dorsal/arxiv
View SchemaInnovation flow through social networks: Productivity distribution
| Authors | T. Di Matteo, T. Aste, M. Gallegati |
|---|---|
| Categories | |
| ArXiv ID | physics/0406091 |
| URL | https://arxiv.org/abs/physics/0406091 |
| DOI | 10.1140/epjb/e2005-00332-y |
| Journal | The European Physical Journal B 47 (2005) 459-466 |
Abstract
A detailed empirical analysis of the productivity of non financial firms across several countries and years shows that productivity follows a non-Gaussian distribution with power law tails. We demonstrate that these empirical findings can be interpreted as consequence of a mechanism of exchanges in a social network where firms improve their productivity by direct innovation or/and by imitation of other firm's technological and organizational solutions. The type of network-connectivity determines how fast and how efficiently information can diffuse and how quickly innovation will permeate or behaviors will be imitated. From a model for innovation flow through a complex network we obtain that the expectation values of the productivity level are proportional to the connectivity of the network of links between firms. The comparison with the empirical distributions reveals that such a network must be of a scale-free type with a power-law degree distribution in the large connectivity range.
{
"annotation_id": "1c623953-faa3-42bb-aa35-ab2971b5bfd8",
"date_created": "2026-03-02T18:00:50.273000Z",
"date_modified": "2026-03-02T18:00:50.273000Z",
"file_hash": "5f9e34e494189f22c268aa5468ed0d7c5305fde4600699bc791f8c675756295f",
"private": false,
"record": {
"abstract": "A detailed empirical analysis of the productivity of non financial firms\nacross several countries and years shows that productivity follows a\nnon-Gaussian distribution with power law tails. We demonstrate that these\nempirical findings can be interpreted as consequence of a mechanism of\nexchanges in a social network where firms improve their productivity by direct\ninnovation or/and by imitation of other firm\u0027s technological and organizational\nsolutions. The type of network-connectivity determines how fast and how\nefficiently information can diffuse and how quickly innovation will permeate or\nbehaviors will be imitated. From a model for innovation flow through a complex\nnetwork we obtain that the expectation values of the productivity level are\nproportional to the connectivity of the network of links between firms. The\ncomparison with the empirical distributions reveals that such a network must be\nof a scale-free type with a power-law degree distribution in the large\nconnectivity range.",
"arxiv_id": "physics/0406091",
"authors": [
"T. Di Matteo",
"T. Aste",
"M. Gallegati"
],
"categories": [
"physics.soc-ph",
"cond-mat.stat-mech",
"physics.gen-ph"
],
"doi": "10.1140/epjb/e2005-00332-y",
"journal_ref": "The European Physical Journal B 47 (2005) 459-466",
"title": "Innovation flow through social networks: Productivity distribution",
"url": "https://arxiv.org/abs/physics/0406091"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "3e47f123-b4af-42ad-acae-aec876b599d8",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}