The 237-Year Climb

The top panel is a stock; the bottom panel is a flow. The top line is the total amount outstanding on one day a year — the close of the fiscal year, Sep. 30 — plus the Treasury’s daily figure for Oct. 15, 2026. The bars below are a different measurement entirely: what came in minus what went out during each year. They are drawn on one shared horizontal scale so the two can be compared without being confused, and the bars stop where the data stops. This dataset carries an annual receipts-and-outlays figure for 86 of 237 fiscal years on the chart, all of them from fiscal 1940 on.

The dollars are nominal. Treasury and the Office of Management and Budget publish these series in the dollars of each year, and that is how they are drawn — no inflation adjustment, no adjustment for the size of the economy or the population. The rise on the chart therefore reflects prices and the growth of the country as well as borrowing, and nothing here separates them. Debt as a share of gross domestic product, the measure economists usually reach for, is not published here because this repository holds no verified GDP series to divide by; saying so is more useful than estimating one. The “× revenue” scale offers the denominator the data does support — the debt measured in years of total federal receipts — and it carries its own caveat, because that ratio also moves when revenue moves.

The scale toggle changes the story, on purpose. On the linear scale, the modern rise is drawn at full size and two centuries of smaller borrowing lie flat along the bottom — real, but too small to see. On the log scale, every gridline is ten times the one below it, which makes the eighteenth and nineteenth centuries legible and, in the same stroke, makes the modern rise look gentler than the linear picture. Neither view is the honest one on its own. Both are here so that neither can be presented as the whole picture.

No one is graded on this page, and no one is drawn on the chart unless you ask. Two overlays sit behind the scale toggle, both off by default. “Administrations” shades every presidency since 1789 in alternating grey — never party colours — with a line at each oath of office, and prints a table of the debt at the start and end of each one, on the convention stated there. “Events” marks the wars, panics, recessions and the one declared public-health emergency that readers most often ask to see against the curve, each on its date, each with its source. Neither overlay changes what the record shows or attributes it. Debt accumulates from appropriations passed by many Congresses, from taxes set by many Congresses, from interest on borrowing already done, and from wars, recessions and demographics that outlast any officeholder. A slope that begins in one person’s term is not that person’s doing, and this page will not imply otherwise. Where the site does report the debt alongside a term — on the individual president pages — it is labelled context and feeds no score.

Two facts the chart holds that are easy to miss. The total was lower at the end of the fiscal year than at the end of the year before in 80 of the 236 year-to-year changes on this chart, the first in 1794 and the most recent in 1957. And the low point of the whole series is not the start: the smallest total recorded is $33,733.05, in fiscal 1835. Neither is legible on the linear scale, which is what the log scale is there for.