Learning Track: Measuring Development
1 Who This Track Is For
This track is for students who need to turn a contested concept into numbers — and to defend the way they did it. It uses development as its running example, because development is a concept where the disagreements are unusually explicit and the data are unusually good. But the method generalises: the same four-step chain from concept to number, and the same questions about normalization, weighting, and what gets left out, apply to measuring wellbeing, sustainability, democracy, institutional quality, innovation, or poverty.
It assumes you can read R code and run a script. You do not need prior knowledge of development economics. If you are new to R, work through the Getting Started with R track first; if you want to strengthen the data-handling side, the Data Workflow in R track is the natural companion.
By the end of this track you will be able to:
- Explain why there is no theory-neutral measurement, and identify the specific points at which theoretical commitments enter an indicator
- Gather cross-country data from the World Bank, the UNDP, WID.world, Our World in Data, and the Atlas of Economic Complexity, reproducibly and without re-downloading on every run
- Describe what the major development indicators measure, how they are constructed, and what each systematically omits
- Choose and justify an indicator with reference to your research question, and test whether your conclusions survive a different reasonable choice
- Read a composite governance indicator critically: identify the conception it encodes, and judge what a correlation with development outcomes can establish
2 Core Sequence
2.1 1. The Theory-Ladenness of Observation
The Theory-Ladenness of Observation
Why observation is never neutral, and where exactly theory enters: the chain from concept to conception to indicator to aggregation rule. Includes a demonstration in which the same twelve countries are ranked under four conceptions of development — output, health, equity, and sustainability — with the output and sustainability rankings coming out close to inverted. Start here: the rest of the track is much less useful without it.
2.2 2. Accessing Development Data in R
Accessing Development Data in R
The practical toolkit. Find indicator codes with WDIsearch(), download with WDI(), and wrap the call in a caching helper so the data are fetched once and committed alongside your code. Covers the traps that cost the most time: silently retired indicator codes, regional aggregates masquerading as countries, and joining sources on country names instead of ISO-3 codes.
2.3 3. Income, Output, and Complexity Indicators
Income, Output, and Complexity Indicators
What GDP and GNI per capita actually measure, why the real/nominal and PPP adjustments change the answer, and where the gap between GDP and GNI becomes economically informative. Then the Economic Complexity Index as a measure of productive capabilities rather than income flows — including a test of whether its relationship with income is really just resource wealth in disguise.
2.4 4. Measuring Inequality and Poverty
Measuring Inequality and Poverty
Building a Lorenz curve and a Gini coefficient from first principles, then the measures that report contested parts of the distribution directly: top shares and the Palma ratio. Covers why wealth is more concentrated than income in every country measured, and why two Gini coefficients from the same dataset are often not comparable. The second half turns to poverty lines and the Multidimensional Poverty Index, and to the live dispute over what the global poverty trend actually shows.
2.6 6. Measuring Institutions with the Worldwide Governance Indicators
Measuring Institutions with the Worldwide Governance Indicators
The track’s method applied end to end to a single, unusually contested indicator: institutional quality. This blog is about what the Unobserved Components Model behind the WGI assumes, which conception of “good governance” the six dimensions encode, and why four of the six dimensions turn out to be nearly interchangeable across countries but not within them. The entry closes with the four readings — causation, reverse causation, common cause, and circular measurement — that are possible, but hard to distinguish explanations for the governance–income correlation.
3 How to Use This Track
If you are writing an empirical paper, read page 1 before choosing your variables, then page 2, then whichever of pages 3–6 covers your indicators. Page 1 is what lets you write a methods section that argues for its measurement choices rather than merely reporting them.
If you are preparing for an exam in development or ecological economics, pages 3–6 are the substantive content, and page 1 supplies the critical framing that distinguishes a good answer from a list of definitions.
If you only want the practical skill, page 2 stands alone and will save you hours the first time you need cross-country data.
4 Going Further
Measurement is where empirical work starts, not where it ends. Once you have defensible variables:
- Reasoning with DAGs — how to think about what your correlations can and cannot establish
- Panel Data and Fixed Effects in R — exploiting within-country variation, the natural next step for the cross-sectional relationships shown throughout this track
- Publication-Ready Figures with ggplot2 — turning the exploratory plots here into something publishable
Citation
@inreference{graebner2026-measuring-development,
author = {Gräbner-Radkowitsch, Claudius},
editor = {Gräbner-Radkowitsch, Claudius},
title = {Learning {Track:} {Measuring} {Development}},
booktitle = {Method and Inquiry},
date = {2026-09-16},
url = {https://euf-methodology-hub.netlify.app/tracks/measuring-development},
langid = {en}
}