Milan active-travel transformations (2016–2021): Interactive period maps from processed municipal data

Overview

This analysis documents active-travel transformations in Milan (2016–2021) using official municipal datasets from the Comune di Milano and a custom dataset of Piazze Aperte created for this study.

Data sources

  • Official municipal data (Comune di Milano)
    • Cycling infrastructure (lines): ciclabile sede propria, ciclabile segnaletica
    • Pedestrian-priority infrastructure (polygons): DT (divieto di transito), AP (area pedonale)
  • Custom dataset (this study)
    • Piazze Aperte (polygons): manually digitised public-space interventions
      Only interventions flagged for analysis were retained (31 features), and classified as:
      • Permanent
      • Tactical

Data processing

Completion of missing periods using GSV (pedestrian-priority areas)

For pedestrian-priority polygons, implementation timing was first extracted from the official field val_inizio when available.
Some features had missing or unusable dates.

These cases were reviewed manually using Google Street View (GSV), and an approximate implementation period was assigned based on visual evidence.
The resulting period labels were then re-imported and joined back to the spatial dataset.

For features where a year was available but no manual period was assigned, period attribution followed the year-based rules described below.

Period attribution (alignment with election cycles)

To enable a consistent temporal comparison linked to political mandates in Milan, all interventions were classified into broad implementation periods aligned with municipal election cycles.

The relevant municipal elections took place in June 2016 and October 2021.
Because these elections occur mid-year (2016) and late-year (2021), and because implementation timing is not consistently available at month-level precision across the datasets, it is not possible to apply strict “before/after election month” cut-offs for all features.
We therefore use election years as anchor points, interpreting the periods as approximations of the electoral cycles.

We defined three periods:

  • pre2016: interventions dated before 2016
  • 2016–2021: interventions dated between 2016 and 2021 (inclusive)
  • post2021: interventions dated after 2021 (not used in the main analysis)

Justification of cut-offs

  • Why pre2016 (and not pre2017):
    The election initiating the new cycle happened in June 2016, meaning that 2016 is a transition year spanning two mandates.
    Since implementation timing is not consistently available at month-level precision across datasets, we cannot reliably split 2016 into pre- and post-election interventions.
    We therefore use 2016 as the anchor year of the new cycle and define pre2016 as interventions dated before 2016, avoiding the systematic classification of all 2016 interventions as baseline infrastructure.

  • Why 2016–2021 (and not 2016–2022):
    The following election took place in October 2021, so interventions dated 2022 onwards clearly belong to the post-election cycle.
    Because we cannot reliably separate late 2021 from early 2022 across all datasets (given the lack of consistent month-level dates), we treat 2022+ as post2021 to avoid mixing post-election implementation into the 2016–2021 period.

Overall, these cut-offs provide a consistent and interpretable framework to compare infrastructure implementation across political cycles, without claiming month-level precision that is not available for all interventions.

Interactive maps

Interactive maps were used to support validation and exploratory analysis at both the city and small-area level.

For cycling, pedestrian-priority infrastructure, and Piazze Aperte, three maps were produced:

  1. Infrastructure implemented before 2016
  2. Infrastructure implemented up to 2021 (pre2016 + 2016–2021)
  3. Changes between periods, highlighting interventions implemented during 2016–2021

Census precinct boundaries for Milan (created by Gabriele Pinto) could be activated as an optional layer.
Maps also included layer toggles by infrastructure type and direct GSV links to support manual validation.

Legend
Cycling
Sede propria
Segnaletica
Aree pedonali
Area pedonale
Divieto di transito
Piazze Aperte
Permanente
Tattica