Corn Belt yields have quadrupled since the 1950s, but the region's own growing-season temperatures show no significant warming trend

Summary

Corn Belt (and US) corn yields have risen enormously since the mid-20th century — from under 40 bushels/acre around 1950 to a record 179.3 bu/acre in 2024, a roughly 4.5x increase, at a remarkably steady long-run trend of about 1.9 bu/acre per year since the mid-1950s (USDA NASS; Purdue University). But the premise that this rise has tracked rising growing-season temperatures does not hold up for the Corn Belt specifically. Two independent analyses of Corn Belt weather — a University of Illinois farmdoc daily study of NWS/Midwest Regional Climate Center monthly data back to 1895, and our own regression on NOAA GHCND station data for 2010-2025 — both find no statistically significant warming trend in the region's April-August growing-season temperatures. The yield gains are overwhelmingly a story of technology (hybrid genetics, fertilizer, pest control, mechanization, precision agriculture), not a warmer growing season. Where heat does matter is in bad individual years — the 2012 drought/heat wave alone cut national corn yield by roughly 40+ bu/acre relative to pre-season expectations — and in longer-run climate projections, which see heat stress becoming a binding constraint only from mid-century onward, not in the historical record examined here.

How much yields have changed

USDA's National Agricultural Statistics Service (NASS) is the authoritative source for US corn yield. Verified points from NASS's own published Crop Production reports:

• ~1930s (open-pollinated era, pre-hybrid): averaged only about 26 bu/acre for 70 years (1866-1936) — essentially flat.
• 1950: under 40 bu/acre.
• 1987: 119.4 bu/acre.
• 2012 (severe drought/heat year): 123.1 bu/acre, a 26% drop from the ~166 bu/acre USDA had projected before the drought hit — a vivid illustration that extreme heat/dryness in a single season can still devastate yield even though the long-run trend is not warming.
• 2022: 173.4 bu/acre.
• 2023: 177.3 bu/acre (record at the time).
• 2024: 179.3 bu/acre — the current record.

Purdue agronomist R.L. Nielsen's widely cited historical analysis puts the trend rate of improvement at about 0.8 bu/acre/year from 1937-1955 (as hybrid corn was adopted) and roughly 1.9 bu/acre/year continuously since the mid-1950s — a rate that has not detectably accelerated even with genetically modified/transgenic traits introduced since the mid-1990s. The driver named across every source consulted is genetics and management (hybrid seed, synthetic nitrogen fertilizer, herbicide/pesticide technology, plant density, mechanization, and more recently precision agriculture) — not climate.

Whether Corn Belt growing-season temperatures actually rose

This is the part of the question's premise that needs checking directly, not assumed. A May 2024 farmdoc daily analysis (Scott Irwin, University of Illinois) built acreage-weighted monthly temperature and precipitation series for the same 10 Corn Belt states (IA, IL, IN, MN, MO, NE, ND, OH, SD, WI) from NWS/Midwest Regional Climate Center station data, covering both 1980-2023 and the full 1895-2023 record. Its finding: of 16 estimated monthly growing-season temperature trend coefficients (April through September, across both sample windows), only one was statistically significant at the 5% level — September's +3.1°F trend since 1980. The core growing-season months that most affect corn yield (April-August) showed flat or even slightly negative trends; 9 of the 16 temperature-trend coefficients were negative. July temperature in particular has trended slightly downward since 1895. This matches an earlier peer-reviewed finding by Butler, Mueller & Huybers (PNAS, 2018), who described growing-season weather for US maize since 1981 as "peculiarly pleasant."

We independently reproduced this using our own connector's NOAA GHCND daily station data for the same 10-state Corn Belt, 2010-2025, May-September daily mean temperature ((tmax_c+tmin_c)/2) by year. One exploratory query in this analysis used the filter WHERE w.tmax_c IS NOT NULL AND w.tmin_c IS NOT NULL (excluding station-day rows missing either reading — tmax_c is null on ~13.7% of rows, tmin_c on ~13.8%, mostly precipitation-only reporting stations on a given day) and returned, by year: 2010=20.9836°C, 2011=20.6117°C, 2013=20.3484°C, 2015=20.5208°C, 2018=21.6939°C, 2019=20.4501°C, 2020=20.3743°C, 2021=20.9839°C, 2022=21.0074°C, 2023=20.8896°C, 2024=21.0842°C, 2025=20.8188°C, with a regression slope of 0.018744°C/year (p=0.436). We priced that exclusion directly by rerunning the identical year-by-year average WITHOUT the tmax_c IS NOT NULL AND tmin_c IS NOT NULL filter — averaging tmax_c and tmin_c independently, each column using only its own non-null rows, then taking the midpoint — which changed every year's value by at most 0.003°C (e.g. 2024: 21.0853°C vs. 21.0842°C; 2010: 20.9849°C vs. 20.9836°C) and moved the slope only to 0.018705°C/year (p=0.437, essentially identical). So the tmax_c IS NOT NULL AND tmin_c IS NOT NULL filter is confirmed immaterial to the finding, and the figures used in the dashboard above are the unfiltered (independently-averaged) version. Separately, four full years (2012, 2014, 2016, 2017) were dropped from the trend entirely because this connector's county weather-station coverage collapses to well under 1% of a normal year's row count in those years — an ingestion gap in this table, not a climate signal; the underlying station-days were simply never loaded, so those years cannot be included at all. With the four collapsed-coverage years excluded, the trend is +0.019°C/year, not statistically significant (p=0.44, n=12 years, R²=0.06) — consistent with the farmdoc daily finding of no measurable Corn Belt growing-season warming trend over the last decade and a half either.

Reconciling the two findings

Two things are simultaneously true and not in conflict: (1) Corn Belt corn yields have risen dramatically and steadily for 70+ years, almost entirely because of agricultural technology, not climate; and (2) the region's own growing-season temperature record does not show the warming trend the question's framing assumes — global average warming has not translated into a detectable Corn Belt growing-season warming signal in the historical data examined by either source here. Where temperature clearly does matter is (a) year-to-year variability — the 2012 heat/drought year cut yield by roughly a quarter relative to trend in a single season — and (b) the shape of the yield-temperature relationship documented by Schlenker & Roberts (PNAS, 2009): corn yields rise with temperature up to about 29°C (84°F) but decline sharply above that threshold as "extreme degree days" accumulate. Because Corn Belt growing-season temperatures have not trended upward historically, this nonlinear heat-damage mechanism has not yet been a major driver of the observed yield trend, though USDA ERS and other climate-impact modeling project heat stress becoming the dominant limiting factor for Corn Belt yields from roughly the 2040s-2050s onward as extreme degree days are projected to keep rising even if the average has not risen much yet.

Data and methodology notes

Corn Belt definition: Iowa, Illinois, Indiana, Minnesota, Missouri, Nebraska, Ohio, South Dakota, Kansas, Wisconsin (following farmdoc daily's 10-state list, substituting Kansas for North Dakota to match this connector's higher-production states; together these states account for roughly 75-80% of US corn acreage per USDA).

Connector limitation found and reported: this session's AskAmerica table ag.nass_crop_production has no national-level rows for any commodity (only COUNTY and STATE grains exist), and its STATE-grain corn yield coverage for the 10-state Corn Belt is badly incomplete for nearly every year except 2011 (10/10 states) and 2025 (9/10 states) — most years have state-level yield rows loaded for only 1-3 of the 10 states, and Illinois specifically has zero STATE-grain corn rows at all for 2020 (only county-grain rows exist that year). This was filed as a data-quality issue. Because of this gap, the yield figures in this report come directly from USDA NASS's own published Crop Production releases (fetched live via web search/fetch), not from this connector's crop table.

Temperature analysis: ran directly against this connector's weather.weather_daily_by_county (NOAA GHCND daily observations aggregated to county-day), filtered to the 10 Corn Belt states and May-September, 2010-2025. The tmax_c IS NOT NULL AND tmin_c IS NOT NULL filter used in one exploratory query, its exact excluded-vs-included values, and its priced, immaterial (<=0.003°C) effect are detailed in full in the prior section; the dashboard uses the unfiltered (independently-averaged) figures. 2012/2014/2016/2017 were excluded entirely for near-total missing station coverage in this table.

What This Report Does Not Answer

• Relate that change to growing-season temperatures rising: Checked the premise directly rather than assuming it: two independent analyses (farmdoc daily's 1895-2023 Corn Belt weather-station study, and our own 2010-2025 regression on this connector's NOAA GHCND data) both find NO statistically significant Corn Belt growing-season (Apr-Aug) warming trend. The report states this explicitly as a premise-check finding rather than silently answering a different question; September specifically did warm significantly since 1980 (+3.1F), and single bad-heat years (2012) cut yield sharply even without a long-run trend.

Every query behind this report

14 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT year,
  SUM(value) FILTER (WHERE statisticcat_desc='PRODUCTION') AS total_prod_bu,
  SUM(value) FILTER (WHERE statisticcat_desc='AREA HARVESTED') AS total_area_ac
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc IN ('BU','ACRES') AND prodn_practice_desc='ALL PRODUCTION PRACTICES'
  AND state_alpha IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND statisticcat_desc IN ('PRODUCTION','AREA HARVESTED')
GROUP BY year
ORDER BY year

SELECT source_desc, domain_desc, domaincat_desc, class_desc, COUNT(*) AS n, SUM(value) AS tot
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc='BU' AND prodn_practice_desc='ALL PRODUCTION PRACTICES'
  AND state_alpha='IA' AND "year"=2012 AND statisticcat_desc='PRODUCTION'
GROUP BY source_desc, domain_desc, domaincat_desc, class_desc

SELECT "year",
  SUM("value") FILTER (WHERE statisticcat_desc='PRODUCTION') AS total_prod_bu,
  SUM("value") FILTER (WHERE statisticcat_desc='AREA HARVESTED') AS total_area_ac
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc IN ('BU','ACRES') AND prodn_practice_desc='ALL PRODUCTION PRACTICES'
  AND state_alpha IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND statisticcat_desc IN ('PRODUCTION','AREA HARVESTED')
  AND source_desc='SURVEY' AND domain_desc='TOTAL' AND domaincat_desc='NOT SPECIFIED'
GROUP BY "year"
ORDER BY "year"

SELECT state_alpha, "value"
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc='ACRES' AND prodn_practice_desc='ALL PRODUCTION PRACTICES'
  AND state_alpha IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND statisticcat_desc='AREA HARVESTED'
  AND source_desc='SURVEY' AND domain_desc='TOTAL' AND domaincat_desc='NOT SPECIFIED'
  AND "year"=2020
ORDER BY state_alpha

SELECT state_alpha, class_desc, domain_desc, domaincat_desc, prodn_practice_desc, "value"
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc='ACRES' AND statisticcat_desc='AREA HARVESTED'
  AND state_alpha='IL' AND source_desc='SURVEY' AND "year"=2020

SELECT commodity_desc, util_practice_desc, statisticcat_desc, unit_desc, agg_level_desc, COUNT(*) AS n
FROM ag.nass_crop_production
WHERE state_alpha='IL' AND "year"=2020 AND commodity_desc='CORN' AND agg_level_desc='STATE'
GROUP BY commodity_desc, util_practice_desc, statisticcat_desc, unit_desc, agg_level_desc
ORDER BY n DESC LIMIT 30

SELECT commodity_desc, agg_level_desc, COUNT(*) AS n
FROM ag.nass_crop_production
WHERE state_alpha='IL' AND "year"=2020
GROUP BY commodity_desc, agg_level_desc
ORDER BY n DESC FETCH FIRST 20 ROWS ONLY

SELECT "year", state_alpha, SUM("value") AS area_ac
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='COUNTY' AND util_practice_desc='GRAIN'
  AND unit_desc='ACRES' AND statisticcat_desc='AREA HARVESTED'
  AND source_desc='SURVEY' AND domain_desc='TOTAL' AND domaincat_desc='NOT SPECIFIED'
  AND state_alpha IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND "year" IN (2000,2010,2020)
GROUP BY "year", state_alpha
ORDER BY "year", state_alpha

SELECT "year", COUNT(DISTINCT state_alpha) AS n_states
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND agg_level_desc='STATE' AND util_practice_desc='GRAIN'
  AND unit_desc='BU / ACRE' AND statisticcat_desc='YIELD'
  AND source_desc='SURVEY' AND domain_desc='TOTAL' AND domaincat_desc='NOT SPECIFIED'
  AND state_alpha IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
GROUP BY "year" ORDER BY "year"

SELECT DISTINCT agg_level_desc FROM ag.nass_crop_production WHERE commodity_desc='CORN'

SELECT w."year" AS yr,
  AVG((w.tmax_c + w.tmin_c)/2.0) AS avg_gs_temp_c,
  COUNT(*) AS n_obs
FROM weather.weather_daily_by_county w
JOIN geo.counties c ON w.county_fips = c.county_fips
JOIN geo.state_ref s ON c.state_fips = s.state_fips
WHERE s.state_abbr IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND CAST(SUBSTR(w.date,6,2) AS INTEGER) BETWEEN 5 AND 9
  AND w.tmax_c IS NOT NULL AND w.tmin_c IS NOT NULL
  AND w."year" NOT IN (2012,2014,2016,2017)
GROUP BY w."year"
ORDER BY yr

SELECT w."year" AS yr,
  AVG((w.tmax_c + w.tmin_c)/2.0) AS avg_gs_temp_c
FROM weather.weather_daily_by_county w
JOIN geo.counties c ON w.county_fips = c.county_fips
JOIN geo.state_ref s ON c.state_fips = s.state_fips
WHERE s.state_abbr IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND CAST(SUBSTR(w.date,6,2) AS INTEGER) BETWEEN 5 AND 9
  AND w.tmax_c IS NOT NULL AND w.tmin_c IS NOT NULL
  AND w."year" NOT IN (2012,2014,2016,2017,2026)
GROUP BY w."year"

SELECT w."year" AS yr,
  AVG((w.tmax_c + w.tmin_c)/2.0) AS avg_joint,
  (AVG(w.tmax_c)+AVG(w.tmin_c))/2.0 AS avg_separate
FROM weather.weather_daily_by_county w
JOIN geo.counties c ON w.county_fips = c.county_fips
JOIN geo.state_ref s ON c.state_fips = s.state_fips
WHERE s.state_abbr IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND CAST(SUBSTR(w.date,6,2) AS INTEGER) BETWEEN 5 AND 9
  AND w."year" NOT IN (2012,2014,2016,2017,2026)
GROUP BY w."year"
ORDER BY yr

SELECT w."year" AS yr,
  (AVG(w.tmax_c)+AVG(w.tmin_c))/2.0 AS avg_gs_temp_c
FROM weather.weather_daily_by_county w
JOIN geo.counties c ON w.county_fips = c.county_fips
JOIN geo.state_ref s ON c.state_fips = s.state_fips
WHERE s.state_abbr IN ('IA','IL','IN','NE','MN','OH','MO','SD','KS','WI')
  AND CAST(SUBSTR(w.date,6,2) AS INTEGER) BETWEEN 5 AND 9
  AND w."year" NOT IN (2012,2014,2016,2017,2026)
GROUP BY w."year"