Which Came First, the Chicken or the Egg? What About Ducks?: Granger Causality Using Philippine Poultry Data

Research by: John Francis T. Diaz

Executive Summary

This study revisits the age-old causality puzzle of whether the chicken or the egg came first, extending it to include ducks using Philippine poultry production data. By applying the Granger causality test—a statistical technique that examines if one time series can predict another—the research explores temporal causality using quarterly data from 1980 to 2014, sourced from the Philippine Statistics Authority. The focus is on assessing directional causality among chickens, ducks, and their eggs.

The paper builds on the foundational work of Thurman and Fisher (1988), who used U.S. poultry data to statistically suggest that eggs preceded chickens. Likewise, genetic studies such as Eriksson et al. (2008) support the idea that mutations in eggs led to the emergence of chickens. By incorporating Granger causality in the context of Philippine data, the current study aims to reinforce the robustness and versatility of this econometric tool, even in contexts influenced by biological and genetic factors.

The methodology involves regressing production levels of chickens and ducks against their own lagged values and the lagged values of egg production, and vice versa. If one variable consistently helps predict another, Granger causality is established. Specifically, the study analyzes four regression models: chicken production as a function of its own lags and lagged egg values; egg production as a function of lagged egg and chicken values; and parallel models for ducks and duck eggs. The goal is to determine if a unidirectional causal relationship exists—e.g., from eggs to chickens or from chickens to eggs.

Empirical results show strong statistical evidence that in the case of chickens, egg production Granger-causes chicken production. This conclusion holds across one to three lags, with the first lag showing the highest model fit (R-squared of 0.858 and the lowest Akaike Information Criterion). This finding reaffirms earlier statistical and biological conclusions that chickens originate from eggs. Conversely, the analysis of ducks yielded inconclusive results. While past duck production was a strong predictor of current duck output, the same could not be consistently said for duck eggs. Only at the second lag did duck eggs significantly influence duck production, and even then, the model fit was poor, with the lowest R-squared and highest AIC among the models tested. This suggests no robust Granger causality from duck eggs to ducks.

Regarding egg production models, the data showed that chicken eggs were consistently determined by their own past values rather than past chicken numbers, further reinforcing the egg-first argument for chickens. For ducks, egg production was also self-determined with no significant influence from duck production, again pointing to the lack of clarity in causality direction.

The paper concludes by emphasizing the methodological value of Granger causality in addressing complex causality questions, even those outside typical economic frameworks. The study suggests future applications of Granger causality in diverse fields such as macroeconomics, finance, and agriculture to further explore complex causal relationships.

To cite this article: Diaz, J. F. (2025). Which came first, the chicken or the egg? What about ducks?: Granger causality using Philippine poultry data. Journal of Economic Statistics, 3(1), 18. https://doi.org/10.58567/jes03010001

To access this article: https://doi.org/10.58567/jes03010001

About the Journal

Journal of Economic Statistics (JES) aims to encourage theoretical and empirical research in economics and statistics, especially the application of statistics in solving economic problems.

Journal Metrics