Method, sources and limits

What every number means, where it comes from, how the two scores are calculated, and what this model cannot tell you.

The political score

Each election point carries four readings on one shared scale running from −100 (left) to +100 (right). Zero is the centre road: a permanent, unmoving reference line.

• Executive — the president’s ideological location. The earliest presidents include estimates where a directly comparable score is not available.
• House and Senate — the ideological centre of each chamber at that point. The two chambers can move differently because membership, election cycles and constituencies differ.
• Court — Martin–Quinn median scores from 1937 forward. Earlier values are appointment-based estimates and should be treated as lower-confidence indicators.

Default political score = Executive × 0.30 + House × 0.25 + Senate × 0.25 + Court × 0.20

The underlying congressional and presidential values are Voteview DW‑NOMINATE first-dimension figures, rescaled onto the −100 to +100 axis. The score is relative to the model’s national ideological scale. It is not a moral ranking, an approval rating, a vote share, or a direct measure of how liberal or conservative the public was.

Why party labels are not enough

Party coalitions and the meaning of left and right changed repeatedly. The route therefore follows comparable ideological readings rather than assuming that a Federalist, Democrat, Whig or Republican label meant the same thing in every century.

The economic score

Four parts combine into one prosperity reading on the same −100 to +100 axis.

• Output per person — inflation-adjusted output growth after accounting for population, so that population growth alone does not read as prosperity.
• Income and wages — whether household purchasing power and worker compensation were improving or deteriorating.
• Employment — job availability and labour-market distress. Deep depressions and sudden shutdowns produce strong negative readings.
• Price stability — rewards reasonably stable prices and penalises severe inflation or disruptive deflation. It carries less weight because stable prices matter less than lost output or lost work.

Default prosperity score = Output × 0.35 + Income × 0.25 + Employment × 0.25 + Prices × 0.15

The economic series is built from a set of anchor years with values interpolated between them. It is a visual historical estimate, not a formal national-accounts series.

What the journey’s visuals mean

• Flat centre road — a permanent zero reference. It never bends politically and never rises economically.
• Horizontal movement — left or right distance is the weighted political score. The route crosses the road when the score changes sides.
• Vertical movement — height above or below the road is the prosperity score. Decorative terrain roughness surrounds the route but does not change the score.
• Ball size in chase views — the chase camera stays tied to the flat baseline, so prosperity hills bring the ball closer and valleys push it further away.
• Cross-line — connects the ball to the centre road at the same year, making political distance and economic elevation easier to compare.
• Historical text — short context describing events that may have influenced the current point or the next several years. It is explanatory background, not a causal claim.

Limits and interpretation

Notes on the source records

• President names in the underlying records are raw roll-call register strings, written “first middle, suffix last” with inconsistent capitalisation — for example James, Jr. Madison and William, Jr. Mckinley. The app displays a tidied common name and keeps the original string on each point’s detail panel. No score, date or record was altered.
• Hiram Ulysses Grant is the birth name recorded for the president universally known as Ulysses S. Grant. The app shows the familiar name and preserves the source string.
• Each point is labelled with the president who served during the Congress that the election produced. That is why the 1864 point reads Andrew Johnson even though Lincoln won the 1864 election: Johnson held the office for most of the resulting Congress. The convention is consistent across the whole series.
• Thomas Jefferson’s executive reading sits close to zero because his DW‑NOMINATE first-dimension estimate is near the midpoint of that era’s scale. It reflects the source measure, not a claim that Jefferson was a centrist by today’s standards.

The compass quiz

The 46 statements on the Compass page, their five-point answer scale and the arithmetic that turns them into three axis scores are the work of SapplyValues, which in turn builds on 8values. They are used here under the MIT licence and reproduced word for word — the questions are the measuring instrument, so rewording them would quietly change what the test measures. The original source is on GitHub.

What is ours: the presentation, and everything that compares a result with the historical record. Note that only the economic axis is matched against history. The model on this site scores institutions on a single left–right dimension, so it has nothing to compare an authority or social reading against; those two are reported on their own terms and go no further.

The quiz runs entirely in your browser. Answers are kept in this device’s local storage so a half-finished quiz survives a reload, and they are never transmitted anywhere.

Keyboard shortcuts

← →

Step through years on the chart or the journey

Home / End

Jump to the first or last point on the chart

Space

Pause or resume the journey

M

Open or close the journey’s top-down map

V

Cycle the journey camera

Esc

Close the point detail panel