Seven Ways to Read Per-Pupil Spending

Every reading at a glance

Each row is one way of reading the same spending. The level is the latest school year (—). The change runs from — to — (from — for spending by what it buys, the first year the state publishes it), in constant dollars where the row is in dollars. Rank 1 is the lowest figure. Totals and student counts are sizes, so they are not ranked.

Source: every figure is read from ct_per_pupil_lenses.json, which Data/build_per_pupil_lenses.py builds from the CT State Department of Education’s Net Current Expenditures Per Pupil (ct_ncep_towns.json), EdSight staffing (administration.json), EdSight Education Financial System spending by function (efs_district_function.json), the need-weighted series (school_spending_need_weighted.json), BLS CPI-U and Employment Cost Index, and ACS median household income. Method notes are at the foot of the page.

1–2. Per pupil as published, and after inflation

The state’s Net Current Expenditures Per Pupil three ways: in the dollars of each year, in — dollars using consumer prices (CPI-U), and in — dollars using what school employers pay for staff (the Employment Cost Index for public schools). Most school spending is salaries and benefits, so the second adjustment asks what the same staff would cost in each year; where the two indexes grew at different rates, the two lines part. The dashed gray line is the median town after consumer-price inflation.

Source: CT State Department of Education, Net Current Expenditures Per Pupil (ncep), via ct_ncep_towns.json. BLS CPI-U CUUR0000SA0 (us_cpi_u.json) and BLS Employment Cost Index, state and local government, elementary and secondary schools, total compensation (eci_sl_schools in us_gov_cost_indexes.json). A school year uses the index for the calendar year it begins in.

3. Which half of the ratio moved

Total school spending after inflation, the number of resident students the town pays for, and spending per pupil after inflation, each set to 100 in —. When the student line falls and the spending line stays flat, the per-pupil line rises by the amount the drop in students alone would produce. The dashed gray line is the median town’s per-pupil line.

Source: net current expenditures (nce) and average daily membership of resident students (adm), the two halves of the state’s per-pupil figure, via ct_ncep_towns.json; deflated with CPI-U. For a town whose regional district runs every grade, the totals are the district’s across every year, so earlier and later years compare like with like.

4. Adults per 100 students

Full-time-equivalent staff for every 100 students, by job, in the districts that teach the town’s students. This counts people rather than dollars, so it is not moved by pay rates or inflation. More adults per student can come from keeping staff as enrollment falls or from adding staff; the count does not say which. It is not class size: specialists, special-education teachers and aides are counted, and the state publishes no class-size figure. The gray line is the median town’s total.

Source: CT State Department of Education, EdSight full-time-equivalent staffing by assignment category and October 1 enrollment, via administration.json. For a town taught by more than one district (its own elementary district and a regional high school, say), the figure is each district’s ratio weighted by how many of the town’s students it teaches, from the apportionment in school_spending_need_weighted.json.

5. What the money buys

Spending per pupil by function, in — dollars (CPI-U), from —. “Classroom and student support” is instruction, guidance, health, library and curriculum support. “Buildings and buses” is operating and maintaining buildings, and student transportation: costs tied to buildings and routes rather than to the number of students. “Administration and central services” is the board and superintendent’s office, principals’ offices, business and technology services. The gray line is the median town’s total.

Source: CT State Department of Education, EdSight Education Financial System, expenditures by function, via efs_district_function.json; weighted across the districts that teach the town’s students as above. This is a wider total than the per-pupil figure in charts 1–3: it includes transportation, which net current expenditures leave out.

6. Counting students the way the state’s aid formula does

Spending per pupil three ways: as published; per “need-weighted” pupil, where a student eligible for free or reduced-price meals counts as 1.3 students (more in a town where over 60% are eligible) and a multilingual learner as 1.25, as in the state’s Education Cost Sharing formula; and that figure with local differences in labor costs removed (the federal Comparable Wage Index for Teachers). A town with more high-need students, or in a higher-wage labor market, reads lower on the second and third lines. All in the dollars of each year. The dashed gray line is the median town per need-weighted pupil.

Source: school_spending_need_weighted.json (town grain, net current expenditures basis; weights from CGS 10-262f(25); NCES EDGE Comparable Wage Index for Teachers). The state’s current weights are applied to every year for comparability, including years before they took effect.

7. The median town, every reading at once

For each reading, the median across Connecticut towns in each year, set to 100 in —. Dollar readings are after consumer-price inflation. Where the lines spread apart, the readings tell different stories about the same years.

Source: as above. Each line is the median of the towns with a figure in that year; the towns of one K-12 regional district share a figure and each counts once.

Growing towns and shrinking towns

The median change from — to — in each reading, with towns grouped by how much their resident student count changed. Spending changes are after consumer-price inflation; spending by what it buys is from —.

Source: as above. Group sizes and boundaries are in the tooltips and in metadata.enrollment_groups of ct_per_pupil_lenses.json.

Higher-income towns and lower-income towns

The same median changes, with towns split into five equal-sized groups by median household income (ACS — release), lowest-income fifth first.

Source: as above, with US Census ACS 5-year median household income (B19013) via ct_town_income.json. The top group’s highest incomes are printed by the Census at the top of its published range; that does not change which fifth a town is in.

Enrollment change against each reading, every town

Each dot is a town: how much its resident student count changed from — to — (across), against how much the chosen reading changed (up). Pick a reading:

Source: as above. The dashed lines are zero change. Towns without both figures are left out and named in the summary.

Every town, every reading

Levels in — and changes since —. Click a column heading to sort. “Position spread” is how far apart a town’s highest and lowest positions are across the five level readings (per pupil, per need-weighted pupil, labor-cost adjusted, adults per 100 students, classroom dollars per pupil), in percentile points: 0 is the lowest town on a reading and 100 the highest, so readings that cover different numbers of towns compare on one scale. A large spread means the readings disagree about where the town stands.