Italian Agrifood: lots of opinions and little certainty in a sea of data
The questions
Information sources on the Italian agriculture industry abound, overlap, and rarely share definitions, methodologies or units of measurement. What’s more, these sources often produce data that are difficult to interpret and compare. In Italy, farms are small, specific skills are lacking, and accounting rules for the industry are complicated, mainly based on bureaucratic automatisms. All this has stunted the evolution of a culture of data collection among people who work in agriculture. Farmers, for their part, frequently find themselves unprepared to answer highly detailed questions they’re asked in sector studies, so they resort to estimates – or guesstimates. This only adds to the variability and subtracts from reliability of the information.
As an example, we can refer to the fact that the number of farms operating in Italy ranges from 718,000 to 1,500,000 based on how the word ‘farm’ is defined. And Annual Work Units (AWU) deviate by 250,000 depending on who is doing the counting. Added to this, some data are collected annually, others every three years (though the latter are equally important and may be linked to the former). The risk is that the information collected by the bodies tasked with the job end up watering down the action of the government and entrepreneurs, who could benefit from more coherent and standardized data on economic, structural and contextual aspects of the agriculture industry. Such data could help them face the growing pressure driving demand for land and food, economic uncertainty and threats linked to the climate emergency.
Fieldwork
Among all most strategic sectors for contending with global challenges, today and in the coming years, agriculture is without a doubt the most complex. In light of this, the industry warrants precise monitoring that reflects all its constituent elements. In fact, public institutions and private players rely on data to understand what’s happening in agriculture by applying objective and generally accepted parameters; these data also serve to allow them to map out and evaluate the lines of action they take.
In a study by AGRI Lab at SDA Bocconi School of Management entitled, “The State of Italian Agriculture and Sustainability Criteria,” Antonio Tencati and Manlio De Silvio shine a light on the more glaring discrepancies in existing data on the industry. In the Italian context, in fact, there are 48 national bodies and 9 international organizations (public bodies, foundations, associations, and private enterprises), all responsible for collecting data on the agriculture industry.
ISTAT (Italian National Statistics Institute) provides the industry’s primary database. To name just the main studies produced by the Institute: a 10-year census (which since 2020 is conducted annually); national analyses on economic results, production and organization of farms; an annual census survey on the characteristics of PDO, PGI and TSG producers; an estimate of the production and land, and one on the profit & loss of agriculture. Along with ISTAT, other public bodies also monitor the industry. CREA (Council for Research in Agriculture and Analysis of the Agricultural Economy) under the supervision of the Ministry of Agriculture, Food and Forest Policies, manages an information system centering on economic, structural and technical aspects; a quarterly statistical analysis on new farms and farm closures is run by Unioncamere (Italian Union of Chambers of Commerce), or more specifically with regard to data collection, Infocamere (a joint-stock consortium company of the Italian Chambers of Commerce) ISMEA (Institute of Services for the Agrifood Market), publishes periodic surveys on structural and market variables in the agriculture and food industries and coordinates permanent observatories on entrepreneurship among women and young people. Supporting public bodies are numerous third-party organizations (Confagricoltura, Coldiretti, the Edison Foundation, the Qualivita Foundation, the Symbola Foundation and many more) which focus on specific issues such as employment, certified productions or food products that are national specialties.
Each source takes a deep dive into one or more topics, adopting the most appropriate methodologies to suit their purposes. But diverse approaches as far as data collection and processing inevitably lead to appreciable variability in the resulting information. In some cases, in fact, the gaps are so substantial we might think that the monitoring bodies are benchmarking to divergent definitions in designing their analyses, giving us an information output that paints a confusing picture.
The original intent of data collection is the cornerstone on which data collection methods and units of measurement are built. As an example, let’s take the distribution of farms in terms of economic output. 9Because ISTAT and CREA adopt different segmentation thresholds, they come back with outlooks that are not comparable. The variances that emerge can be limited, such as the number of farms with standard production above 500,000 euro, which according to the two data sources differs by little more than 420 units, or more sizeable, up to 63,000 units in the tally of farms with production ranging from 50,000 to 99,999 euro.
The unit of measurement and the frequency of data collection play an equally critical role, as employment data illustrate. The sources in question alternately cite people, workers, AWUs or days, and run some studies annually, and others every three years or more. Confagricoltura (the national organization of farmers) compares AWUs of the main agricultural countries in the European Union, while ISTAT’s “survey on the workforce,” in contrast, serves to track the distribution of workers by age and professional position in Italian industries. Even ISTAT selects different units of measurement for its different studies. For instance, assuming we can consider ‘workers’ and ‘people’ one and the same, and we compare the respective numbers, we find divergent data: ISTAT studies from 2013 talk about 799,000 workers but 3,559,081 people. And when we compare data catalogued with the same label for the basic unit of measurement, we can find dissimilar calculation methods, especially from source to source. For example, matching ISTAT’s REA study on the economic performance of farms against the analyses by Confagricoltura on AWUs, we come up with a divergence of more than 250,000 units.
With regard to more specific topics, fewer available sources minimize uncertainty in identifying salient information, and implies less methodological variety as well. This is the case on issues such as entrepreneurship among women and young people, which is effectively tracked by ISMEA. More standardized data are generated also when there are official, generally accepted definitions, such as criteria for labelling products with PDO, PGI and TSG certification.
Moving forward
So we need to know what the aims and the methodologies underpinning the main sources of information in order to identify the data that are useful for our purposes. Although not all databanks offer the same width and depth, using them as supplementary sources is currently the most effective way to get a snapshot of the dynamics of a multifaceted industry such as agriculture, characterized by complex value chains and a great variety of productions.
Despite the fact that the landscape is already vast, there are still many dynamics that warrant more in-depth monitoring: legality in production chains; price trends and competitiveness of productions; the adoption of contract models for supply chains and the distribution of value along production chains. In addition, as widely discussed in the media, as of now little investigation has been done into the use of advanced farming practices in the field of cultivation and soil management, such as precision agriculture and mechanization, along with everything relating to the link between the agricultural industry and the climate emergency.
By coordinating information sources, we could help expand the perimeter of study to understand these issues. What’s more, to shore up data collection processes and the reliability of the resulting information, it is still indispensable to support the people who work in the industry, building a culture of data. This is based on competencies, technologies and tools for tracking data that are simple, basic, and suited to the reality of farming.
In light of this consideration, AGRI Lab at SDA Bocconi has recently developed the CoCoA Business Simulation, a management control system software that can support farms in collecting data and measuring the profitability of various activities.
