mws-lab Documentation

A complete reference for mws-lab — Water Scarcity in Morocco: the methodology, the data sources, the OLS model, the diagnostics, the repository structure, the analysis pipeline, and the validation checks.

Overview

mws-lab is an exploratory analysis of the climatic, agricultural, and socioeconomic drivers of national freshwater withdrawals in Morocco (2000–2025). It uses five publicly available World Bank indicators, supplemented by national indicators from Morocco's 2025 State of the Environment Report.

"Morocco's renewable freshwater per capita fell from roughly 2,560 m³ per year in 1960 to about 620 m³ in 2020, well inside the United Nations' absolute water-scarcity threshold."

Methodology

The analysis is exploratory and associational. It is not designed to establish causal attribution. All relationships are reported as associations, not as causal effects.

Analytical Workflow

1. Data Acquisition     →  World Bank WDI + SoE 2025
2. Data Harmonization   →  Units · periods · metadata
3. Quality Control      →  Missingness · outliers · source
4. Exploratory Analysis →  Descriptives · trends
5. Empirical Modelling  →  Correlation · OLS (validated)
6. Robustness Checks    →  Diagnostics · specifications
7. Interpretation       →  Policy implications · limits

Regression Specification

FW_t = β0 + β1·IRRIG_t + β2·POPGR_t + β3·YIELD_t + β4·ARABLE_t + ε_t

β̂ = (XᵀX)⁻¹ Xᵀy

Data Sources

1. World Bank WDI

2. Morocco State of the Environment Report 2025

Renewable water resources per capita, potential and mobilizable water resources, groundwater abstraction, sectoral water allocation, seasonal rainfall, and national dam-filling rates for 2025 and 2026.

OLS Model

The regression is estimated by ordinary least squares on the complete-case window 2004–2019 (n = 16), limited by the irrigated-land series.

R² = 0.812          (adjusted R² = 0.744)
κ(X) ≈ 78,500       (condition number — severe multicollinearity)
n = 16              (complete-case window 2004–2019)

Reproducibility

To reproduce the analysis:

# 1. Install the package
pip install mws-lab

# 2. Fetch the World Bank data
python -c "from mws_lab import fetch; fetch.main()"

# 3. Or: fetch manually
#    Download the five WDI indicators for Morocco, 2000–2025
#    Keep missing values missing
#    Restrict the regression to 2004–2019 (n = 16)
#    Estimate by OLS; report R², adjusted R², κ(X), CIs, VIFs

The full manuscript, all seven figures, and the analysis pipeline are available in the repository and on Zenodo/Figshare.

Key Findings

Limitations

Repository Structure

mws-lab/
├── README.md
├── AUTHORS.md
├── CHANGELOG.md
├── CITATION.cff
├── CONTRIBUTING.md
├── LICENSE
├── MANIFEST.in
├── VERSION
├── pyproject.toml
├── requirements.txt
├── setup.py
├── setup.cfg
├── upload_morocco.py
│
├── mws_lab/                      # Python package
│   ├── __init__.py
│   ├── fetch.py
│   └── data/
│       ├── worldbank_data.csv
│       └── input.json
│
├── DATA/
│   └── raw/
│       ├── fetch_worldbank_data.py
│       ├── worldbank_data.csv
│       └── input.json
│
├── ANALYSIS/
│   ├── ANALYSIS_REPORT.md
│   └── analysis_insights.json
│
├── PUBLICATION/
│   ├── water_scarcity_morocco.pdf
│   ├── figures/
│   ├── Netlify/
│   │   └── public/
│   ├── credit_table.md
│   ├── manifest.json
│   └── netlify.toml
│
├── REPORTS/
├── EXTERNAL_REVIEW/
├── VALIDATION/
└── docs/
    ├── ITERATION_LOG.md
    └── comprehensive_audit_report.json

Analysis Pipeline

The Python package mws-lab includes a fetch script (mws_lab/fetch.py) that downloads the five World Bank indicators.

# Install
pip install mws-lab

# Fetch data
python -c "from mws_lab import fetch; fetch.main()"

# Or use the console script
mws-lab-fetch

The package includes:

• mws_lab/fetch.py — reproducible data-fetch script
• mws_lab/data/worldbank_data.csv — World Bank WDI extract
• mws_lab/data/input.json — indicator configuration

Validation

Five validation checks are executed at v1.0.0. All pass.

Scope & Limits

What mws-lab supports

• Reproducible data extraction from World Bank WDI.
• OLS regression with full diagnostics (VIF, condition number, CIs).
• Pairwise Pearson correlations on pairwise-complete observations.
• Open manuscript (CC-BY-4.0) and open code (MIT).
• Python package (PyPI: mws-lab).

What mws-lab does not support

• Causal inference (no quasi-experimental design).
• Basin-level disaggregation (national indicators only).
• Direct climate variables in the regression.
• Sub-annual or seasonal analysis.
• Cross-country generalization.

Citation

Zenodo Archive

@software{baladi2026mws_zenodo,
  author       = {Baladi, Samir},
  title        = {{Water Scarcity in Morocco}: A Multivariate Analysis
                  of Climatic, Agricultural, and Socioeconomic Drivers},
  year         = {2026},
  publisher    = {Zenodo},
  version      = {1.0.0},
  doi          = {10.5281/zenodo.23884236},
  url          = {https://doi.org/10.5281/zenodo.23884236},
  note         = {Ronin Institute / Rite of Renaissance}
}

APA (inline)

Baladi, S. (2026). Water scarcity in Morocco: A multivariate analysis of climatic, agricultural, and socioeconomic drivers (Version 1.0.0). Zenodo. https://doi.org/10.5281/zenodo.23884236

Status