From award years 2019–20 through 2024–25, statistics rose 13.82%. Applied mathematics fell 6.12%, mathematics fell 24.01%, and economics fell 39.99%. The endpoint is provisional, so the comparison is a current reading rather than a final trend verdict.

The official field-to-career crosswalk connects mathematics and applied mathematics to careers such as mathematicians, statisticians, and data scientists; applied mathematics also connects to actuaries. Economics connects to economists, research assistants, survey researchers, managers, and teachers. These are possible preparation paths, not observed graduate destinations.

Data scientists had 262,440 jobs, a $120,230 May 2025 median wage, and 33.5% projected growth for 2024–34. Actuaries had a $130,000 median and 21.8% growth. Mathematicians and economists were much smaller occupations, both with master’s degrees listed as typical entry education and projected growth near zero.

Occupation-level observations were highest here for data scientists and mathematicians. The detailed source also separates Claude.ai and first-party API interaction shares. Both describe observed Claude use in their source populations; neither measures the share of all work, productivity, displacement, or future employment.

Statistics degrees grew while the other fields contracted

Statistics bachelor’s degrees increased 13.82% from 2020–21 to 2025–26.

Applied mathematics decreased 6.12%, mathematics decreased 24.01%, and economics decreased 39.99%.

These are national first-major bachelor’s degrees reported through IPEDS.

The 2025–26 figures are provisional and may change.

The chart’s downloadable data includes the comparison-group label in every row.

Career patterns differ more than the field names suggest

The CIP–SOC crosswalk links mathematics and applied mathematics to mathematicians, statisticians, and data scientists.

Applied mathematics also links to actuaries.

Statistics links to actuaries, data scientists, statisticians, survey researchers, and related research or teaching careers.

Economics links to economists, economics teachers, research assistants, survey researchers, managers, and secondary teachers.

These links describe structural preparation. They do not measure graduate placement or employment probability.

Statistics and applied mathematics connect to faster-growing examples

Data scientists had 262,440 jobs and a $120,230 median wage in May 2025.

BLS projects 33.5% growth for data scientists from 2024 through 2034. Typical entry education is a bachelor’s degree.

Actuaries had a $130,000 median wage in May 2025 and projected growth of 21.8% through 2034.

Statisticians had a $105,650 median wage in May 2025 and projected growth of 8.5%.

Typical entry education is a bachelor’s degree for actuaries and a master’s degree for statisticians.

Mathematics and economics include narrower graduate-level careers

Mathematicians had 2,030 jobs and a $126,710 median wage in May 2025.

Their projected growth was −0.7% through 2034, with a master’s degree listed as typical entry education.

Economists had 17,790 jobs and a $124,720 median wage in May 2025.

Their projected growth was 1.2% through 2034, with a master’s degree listed as typical entry education.

Economics also links to social-science research assistants. That occupation had a $61,990 median wage in May 2025.

The two AI datasets show different measures

The occupation-level Anthropic measure reported scores of 0.4605 for data scientists and 0.4241 for mathematicians.

It reported 0.2418 for economists and 0.2107 for statisticians in the June 2026 release.

The detailed dataset separates Claude.ai from first-party API interactions during May 1–June 1, 2026.

Claude.ai shares were 0.98% for mathematicians, 0.61% for data scientists, and 0.33% for statisticians.

The Claude.ai share for economists was 0.09% during that period.

API shares were 0.77% for statisticians, 0.41% for mathematicians, and 0.39% for data scientists.

The API share for economists was 0.11%.

These percentages describe shares of published Claude interactions. They do not represent shares of all work.

Scorecard earnings could not be published safely

The Scorecard benchmark contains broader-field earnings, but it does not expose the outcome period or dated release.

Those earnings figures are omitted rather than presented as current outcomes.

BLS projections and observed Claude use remain separate measures. Neither establishes shortages, displacement, or AI employment effects.