Giffen Good Check Calculator from Price and Quantity Data

This calculator checks observed price and quantity pairs for that positive response by fitting a straight-line demand relationship with ordinary least squares. It reports the slope, correlation, fit quality, and the direction of adjacent changes, then gives a clear classification. The result is a diagnostic for your dataset, not proof that price alone caused the observed quantities.

Prepare observations that support a meaningful comparison

Enter at least two observations for the same good, recording a positive own price and a nonnegative quantity demanded in each row. More observations across several distinct prices produce a more informative check than a single before-and-after comparison. Try to use the same market, time scale, product definition, and quantity unit throughout. The method assumes that rows are reasonably comparable, so mixing weekly quantities with monthly quantities or package prices with per-unit prices will distort the result. Other demand influences matter as well. Income, population, seasonality, promotions, product quality, rationing, expectations, and prices of substitutes or complements can all move quantity while price changes. If those factors vary sharply across rows, a positive fitted slope may reflect confounding rather than Giffen behavior. Repeated observations at the same price are accepted and contribute to the regression, but the dataset must contain at least two distinct prices. Use the optional slope tolerance when tiny positive movements should be treated as practically negligible because of measurement resolution or sampling noise. The tolerance uses quantity units per one price unit, so choose it in the same units as your data rather than as a percentage.

Understand how the check reaches its conclusion

The calculator fits quantity demanded as a linear function of own price using ordinary least squares. Its slope estimates the average change in quantity associated with a one-unit increase in price across the supplied observations. The result is classified as behaving like a Giffen good when that slope is strictly greater than the chosen slope tolerance. A zero or negative slope does not pass. The output includes the intercept for reproducing the fitted line, covariance and correlation for describing direction and strength, and R-squared for showing how much of the observed quantity variation the simple line accounts for. Correlation ranges from negative one to positive one, while R-squared ranges from zero to one. These statistics are descriptive, not a significance test. The calculator also sorts observations by price and counts adjacent pairs whose quantities rise, fall, or remain unchanged. Those counts make mixed patterns visible, but they do not replace the regression classification. Equal-price neighbors are excluded from adjacent comparisons because they contain no own-price change. All reported decimals are rounded deterministically, and the original rows are never reordered or modified outside the calculation.

Interpret a positive result with economic caution

A positive result means the fitted relationship in the submitted sample has quantity rising with own price by more than the tolerance. That is consistent with the defining observable pattern of Giffen behavior, but it is not sufficient to establish that the good is truly Giffen. Economic identification requires separating the substitution effect from the income effect and controlling for competing explanations. Classical Giffen behavior is associated with an inferior good for which the negative real-income effect of a price increase is strong enough to outweigh substitution away from the now more expensive good. Observational records rarely isolate that mechanism on their own. Review the correlation and R-squared alongside the classification: a slightly positive slope with weak fit and conflicting adjacent changes deserves much less confidence than a stable pattern collected under controlled conditions. Consider segmenting data by household income, location, product grade, or period, and use econometric controls or an experimental design when the decision is consequential. A negative result only says this dataset does not show a positive response above the threshold; it does not prove that Giffen behavior is impossible in every population or price range. The browser check is free, while an API request is billed at $0.002.