ALPHAMAPS predicts the future with open data

Four models reveal where most current trends are headed

Most macro level open data sets reveal linear trends – relatively stable trends of growth, decline, or stabilisation – on a medium-term time frame. These trends can be extrapolated with a simple linear regression algorithm.

For example, Eurostat’s organic farming data set shows that the share of EU agricultural land under organic farming has been growing at an almost constant rate (of around 6.2% per year) for the past decade.

Whereas the number of persons active in agriculture, as a share of total active EU population, has been declining at an almost constant rate (of around 2.3% per year) during the same period.

A few data sets follow exponential or logarithmic trends – trends that accelerate or decelerate over time. These trends can also be extrapolated with exponential or logarithmic algorithms.

For example, the spread of broadband internet across the EU started with a bang in the mid-2000s and has decelerated ever since (notwithstanding continued increases in connection speed).

Finally, certain data sets, especially economic ones, are subject to the “rhymes of history” – a reference to the famous adage “History never repeats itself, but it rhymes.”, attributed to Mark Twain.

For example, when a recession hits, household savings as a share of GDP go up. Historical trend “shocks” can be replicated to forecast the future trend of an ongoing shock and adapted to reflect the apparent differences between the two shocks (such as speed and intensity). In this example, the economic impact of Covid19 has been approximated to an accelerated version of the Great Recession, replicating the changes in the data between 2007 and 2013 to the years 2020, 2021 and 2022.

Published on: 15 June 2021

This site makes use of essential cookies. By continuing to navigate this site, you agree to our Cookie Policy and Terms of Service. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.Accept & continue