Using Chinese provincial data, we examine the relationship between relative prices and wages, which can be derived on the basis of the Balassa-Samuelson theorem. First, considering cross sectional dependency in our data, we implement recently developed panel unit root tests. We report strong evidence of non-divergence in the prices of provinces vis-a-vis those of the benchmark (Beijing) and weaker evidence in wages. In the latter case, the results seem sector-specific, suggesting that wages in some industries are not following the trend in Beijing. Secondly, using sectorlevel data, we find evidence that relative prices can be explained by relative wages.

JEL classification: E300, F300