A recent influential paper (O'Connell 1998) argues that panel data evidence in favor of purchasing power parity disappears once test procedures are altered to accommodate heterogenous cross-sectional dependence among real exchange rate innovations. We present evidence to the contrary. First, we...