Does PPP hold for Big Mac price or consumer price index? Evidence from panel cointegration

This paper examines the validity of purchasing power parity (PPP) using CPI and Big Mac prices. The benchmark model, i.e., the OLS method, which does not take nonstationarity into account, rejects the hypothesis of PPP regardless of prices used. We next use the panel cointegration method to consider the nonstationary nature of variables. Estimated results for CPI are mixed. The PPP is rejected when the nominal exchange rate is employed as the dependent variable but is not rejected when the price ratio is used as the dependent variable. By contrast, the PPP is overwhelmingly not rejected when the Big Mac price is used. Last, we remove the production bias and re-examine the same issue by using panel cointegration. The PPP is again decisively rejected when CPI price is used but not for Big Mac price. Accordingly, Big Mac price is more supportive to the validity of PPP than CPI price.