<!-- class: inverse, center, title-slide, middle --> class: center, middle <style> g { color: rgb(0,130,155) } r { color: rgb(174,77,41) } y { color: rgb(177,148,40) } </style> # Lecture 09: Geospatial Data Sciences # and Economic Spatial Models ## <img src="figs/bse_primary_logo.png" style="width: 35%" /><br><br>Bruno Conte ## 04-05/Mar/2026 --- # Geospatial Data and Spatial Models: Schedule 1. ~~Introduction to (spatial) data and programming in `R`~~ **[08/Jan/2026]** 2. ~~Week 2-4: Vector spatial data~~ **[14 - 29/Jan/2026]** 3. ~~Week 5-7: Raster spatial data + (basic) interactive tools~~ **[05 - 19/Feb/2026]** 4. Week 8-10: Spatial models and applications with data **[25/Feb - 12 Mar/2026]** - ~~Week 8: Introduction to economic spatial models~~ - Week 9: Linking models to (spatial) data - Week 10: Models, data, and counterfactual simulations<br> <br> 5. <span style="color: rgb(177,148,40)">Take-home exam</span> **[27/Mar/2026]** --- # Spatial Models and Spatial Data Last class we learned the basics of <g>spatial models</g> - Economic theory + geographic features = <u>spatial distribution of economic activity</u> - Powerful tool for evaluating counterfactual experiments related to spatially linked shocks - I.e., treatment effects with spatial spillovers/general equilibrium effects <r>This class</r>: how to link these models to spatial data --- # Spatial Models and Spatial Data: Roadmap 1. <u>Illustrative context</u>: two examples of major <g>spatial shocks</g> - Expansion of trasportation infrastructure in the US and Brazil - How to read that through the lens of a spatial model? 2. <u>Linking model to data</u>: - How to summarize (spatial) <r>general equilibrium effects</r> of these policies? - How to <y>implement it</y> with real-world spatial data? --- class: center, middle # Getting Started: Two Examples of Spatial # Shocks or Policies --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - <u>Question</u>: what were the <r>welfare gains</r> from expansion of railroads in the 19th century US? -- .pull-left[<img src="figs/donaldson_hornebeck1.png" style="width: 100%" />] .pull-right[<img src="figs/donaldson_hornebeck2.png" style="width: 100%" />] --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - <u>Question</u>: what were the <r>welfare gains</r> from expansion of railroads in the 19th century US? .pull-left[<img src="figs/donaldson_hornebeck1.png" style="width: 100%" />] .pull-right[<img src="figs/donaldson_hornebeck3.png" style="width: 100%" />] --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - <u>Question</u>: what were the <r>welfare gains</r> from expansion of railroads in the 19th century US? - Quantification goes <y>beyond local treatment effects</y> (say, buffer of road density) - Market integration affects all locations through spatial linkages (and GE effects)! -- Donaldson and Hornbeck (2016): - Spatial model `\(\rightarrow\)` theory-based measures to summarize these GE effects - Combine it with data to quantify aggregate wefare gains - And hypothetical consequences under counterfactual scenarios --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - How to interpret the (effects of the) railroad expansion through the lens of a spatial model? Recall the economic activity in location `\(i\)`, `\(v_i L_i\)`, in equilibrium is `$$v_i L_i= X_i = \sum_{n \in S} \pi_{ni} \times v_n L_n,$$` where `\(\pi_{ni} \in [0,1] \equiv\)` share of `\(i\)`'s production exported to `\(n\)`. --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - How to interpret the (effects of the) railroad expansion through the lens of a spatial model? Recall the economic activity in location `\(i\)`, `\(v_i L_i\)`, in equilibrium is `$$v_i L_i= X_i = \sum_{n \in S} \pi_{ni} \times v_n L_n,$$` where `\(\pi_{ni} \in [0,1] \equiv\)` share of `\(i\)`'s production exported to `\(n\)`. Recall that `$$\pi_{ni} = \frac{A_i \left( w_i d_{ni} \right)^{-\theta}}{\sum_{s \in S} A_s \left(w_s d_{ns} \right)^{-\theta}},$$` what happens if `\(n\)` and `\(i\)` become integrated? And <r>if `\(n\)` and `\(s\)`</r>? --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"* - How to interpret the (effects of the) railroad expansion through the lens of a spatial model? - Important to account the effect of all (new) trade connections - Innovative approach compared to usual buffer-type of local effects - The authors propose a **<r>market access measure** that summarizes all connections `$$MA_i \approx \sum_{n \in S} d_{ni}^{-\theta} \times L_n$$` It is a weighted-average of all markets `\((L_n)\)`, where trade costs `\(d_{ni} \equiv\)` weigths - It captures how location `\(i\)` is well-connected (can access) all other markets in the economy! --- # Spatial Models and Spatial Data: Market Access Example .center[<img src="figs/example_mkt_access_1.png" style="width: 100%" />] --- # Spatial Models and Spatial Data: Market Access Example .center[<img src="figs/example_mkt_access_2.png" style="width: 100%" />] --- # Spatial Models and Spatial Data: Example 1 Donaldson and Hornbeck (2016): *"Railroads and American economic growth: A <g>market access</g> approach"*. .center[<img src="figs/donaldson_hornebeck4.png" style="width: 70%" />] How much `\(MA\)` captures <r>on top of local (buffer) effects</r>? --- # Spatial Models and Spatial Data: Example 2 Morten and Oliveira (2023): *"The Effects of Roads on Trade and Migration: Evidence from a Planned Capital City"*. .pull-left[ <u>Context</u>: major road infrastructure improvement in Brazil <u>Questions</u>: - What were the welfare gains from it? - Did <r>trade and migration</r> integration have different roles? <u>Method</u>: estimated aggregated welfare gains with MA approach ] .pull-right[ <img src="figs/melanie1.png" style="width: 80%" /> ] --- # Spatial Models and Spatial Data: Example 2 Morten and Oliveira (2023): *"The Effects of Roads on Trade and Migration: Evidence from a Planned Capital City"*. .pull-left[ <img src="figs/melanie2.png" style="width: 80%" /> ] .pull-right[ <img src="figs/melanie3.png" style="width: 80%" /> ] --- # Spatial Models and Spatial Data: Example 2 Morten and Oliveira (2023): *"The Effects of Roads on Trade and Migration: Evidence from a Planned Capital City"*. .pull-left[ <img src="figs/melanie2.png" style="width: 80%" /> ] .pull-right[ <img src="figs/melanie4.png" style="width: 80%" /> ] --- # Spatial Models and Spatial Data: Example 2 Morten and Oliveira (2023): *"The Effects of Roads on Trade and Migration: Evidence from a Planned Capital City"*. .center[ <img src="figs/melanie5.png" style="width: 80%" /> ] --- class: center, middle # Hands-on: Implementing Market Access # Concepts with Real-World Data --- # Spatial Models and Spatial Data: Hands-on .pull-left[ <u>Task</u>: compare the relationship <g>between economic activity</g> (population) and - Local road density - Market access Similar to Donaldson and Hornbeck (2016), <r>stronger relationship</r> with market access! <u>For that</u>, download the [<u>data</u>](https://www.dropbox.com/scl/fi/zxt6fcjgvr41o3sf3mpz0/data_for_class09.rdata?rlkey=9ifr0ntmtdv9futii4r4wzh8i&dl=1) and follow the next steps ] .pull-right[ <img src="figs/exercise_mktaccess_4.png" style="width: 100%" /> ] --- # Spatial Models and Spatial Data: Hands-on .pull-left[ 1. Retrieve the location (coordinate) of the town in each gridcell 2. Construct a bilateral distance matrix ] .pull-right[ <img src="figs/exercise_mktaccess_1.png" style="width: 100%" /> ] --- # Spatial Models and Spatial Data: Hands-on .pull-left[ 1. Retrieve the location (coordinate) of the town in each gridcell 2. Construct a bilateral distance matrix 3. Calculate `\(MA_i = \sum_{n \in S} d_{ni}^{-\theta} \times L_n\)`: - `\(L_n \equiv\)` population - `\(d_{ni} = e^{\text{distance}(i,n) \times 0.001}\)` - `\(\theta = 6.5\)` (Morten and Oliveira, 2023) ] .pull-right[ <img src="figs/exercise_mktaccess_3.png" style="width: 100%" /> ] --- # Spatial Models and Spatial Data: Hands-on .pull-left[ 1. Retrieve the location (coordinate) of the town in each gridcell 2. Construct a bilateral distance matrix 3. Calculate `\(MA_i = \sum_{n \in S} d_{ni}^{-\theta} \times L_n\)`: - `\(L_n \equiv\)` population - `\(d_{ni} = e^{\text{distance}(i,n) \times 0.001}\)` - `\(\theta = 6.5\)` (Morten and Oliveira, 2023) 4\. Retrieve road density within gridcells, correlate both ] .pull-right[ <img src="figs/exercise_mktaccess_5.png" style="width: 100%" /> ] --- class: center, middle # See you next class! --- # References - Donaldson, D. and Hornbeck, R., 2016. Railroads and American economic growth: A “market access” approach. *The Quarterly Journal of Economics*, 131(2), pp.799-858. - Morten, M. and Oliveira, J., 2023. The effects of roads on trade and migration: Evidence from a planned capital city. NBER Working Paper, 22158, pp.1-64.