A choropleth map is a thematic map that fills predefined geographic areas with colors or shades proportional to a statistical value measured in each area, such as unemployment rate by county or vaccination coverage by country. In data visualization within the Data discipline, it is the standard way to show how a variable varies across administrative or statistical regions.
The name combines the Greek words for area and multitude. Its strength is immediate recognition of spatial patterns, clusters and regional outliers; its weakness is that area size, classification choices and boundary definitions can all distort what readers perceive.
How a choropleth map encodes data
A choropleth map works by joining a table of values keyed by region to a boundary geometry and mapping each value to a color. The join usually relies on standardized codes, such as ISO 3166 country codes, FIPS codes for United States counties or NUTS codes for European regions, because matching on names fails with spelling variants.
The geometry comes from formats such as GeoJSON, TopoJSON or shapefiles. Each polygon receives a single uniform fill, which means the map implicitly assumes the value is evenly distributed within each area. That assumption is rarely true, and it is the source of several interpretation problems discussed below.
The following table summarizes the components every choropleth map requires.
| Component | Role | Typical choice |
|---|---|---|
| Enumeration units | Areas that receive a color | Countries, states, counties, postal zones |
| Attribute data | Value per area | Rates, ratios, densities, indices |
| Classification | Groups values into classes | Quantiles, natural breaks, equal intervals |
| Color scheme | Maps classes to colors | Sequential or diverging ramp |
| Projection | Flattens the globe | Equal-area projection preferred |
| Legend | Explains class ranges | Labeled color swatches |
Normalization and the choice of variable
A choropleth map should almost always display normalized values such as rates, percentages or densities, not raw counts. Large or populous regions naturally contain more of nearly everything, so a map of total cases or total sales mostly reproduces a population map and hides the real pattern.
Normalizing by population, land area or households converts totals into comparable measures, for example cases per 100,000 residents or median income. When raw counts genuinely matter, a proportional symbol map with sized circles placed on each region communicates magnitude more honestly than area fills.
| Variable | Suitable for a choropleth | Better alternative if not |
|---|---|---|
| Unemployment rate | Yes | Not needed |
| Population density | Yes | Not needed |
| Total population | No | Proportional symbol map |
| Number of stores | No | Dot map or rate per capita |
| Median household income | Yes | Not needed |
Classification methods and color schemes
The classification method decides which values share a color, and it can change the visual story more than any other design choice. A classed choropleth typically uses five to seven classes, because readers struggle to distinguish more shades reliably, while an unclassed choropleth maps values to a continuous ramp.
| Method | How classes are formed | Best use | Risk |
|---|---|---|---|
| Equal interval | Value range split evenly | Uniformly spread data | Most areas in one class if skewed |
| Quantile | Equal count of areas per class | Ranking and relative position | Similar values split apart |
| Natural breaks (Jenks) | Minimizes variance within classes | Clustered distributions | Hard to compare across maps |
| Standard deviation | Distance from the mean | Deviation from average | Assumes roughly normal data |
| Manual breaks | Analyst-defined thresholds | Policy or regulatory limits | Can be chosen to mislead |
Color schemes must match the data type. Sequential schemes run from light to dark for values that go from low to high, diverging schemes use two hues around a meaningful midpoint such as zero change or a national average, and qualitative schemes suit categories only. Perceptually uniform palettes, such as those popularized by ColorBrewer and viridis, keep equal data steps looking like equal visual steps and remain readable for people with color vision deficiency.
Choropleth maps compared with other spatial charts
A choropleth map is the right choice when data are aggregated to known boundaries and the question concerns regional differences. Other map and chart types answer different questions, and a heatmap is often confused with it even though it shows continuous intensity rather than administrative units.
| Chart type | What it shows | When to prefer it |
|---|---|---|
| Choropleth map | Rate per predefined area | Regional comparison of normalized values |
| Proportional symbol map | Magnitude as symbol size | Raw totals by location |
| Dot density map | Individual occurrences | Distribution within areas |
| Geographic heatmap | Smoothed point intensity | Hotspots ignoring boundaries |
| Cartogram | Areas resized by value | Correcting large-area dominance |
| Ranked bar chart | Exact values per region | Precise comparison without geography |
Pitfalls and misleading choropleth maps
The most persistent distortion is large-area dominance: sparsely populated but extensive regions occupy most of the visual space, so their values appear more important than those of small, dense regions. Election maps of large countries are the classic example, where geography rather than voters dominates the impression.
- Modifiable areal unit problem (MAUP): the same data aggregated to different boundaries can produce opposite patterns.
- Ecological fallacy: a high regional rate does not mean every individual in that region shares the trait.
- Small-number instability: rates in areas with few residents fluctuate wildly; smoothing or suppression is advisable.
- Non-equal-area projections such as Web Mercator inflate high-latitude regions and exaggerate their visual weight.
- Missing data needs a distinct neutral pattern, never a color that belongs to the scale.
Good practice is to state the classification method and data source period in the legend or caption, pair the map with a sorted bar chart for exact reading, and test whether a different classification reverses the conclusion.
The idea in brief
A choropleth map is a map where each region is colored according to a number. Darker areas usually mean higher values and lighter areas mean lower ones. Weather maps of rainfall by district or news maps of income by state often work this way.
The point is to see regional differences instantly without reading a long table. The catch is that big regions catch the eye even when few people live there, so a good map compares rates, not totals, and explains its color steps clearly.
A practical scenario
A public health analyst at a state agency needs to show where childhood vaccination coverage has fallen. She joins clinic records aggregated by county to county boundaries using FIPS codes and computes the percentage of five-year-olds fully vaccinated, rather than the number of vaccinated children.
She uses five quantile classes with a sequential blue palette and marks three counties with fewer than 200 children in a hatched pattern labeled as unstable estimates. The map reveals a contiguous cluster of rural counties below 85 percent coverage that a ranked table had not made obvious. The agency redirects mobile clinic visits to that cluster for the next quarter, and the analyst publishes the map alongside a sorted bar chart so readers can check exact values.