The system generates metric models regression, which is a dynamic linear panel model. That can account for the interplay between broadband coverage and stimulation of economic growth and account for measurement error. It further accounts for other problems that are usually well known in these growth regressions data if real GDP per capita and purchasing power parities are used. Some of the control variables include past economic growth, population growth and proxies for investments and trade orders. These data are proxied by using information on formal sector employments, usually.
Modelling impacts of digitalisation can be very useful for assessing the effects of interventions that aim to increase the digitalisation of rural areas, in particular, the effects on economic growth. The contextual background offered by DESIRA indicated the lack of data at the NUTS 3 level for measuring indicators related to human capital, use of internet, integration of digital technology or digital public services. However, it has shown that it is possible to analyse broadband access at the NUTS 3 level. This opens a new door to impact assessment at this territorial level, where many of these interventions are implemented.
Furthermore, the fixed broadband coverage at different speeds measured at the NUTS 3 level is important for monitoring and evaluation because of the Farm to Fork objective stating that ‘The Commission aims to accelerate the roll-out of fast broadband internet in rural areas to achieve the objective of 100% access by 2025’. The RDI (which includes DESI indicators on fixed broadband coverage) may be a good monitoring and reporting indicator of progress.
Knowledge of econometrics to undertand and use the 'system generalised method of moments estimator' is required to model broadband access.