Market Baseline Calculation

Module 6: Scenarios

Why Calculate a Market Baseline?

Scenario modeling requires establishing the supply-demand balance for white maize across Guatemala’s 22 departments. Before simulating biofortification production coverage scenarios, we must quantify:

  • Production capacity — How much maize can each department produce if all farmers adopt biofortified varieties?
  • Consumption demand — How much maize does each department’s population require?
  • Structural surplus/deficit — Which departments produce more than they consume, and which face structural deficits?

This baseline enables realistic scenario projections that account for geographic heterogeneity in agricultural capacity and population density. It also feeds two structural elements that govern how production reaches consumption across the country: the continuous adoption curves that determine each farmer’s contribution at any national production coverage level, and the gravity-based redistribution model that channels surplus from producing departments to deficit departments along real travel times.

The figures presented here correspond to the baseline scenario combining the 2022 biofortified seed variety (currently deployed in the field) with the current subsidy regime as of April 2026. Alternative seed variants and subsidy regimes are evaluated in the Shiny dashboard.

Data Integration

The market baseline integrates four data sources:

  • Farmer adoption data (from Module 2) — Contains farmer rankings, production capacity, segment assignments, calibrated survey weights, and the per-farmer continuous adoption parameters (p_entry, profile) that govern conversion intensity across production coverage levels.
  • Individual nutrient intake (from Module 1) — Contains population counts by department for consumption demand calculations.
  • Calibrated model configuration (from Module 2) — Contains the six continuous adoption curve parameters (floor_S/M/L, k_S/M/L) calibrated against empirical anchors.
  • Travel time matrix (from the one-shot routing script of this module) — Contains the 22×22 driving-time matrix between departmental capitals and the derived gravity weights.
Table 1: Input datasets for market baseline calculations
Input Datasets. Data sources for market baseline calculations (2022 seed, current subsidy)
Input Datasets
Data sources for market baseline calculations (2022 seed, current subsidy)
Dataset Records Variables Notes
Farmers (02_06_s22_sub00) 7,429 165 Per-farmer adoption scores, p_entry, profile
Model configuration (02_05) — — 6 continuous params: floor_S=0.95, floor_M=0.5, floor_L=0.2, k_S=8, k_M=4, k_L=0.5
Individuals (01_04) 46,017 93 Individual nutritional intake with survey weights
Travel time matrix (06_00) 484 6 22x22 asymmetric OSRM travel time matrix
Source: ENCOVI 2023 (MAGA-calibrated), Module 2 outputs, OSRM (OpenStreetMap routing).

The farmer dataset contains expanded records (from MAGA weight calibration (Ministerio de Agricultura, Ganadería y Alimentación, 2023)) with adoption priority scores and the continuous adoption parameters per farmer, while the individual dataset provides the population base for consumption calculations. The model configuration provides the six scalar parameters that define the sigmoid conversion curves used downstream.

Departmental Baseline Metrics

All calculations use survey weights to produce population-representative estimates. The baseline captures four dimensions for each department: farmer population by technology segment, land area allocation, annual consumption demand, and production capacity (current and maximum biofortified).

NoteConsumption Reference Values

Two consumption benchmarks inform our analysis:

  • Direct human consumption: 110 kg/capita/year — Lower-bound estimate for direct food use only (primarily tortillas).
  • FSI consumption: 174 kg/capita/year — Includes food, seed retention, animal feed, and post-harvest losses.

Consumption coverage calculations use the FSI estimate (USDA Foreign Agricultural Service, 2019) as the denominator, rather than direct human consumption (Ministerio de Agricultura, Ganadería y Alimentación, 2008), since biofortified production must satisfy total domestic demand.

Conversion factors used in the baseline:

  • Kilograms to quintales: ÷ 45.36 (1 qq = 100 lb = 45.36 kg)
  • Quintales to metric tons: × 0.04536 (1 qq ≈ 0.04536 MT)
  • Manzanas to hectares: × 0.7 (1 mz = 0.7 ha)
Table 2: Departmental baseline metrics for scenario modeling
Departmental Market Baseline Metrics. Farmers, land, consumption, and production capacity by department
Departmental Market Baseline Metrics
Farmers, land, consumption, and production capacity by department
Department Total Farmers Hybrid Farmers Non-hybrid Farmers Land (ha) Human Cons. (MT/yr) FSI Cons. (MT/yr) Current Prod. (MT) Max Bio Prod. (MT) Max Coverage (%)
Guatemala 19,316 314 19,002 20,257.6 367,944.1 582,020.6 37,049.1 42,243.4 7.3%
El Progreso 5,848 1,443 4,405 14,198.4 22,449.1 35,510.4 16,651.1 20,450.8 57.6%
Sacatepéquez 5,759 1,293 4,466 8,746.4 42,058.9 66,529.6 14,601.2 18,603.1 28.0%
Chimaltenango 46,641 12,142 34,499 24,384.7 83,299.0 131,763.9 48,320.4 60,009.1 45.5%
Escuintla 18,216 9,108 9,108 7,151.5 90,250.2 142,759.3 23,987.6 23,006.9 16.1%
Santa Rosa 37,379 16,335 21,043 16,149.5 49,575.8 78,419.9 40,210.0 44,477.4 56.7%
Sololá 9,459 2,463 6,997 17,184.8 49,054.4 77,595.1 36,763.5 47,485.1 61.2%
Totonicapán 19,718 3,344 16,374 29,563.7 54,423.5 86,088.1 45,321.0 59,007.5 68.5%
Quetzaltenango 25,551 5,289 20,262 43,607.2 94,319.1 149,195.6 117,061.0 149,819.2 100.0%
Suchitepéquez 13,115 5,122 7,993 8,182.6 65,358.7 103,385.6 25,183.0 28,647.1 27.7%
Retalhuleu 10,577 4,512 6,065 20,542.1 44,278.1 70,039.9 51,225.1 57,172.8 81.6%
San Marcos 54,490 12,403 42,087 52,370.3 133,230.2 210,746.0 115,168.8 149,205.5 70.8%
Huehuetenango 82,011 25,517 56,494 109,568.2 161,586.4 255,600.3 182,515.2 214,062.5 83.7%
Quiché 86,830 19,467 67,362 157,304.0 137,335.7 217,240.0 234,588.4 301,091.4 100.0%
Baja Verapaz 17,846 2,511 15,335 28,163.3 40,232.9 63,641.2 32,683.8 42,466.9 66.7%
Alta Verapaz 106,782 19,081 87,701 177,085.2 154,746.6 244,780.9 240,255.3 305,955.8 100.0%
Petén 42,446 9,241 33,205 249,562.6 62,040.1 98,136.2 483,453.3 592,053.7 100.0%
Izabal 37,863 11,178 26,685 21,081.0 53,807.8 85,114.2 56,915.6 64,666.2 76.0%
Zacapa 16,478 3,438 13,040 16,628.8 31,276.2 49,473.2 35,114.5 44,269.4 89.5%
Chiquimula 50,738 7,498 43,240 46,001.9 52,439.9 82,950.3 68,894.7 86,320.4 100.0%
Jalapa 21,668 2,807 18,861 28,209.9 44,635.6 70,605.4 49,675.7 64,366.3 91.2%
Jutiapa 54,742 14,157 40,585 60,268.9 61,735.3 97,654.0 120,057.7 147,643.1 100.0%
Total — 783,474 188,662 594,812 1,156,213 1,896,077 2,999,250 2,075,696 2,563,024 —
Source: ENCOVI 2023 (MAGA-calibrated). Human Cons. = 110 kg/capita/year (lower-bound estimate). FSI Cons. = 174 kg/capita/year (USDA 2019). Green = surplus (≥100%). Red = deficit (<50%). Production capacity shares cited in the text (e.g., a department's share of national production) are relative to the national total; consumption coverage and surplus/deficit status are relative to each department's own annual FSI consumption demand.

The baseline is geographically concentrated. Petén alone accounts for 23.1% of national maize production capacity, and together with Alta Verapaz and Quiché for 46.8%, while Guatemala department, the most populous, covers 7.3% of its own FSI demand from local production. Six departments (highlighted in green) reach or exceed 100% consumption coverage from local production alone; five (highlighted in red) stay below 50% even at full production coverage.

ImportantStructural Supply-Demand Imbalance

Even at 100% biofortified production coverage, 6 of the 22 departments satisfy their local consumption demand from their own production. The other 16 range from 7.3% to 91.2% coverage. This is what the redistribution model that follows is built for: without surplus moving between departments, a national consumption coverage target cannot be met from a geographically concentrated supply.

Continuous Adoption Model

Production coverage is implemented as a continuous adoption process rather than a binary admit / exclude rule. Each farmer carries two attributes calibrated in Module 2:

  • A per-farmer entry point p_entry — the national production coverage level at which the farmer begins adopting biofortified seed.
  • A behavioural profile S, M, or L (Small, Medium, Large entry-size) — that governs the steepness of conversion intensity from the entry point onwards.

At national production coverage p, every farmer with p_entry ≤ p is in the adoption pool. The fraction of that farmer’s potential biofortified production realised at production coverage p is the conversion ratio:

\[ r_i(p) = \text{sigmoid}(p, p_{entry,i}, \text{floor}_{profile}, k_{profile}) \]

where floor_{profile} is the conversion fraction at the entry point itself, and k_{profile} controls how rapidly conversion rises from the floor towards 1 as p increases. The six scalar parameters (floor_S, floor_M, floor_L, k_S, k_M, k_L) come from the model configuration calibrated in Module 2 against empirical adoption anchors via median absolute deviation. At p = 0 no farmer is in the pool; at p = 1 every farmer is in the pool with r_i(1) = 1 by construction.

The effective biofortified production of farmer i at production coverage p is therefore:

\[ \text{bio}_i(p) = \text{production\_biofortified}_i \times \text{weight\_calibrated}_i \times r_i(p) \]

This formulation captures realistic adoption patterns where early adopters are not uniformly distributed across departments, and where each farmer’s conversion intensity varies by behavioural profile and position along the production coverage curve. It also produces smooth saturation trajectories (shown in the consumption coverage analysis below) that converge to the maximum biofortified production capacity at full production coverage.

Consumption Coverage Before Redistribution

Before introducing inter-departmental redistribution, we analyse how departmental consumption coverage evolves across the national production coverage range (0–100%) based solely on local production. At each production coverage level p, the biofortified production realised in department d is the sum of bio_i(p) across all farmers i located in d. Coverage is the ratio between this realised production and the department’s annual FSI consumption demand.

This analysis surfaces the departments that saturate early (reach 100% local coverage at low national production coverage) — these are the natural supply nodes for the redistribution model that follows.

Line chart with national production coverage (percent) on the x-axis and departmental consumption coverage (percent) on the y-axis, one line per department based on local production only. Non-saturating departments are drawn as faint gray lines, while the departments that saturate are highlighted in color with a marker at the point where each first crosses the dashed 100 percent threshold. A few departments rise steeply and saturate early at low national coverage, marking them as natural supply nodes for redistribution.
Figure 1: Departmental consumption coverage evolution by national production coverage level (before redistribution)
NoteSaturation Timing

Departments that reach 100% consumption coverage at lower national production coverage levels have production capacity that exceeds local demand. These early saturators become natural supply nodes for the redistribution model: once their own demand is satisfied, additional biofortified production from local farmers is available to supply deficit departments elsewhere in the country.

Travel Time Infrastructure

Redistribution between Guatemalan departments is not symmetric in space. The country’s road network is shaped by mountain ranges (Sierra Madre, Sierra de los Cuchumatanes) that create significant travel time asymmetries between departments that are geographically close but poorly connected by road. The gravity-based redistribution model therefore uses real driving times rather than Euclidean distances or simple adjacency.

The 22×22 travel time matrix between departmental capitals is computed once, by a separate one-shot routing script. That script queries the Open Source Routing Machine (OSRM) over OpenStreetMap road network data and produces a deterministic matrix that depends only on the road network. The matrix is stored as a parquet file and re-used across all seed × subsidy combinations of the scenario pipeline.

Departmental Capitals

The reference points for the travel time matrix are the administrative seats (cabeceras departamentales) of each of Guatemala’s 22 departments, as defined by Guatemala’s Instituto Nacional de Estadística (INE). Coordinates were obtained from the GeoNames geographical database (https://www.geonames.org, Creative Commons Attribution 4.0 License), which aggregates data from official national mapping agencies and the OpenStreetMap community. Each coordinate corresponds to the populated place record classified as feature class P (populated place) and feature code PPLA (seat of a first-order administrative division) in GeoNames. Coordinates are in WGS84 (EPSG:4326).

Table 3: Departmental capitals of Guatemala with geographic coordinates
Departmental Capitals of Guatemala. Reference points for OSRM travel time computation
Departmental Capitals of Guatemala
Reference points for OSRM travel time computation
Department Capital (Cabecera) Latitude Longitude
Guatemala Guatemala City 14.6349 −90.5069
El Progreso Guastatoya 14.8536 −90.0672
Sacatepéquez Antigua Guatemala 14.5586 −90.7295
Chimaltenango Chimaltenango 14.6619 −90.8193
Escuintla Escuintla 14.2980 −90.7870
Santa Rosa Cuilapa 14.2765 −90.2984
Sololá Sololá 14.7725 −91.1866
Totonicapán Totonicapán 14.9082 −91.3607
Quetzaltenango Quetzaltenango 14.8347 −91.5183
Suchitepéquez Mazatenango 14.5336 −91.5037
Retalhuleu Retalhuleu 14.5340 −91.6781
San Marcos San Marcos 14.9647 −91.7963
Huehuetenango Huehuetenango 15.3195 −91.4714
Quiché Santa Cruz del Quiché 15.0295 −91.1502
Baja Verapaz Salamá 15.1025 −90.3160
Alta Verapaz Cobán 15.4713 −90.3707
Petén Flores 16.9316 −89.8923
Izabal Puerto Barrios 15.7280 −88.5940
Zacapa Zacapa 14.9719 −89.5294
Chiquimula Chiquimula 14.7979 −89.5449
Jalapa Jalapa 14.6354 −89.9889
Jutiapa Jutiapa 14.2916 −89.8953
Source: GeoNames geographical database (https://www.geonames.org, CC BY 4.0). Coordinates in WGS84 (EPSG:4326).

Travel Time Matrix

The matrix below shows driving times in minutes between every pair of departmental capitals. The matrix is asymmetric — t(i, j) ≠ t(j, i) in general — reflecting real directional differences in road routing, terrain gradients, and directional constraints in Guatemala’s road network.

Heatmap of a 22-by-22 matrix with origin departmental capitals along the x-axis and destination capitals along the y-axis. Each tile is labelled with the driving time in minutes and shaded on a blue gradient that deepens with longer travel times. The pattern is asymmetric across the diagonal, so travel time in one direction differs from the reverse, and the longest times cluster around the most distant, peripheral departments.
Figure 2: Driving travel times between departmental capitals (minutes) — OSRM asymmetric matrix

Gravity Weights

From the travel time matrix we derive a matrix of gravity weights w(i, j) = 1 / travel_time(i, j)^β, with a decay parameter β = 2, the Newtonian specification used in the trade gravity literature (Anderson & van Wincoop, 2003). The summary below characterises the resulting weight distribution over the 462 ordered department pairs, whose driving times run from 21.6 to 552.5 minutes.

Table 4: Gravity weight matrix summary statistics
Gravity Weight Matrix Summary. w(i,j) = 1 / travel_time(i,j)² — Anderson & van Wincoop (2003)
Gravity Weight Matrix Summary
w(i,j) = 1 / travel_time(i,j)² — Anderson & van Wincoop (2003)
Metric Value
Beta (decay parameter) 2
Number of department pairs 462
Min gravity weight 3.28e-06
Max gravity weight 2.14e-03
Mean gravity weight 9.13e-05
Min travel time (min) 21.6
Max travel time (min) 552.5
Mean travel time (min) 215.2
Source: OSRM travel time matrix. Statistics computed over off-diagonal pairs (origin ≠ destination).

Gravity-Based Redistribution Model

The redistribution model channels surplus biofortified production from departments that exceed their local demand to departments that face a deficit. The flow allocation follows the gravity formulation of (Anderson & van Wincoop, 2003), adapted here to inter-departmental commodity movement within Guatemala. The model is implemented in the utility function redistribute_gravity() of the framework’s centralized function file.

For each donor department i with surplus production S_i and each receiver department j with deficit D_j, the share of S_i that flows to j is proportional to:

\[ \text{flow}_{i \to j} \;\propto\; D_j \;\times\; w(i, j) \]

where w(i, j) = 1 / travel_time(i, j)² is the gravity weight defined above. The proportionality gives the allocation two structural properties: receivers with larger deficits attract a larger share of surplus, the demand pull, and receivers further away by road attract a smaller share, the distance friction. Flows are normalised so that the total volume out of i equals S_i. The asymmetric travel time matrix means the flow i → j may differ from j → i, which matters when topographic asymmetry makes one direction more accessible than the other.

The gravity model uses real driving times rather than nominal adjacency, captures the terrain effects that adjacency hides, and evaluates all donor-receiver pairs jointly without requiring an external ordering rule.

Production Conservation

A direct consequence of the proportional allocation is that total biofortified production is conserved through redistribution: the volume leaving each donor equals the volume distributed across receivers. The check below evaluates this at full national production coverage, the most demanding case, where every department is simultaneously active in the redistribution network.

Table 5: Production conservation validation at 100% production coverage
Production Conservation Check. Validation at 100% national production coverage — gravity redistribution
Production Conservation Check
Validation at 100% national production coverage — gravity redistribution
Metric Value
Total bio production before redistribution (qq) 56,504,051
Total bio production after redistribution — capped (qq) 56,504,051
Remaining excess after redistribution (qq) 0
Total accounted (capped + excess) (qq) 56,504,051
Difference (loss/gain) (qq) 0
Difference as % of total production 0%
A non-zero difference indicates production loss or creation through redistribution. Values near zero (< 1 qq) are acceptable due to floating-point arithmetic.

Consumption Coverage After Redistribution

With the gravity model applied, departmental consumption coverage at each production coverage level reflects both local production (from the continuous adoption curves) and received flows from surplus departments (from the gravity allocation). The two outputs below summarise the post-redistribution market structure: the saturation order describes how progressively more departments reach 100% local coverage as national production coverage rises, and the saturation curves describe the smoothed trajectories for every department.

Saturation Order

The table below ranks departments by the national production coverage level at which each first reaches 100% consumption coverage after gravity redistribution. Departments that do not reach saturation even at 100% national production coverage are marked with “—”, indicating insufficient inflow through the gravity network to close their local deficit.

Table 6: Departmental saturation order by production coverage level (after gravity redistribution)
Departmental Saturation Order. Production coverage level at which each department reaches 100% coverage (after gravity redistribution)
Departmental Saturation Order
Production coverage level at which each department reaches 100% coverage (after gravity redistribution)
Order Department Saturation Point
1 Petén 30%
2 Alta Verapaz 90%
3 Jutiapa 90%
4 Quiché 90%
5 Chiquimula 100%
6 Quetzaltenango 100%
— Baja Verapaz —
— Chimaltenango —
— El Progreso —
— Escuintla —
— Guatemala —
— Huehuetenango —
— Izabal —
— Jalapa —
— Retalhuleu —
— Sacatepéquez —
— San Marcos —
— Santa Rosa —
— Sololá —
— Suchitepéquez —
— Totonicapán —
— Zacapa —
Saturation point = national production coverage level at which the department first reaches 100% local consumption coverage post-redistribution. A dash indicates the department does not saturate even at 100% national production coverage.

Saturation Curves

The figure below shows the post-redistribution consumption coverage trajectory for every department across the full production coverage range. Compared with the pre-redistribution curves shown earlier, the gravity model raises the deficit departments as national production coverage grows, and caps the apparent surplus of donor departments at 100%, since their excess beyond local demand has been allocated outwards. At full production coverage the lowest departmental coverage rises to 58.6%, and the largest single gain is 59.3 percentage points in Guatemala, which moves from 7.3% to 66.5%.

Line chart with national production coverage (percent) on the x-axis and departmental consumption coverage (percent) on the y-axis, one line per department after gravity redistribution. Non-saturating departments appear as faint gray lines and saturating departments are highlighted in color, each with a marker where it first crosses the dashed 100 percent threshold. Compared with the pre-redistribution version, deficit departments are lifted closer to saturation while donor departments are capped at 100 percent as their surplus is allocated outward.
Figure 3: Departmental consumption coverage evolution by national production coverage level (after gravity redistribution)

Market Metrics After Redistribution

The final table consolidates the market-level outcomes at full national production coverage, comparing pre- and post-redistribution coverage to quantify the contribution of the gravity model to closing departmental deficits.

Table 7: Market structure metrics — pre vs post gravity redistribution
Market Structure: Pre vs Post Redistribution. Comparison at 100% national production coverage
Market Structure: Pre vs Post Redistribution
Comparison at 100% national production coverage
departamento n_farmers_total land_ha_total max_bio_production_mt max_coverage_before bio_production_after_mt coverage_after coverage_change
Guatemala 19316 20258 42243 7.258 387198 66.53 59.268
El Progreso 5848 14198 20451 57.591 33281 93.72 36.130
Sacatepéquez 5759 8746 18603 27.962 45406 68.25 40.287
Chimaltenango 46641 24385 60009 45.543 101335 76.91 31.364
Escuintla 18216 7152 23007 16.116 83643 58.59 42.474
Santa Rosa 37379 16150 44477 56.717 68666 87.56 30.846
Sololá 9459 17185 47485 61.196 68993 88.91 27.718
Totonicapán 19718 29564 59008 68.543 80397 93.39 24.846
Quetzaltenango 25551 43607 149819 100.000 149196 100.00 0.000
Suchitepéquez 13115 8183 28647 27.709 62253 60.21 32.505
Retalhuleu 10577 20542 57173 81.629 62909 89.82 8.189
San Marcos 54490 52370 149206 70.799 178667 84.78 13.980
Huehuetenango 82011 109568 214063 83.749 242435 94.85 11.100
Quiché 86830 157304 301091 100.000 217240 100.00 0.000
Baja Verapaz 17846 28163 42467 66.729 63345 99.53 32.806
Alta Verapaz 106782 177085 305956 100.000 244781 100.00 0.000
Petén 42446 249563 592054 100.000 98136 100.00 0.000
Izabal 37863 21081 64666 75.976 76301 89.65 13.670
Zacapa 16478 16629 44269 89.481 48459 97.95 8.469
Chiquimula 50738 46002 86320 100.000 82950 100.00 0.000
Jalapa 21668 28210 64366 91.163 69779 98.83 7.667
Jutiapa 54742 60269 147643 100.000 97654 100.00 0.000
Pre-redistribution metrics rely solely on local production at 100% production coverage. Post-redistribution metrics incorporate gravity-weighted flows from surplus departments to deficit departments.

Summary

Baseline Framework

The market baseline characterises the supply-demand structure of white maize across Guatemala’s 22 departments and supplies three structural elements consumed downstream by the scenario precomputation step:

  1. Per-department baseline metrics — Farmer counts by segment, land area, FSI consumption demand, current production, and maximum biofortified production capacity.
  2. Continuous adoption parameters — The six scalar parameters (floor_S/M/L, k_S/M/L) calibrated in Module 2 that define the per-farmer conversion ratio at every national production coverage level.
  3. Gravity redistribution infrastructure — The 22×22 OSRM travel time matrix and the derived gravity weights that channel surplus from donor departments to deficit departments.

Key Findings

  1. Structural geographic heterogeneity — Petén concentrates nearly a quarter of national maize production capacity, while several densely populated departments face structural deficits below 50% of FSI demand even at full local production. This heterogeneity makes inter-departmental redistribution a necessary component of any national consumption coverage target.

  2. Continuous adoption smooths the saturation trajectories — Per-farmer entry points and profile-specific sigmoid curves replace the prior binary adopt / not-adopt rule. Adoption progresses smoothly across national production coverage, with r_i(0) = 0 and r_i(1) = 1 as endpoint constraints by construction.

  3. Gravity redistribution narrows deficits without creating production — The Anderson & van Wincoop formulation channels surplus to deficit departments along real driving times. At full production coverage the lowest departmental coverage rises from 7.3% to 58.6%, while the number of departments reaching 100% stays at 6: the model closes the distance to saturation without adding saturating departments. Total production before and after redistribution matches to 0 qq.

  4. Travel time asymmetry — Inter-departmental driving times run from 21.6 to 552.5 minutes across 462 ordered pairs, and differ by direction. The asymmetric matrix returned by OSRM carries that structure, which a symmetric distance or a simple adjacency rule would not.

Application

The outputs of this script feed the scenario precomputation step, which uses the per-department coverage trajectories to assign biofortified maize consumption to children via the hierarchical priority assignment, and then aggregates the resulting individual nutrient and stunting outcomes at departmental and national level.

Limitations

  • Driving mode only — The travel time matrix assumes road transport. Alternative routes (river, air) are not modelled.
  • Static road network — Travel times reflect the OSM road network at the time of OSRM computation. Major infrastructure changes such as new highways or bridge construction would require the routing script to be run again.
  • Single reference point per department — Travel times are measured between cabeceras departamentales. Intra-departmental distribution costs are not modelled.
  • No seasonal variation — Travel times do not account for seasonal road conditions (rainy season closures, flooding).
Back to top

References

Anderson, J. E., & van Wincoop, E. (2003). Gravity with gravitas: A solution to the border puzzle. American Economic Review, 93(1), 170–192. https://doi.org/10.1257/000282803321455214
Ministerio de Agricultura, Ganadería y Alimentación. (2008). Segundo Informe Nacional sobre el Estado de los Recursos Fitogenéticos de Guatemala. MAGA/UNR.
Ministerio de Agricultura, Ganadería y Alimentación. (2023). Informe de Producción de Granos Básicos. MAGA/DIPLAN.
USDA Foreign Agricultural Service. (2019). Guatemala: Grain and Feed Annual (GAIN Report No. GT19007). USDA Foreign Agricultural Service.