level of Landscape and Cultural Heritage.  Density and Relevance of the Museum Heritage: is defined as the "Number of permanent exhibition structures for 100 km2 (museums, archaeological areas and monuments open to audience), weighted by the number of visitors". There is a negative relationship between "Density and Relevance of the Museum Heritage" and the level of "Landscape and Cultural Heritage". The greater the level of "Density and Relevance of the Museum Heritage" the lower the level of "Landscape and Cultural Heritage". This result can appear counterfactual. But it can be better understood considering that many southern Italian regions that have a widespread diffusion of museum also have low levels of attention for the landscape and cultural heritage due to the lack of social and human capital.

Clusterization
We have applied the k-Means algorithm optimized with the Silhouette Coefficient to investigate the presence of clusters in the sense of "Concern about the deterioration of the landscape". We found two clusters as follows: Specifically, the median value of regions in the Cluster 1 is equal to 13,55, while the median value of regions in the Cluster 2 is equal to 10,6. As we can see there is a great divide between Southern and Northern Italy in the sense of "Concern about the deterioration of the landscape". Northern Italian regions that have generally greater Gdp Per Capita and greater human and social capital also show a greater concern about the deterioration of the landscape.

Machine Learning and Predictions
We have estimated the level of "Concern about the deterioration of the landscape" using eight different machine learning algorithms to predict the future value of the observed values. We use 70% of the dataset as learning rate and the remaining 30% for the prediction. We have ranked the eight different algorithms based on their ability to minimize statistical errors such as "Mean Absolute Error", "Mean Squared Error", "Root Mean Squared Error", "Mean Signed Difference". The order of algorithms based on their ability to minimize errors is as follows:  Using the Tree Ensemble Regression we have predicted the following percentage variation of the variable "Concern about the deterioration of the landscape" :  Piemonte +13,92%;  Lombardia -2,13%;  Veneto +27,59%;  Abruzzo +87,67%;  Molise 222,58%;  Puglia +209,57%.
Finally the mean value of the prediction is equal to 93,20% in the observed regions.

Conclusions
We estimate the Landscape and Cultural Heritage Index among Italian regions in the period 2004-2019 using data from ISTAT-BES. We use Panel Data with Fixed Effects, Panel Data with Random Effects, Pooled OLS, WLS, Dynamic Panel. We present a brief literature review considering the role of communitarian and environmental issues on landscape and cultural heritage preservation. Climate change is one of the main threats for landscape and cultural heritage but also socio-political issues, such as communitarian engagement, have a relevant role in promoting tools for conservation. In the second paragraph we estimate the value of Landscape and Cultural Heritage and we find that it is negatively associated with "Dissatisfaction with the landscape of the place of life", "Illegal building", "Density and relevance of the museum heritage", "Internal material consumption", "Erosion of the rural space due to abandonment", "Availability of urban green", and positively associated with "Pressure from mining activities", "Erosion of the rural space by urban dispersion", "Concern about the deterioration of the landscape", "Diffusion of agritourism farms", "Current expenditure of the Municipalities for culture". In the third paragraph we propose the application of a cluster analysis with the k-Means algorithm optimized with the Silhouette Coefficient and we found two clusters in the sense of "Concern about the deterioration of the landscape". We found that the Italian regions are essentially divided in two main parts: the Northern Italy with high levels of concern and the Southern Italy with lower level of concern. This contraposition suggests that the economic divide in terms of Gdp per capita and social and human capital operates as a determinant for the better performance of Northern Italian regions in respect to Southern Italian regions. Finally, we use eight different machine learning algorithms to predict the level of "Concern about the deterioration of the landscape" and we found that the Tree Ensemble Regression is the best predictor. Furthermore, the predicted values suggest that the level of "Concern about the deterioration of the landscape" is expected to growth significantly.
Our analysis suggests that to promote conservation of landscape and cultural heritage is essential either to invest in cultural expenditure at a local level either to promote economic activities that are more oriented to environmental sustainability such as in the case of agritourism farms. Furthermore, the lack of human and social capital significantly reduces the possibility of southern Italian regions to effectively protect their landscape and cultural heritage. If policy makers are oriented to promote landscape and cultural heritage conservation they should invest more in human and social capital, promoting a deeper consciousness of preservation of cultural and environmental common goods among the communities.  Varianza 'between' = 0 Varianza 'within' = 2,12113 theta medio = 0 Test congiunto sui regressori -Statistica test asintotica: Chi-quadro(11) = 16551,5 con p-value = 0