Network diversity index (NTd)

• Greater connectivity and access to varied resources.
• Enhanced collaboration potential, as multiple perspectives are represented.
• Stronger resilience, reducing dependency on a single category of actors.

By quantifying diversity, the NDI helps practitioners assess whether the network structure supports inclusive decision-making and innovation – key elements for effective local development strategies.

The NTd is a form of Gini’s concentration index. It can be calculated based on the following formula:

where:

• N represents the total number of categories of interests represented in the network;
• p_i is the proportion of the first i categories to the total number of categories; and
• q_i is the proportion of actors belonging to the first i categories to the total number of actors.

The calculation of the index requires the number of actors in the network broken down by the category of interests they represent. The more detailed the categorisation, the more accurate the estimation of the index.

Example – Consider the case of a LAG (LEADER), the categories of interests could be members of the organisation that represent:

1. public administrators
2. private local economic interests
3. social local interests
4. other types 

So, in this simple example, there are four categories of interests, while the ‘number of actors’ is the number of LAG members belonging to each category of interests. The total number of members of the LAG equals the sum of members across all categories of interest. 

Its main characteristics are:

• It makes use of data that are usually readily available or easy to collect.
• The index varies in the range of 0 to 1, assuming the value 0 (no diversity) when there is only one category of actors in the network, and the value 1 (maximum diversity) when all the categories are represented by an equal number of actors.
• It can be applied fairly rapidly and can provide insights about the levels of structural social capital in a network and their changes over time, if it is computed in different moments of the programming period to check for changes. 

• Access to as detailed a composition of the network as possible.
• Additional data, such as the population served by the network, might be useful to better contextualise the calculated indices.

• Farmers, foresters and their associations (other than young farmers and foresters)
  Including both natural and legal persons.
• Young farmers and foresters (below 40 years old) and their associations
• Micro, small and medium enterprises in the processing of agricultural products, e.g. cheese makers, oil presses, etc. Includes also their associations.
• Other micro, small and medium-sized enterprises
  Business organisations outside the primary sector, including their associations. Primary sector SMEs are included in the first category. Farmers, foresters and their associations.
• Large enterprises in the processing of agricultural products
  Large enterprises employ more than 250 employees and/or have more than EUR 50 million annual turnover.
• Other large enterprises
  Business organisations outside the primary sector. Primary sector large enterprises are included in the first category. Farmers, foresters and their associations.
• Public enterprises
  Business organisations wholly or partly owned by the state and controlled through a public authority.
• Professional organisations or associations (e.g. an association of tourist accommodations)
• Trade unions
• Local women associations or organisations
• Local youth associations or organisations
• Other local associations
• Social NGOs (except trade unions and local associations)
• Environmental NGOs
• Protected areas governance bodies
  Governance bodies for NATURA 2000, national parks or reserves and any other protected area.
• Local public authorities (e.g. municipalities and their associations, regional authorities)
  Enterprises wholly or partly owned and controlled by local authorities must be counted only under the public enterprises above.µ

It is important to be very clear regarding the different actors that can be classified under a certain category. Providing definitions (as shown above for ‘Public enterprises’) and preparing examples of how certain actors can be categorised (as shown above for ‘Professional organisations or associations’) can contribute to more accurate categorisation and improved quality of the collected data. 

Step 2 – Collect data 

Organise a survey to collect the number of actors under each category. Consider collecting additional data that may help contextualise the calculated indices. 

Data can be collected for different levels in a network: members of LAGs, the decision-making body, members of networks developed/maintained by LAGs and members of the National Network, etc. 

Step 3 – Sort the categories by the number of the corresponding actors  

Sort the categories in ascending order by the number of actors that belong to each category. 

Step 4 – Calculate the proportion of the first i categories to the total number of categories (p_i)

Calculate the cumulative sum of the number categories. Divide each sum by the total number of categories. See the example in the following table.

Categories	Category 1 (e.g. municipality)	Category 2	Category 3	Category 4	Total
Number of actors	12	14	24	30	80
Cumulative sum of the number of categories	1	2	3	4
pi	1/4 = 0.25	2/4 = 0.5	3/4 = 0.75	4/4 = 1	= 1.5

Step 5 – Calculate the proportion of actors belonging to the first ‘i’ categories to the total number of actors 

Calculate the cumulative sum of the number of actors. Divide each sum by the total number of actors. See the example in the following table.

Step 6 – Calculate the NTd

Calculate the sum of differences between p_i and q_i for each category. Use them to calculate the NTd. See the example in the table below.

Categories	Category 1	Category 2	Category 3	Category 4	Total
pi	1/4 = 0.25	2/4 = 0.5	3/4 = 0.75	4/4 = 1	= 1.5
qi	12/80 = 0.15	26/80 = 0.325	50/80 = 0.625	80/80 = 1
pi - qi	0.25 – 0.15 = 0.1	0.5 – 0.325 = 0.175	0.75 – 0.625 = 0.125	1 – 1 = 0	= 0.4
NTd					1 – (0.4 / 1.5) = 0.733

Step 7 – Interpret the score of the NTd

The NTd measures how evenly different categories of actors are represented within a network.

How to interpret the score:

• Closer to 0 → Low diversity: the network is dominated by one or very few categories.
• Closer to 1 → High diversity: actors are well distributed across all categories.

In our case, the NTd score is 0.733, which suggests high diversity. This means the network includes actors from multiple categories, with a relatively balanced distribution. This diversity often indicates:

• Broader perspectives and resources within the network.
• Greater potential for collaboration and innovation.
• Reduced risk of groupthink or over-reliance on a single category.

While a higher score generally reflects a healthier, more inclusive network, consider the context:

• Are all relevant categories represented?
• Does the distribution align with the network’s goals?

• The NTd is a relatively easy to calculate metric that can provide insights about the levels of structural social capital of a network.
• It can be used in the assessment of LEADER added value, as well as National Network, AKIS, and smart village strategies.
• It can be based on a participatory effort to define the categories of actors together with the network managers or support units, contributing to them assuming ownership of the process.

• Nardone, G., Sisto R., Lopolito A. (2010), Social Capital in the LEADER Initiative: a methodological approach. Journal of Rural Studies, 26, pp. 63-72.

Evaluation support study of the costs and benefits of the implementation of LEADER

In this work, the NTd is used as a quantitative measure to assess the structural social capital within LAGs and their broader networks under the LEADER program. Specifically, NTd has been calculated at different levels, such as within the LAGs’ boards of directors, general assemblies, project promoters and cooperation projects. The NTd was also used to compare LAGs with other rural development initiatives and to analyse the impact of inter-territorial and transnational cooperation projects, providing a nuanced understanding of how network diversity contributes to the added value and effectiveness of the LEADER approach.

For more information, see: 

• Evaluation support study of the costs and benefits of the implementation of LEADER – Final report (2023)