When tolstoy meets leontief: luck, policies, and learning from miracles.

• Economic miracles, such as the Asian miracles, are not statistical accident to be ignored, and they are not as rare as commonly assumed i.e., the distribution of cross-country long-term growth follows a Power law. • Disentangling the role of policies from luck implies that economic miracles are alike, and every failure is a failure in its own way. High and sustained growth depends on maintaining the best possible policies while not being too unlucky. • The relatively few successes in achieving high and sustained growth such as the Asian miracles should constitute a focal point for the study of development policies. • In contrast, results derived from a large pool of development failures should be taken with skepticism, especially if it is based on the study of growth in the short to medium run. We show evidence that development "miracles," such as the Asian Miracles, may not be that rare as the distribution of cross-country long-term growth follows a Power Law. We propose a growth theory, disantangling the role of policies from luck, which is consistent with this stylized fact and predicts that high growth depends on maintaining the best possible policies while not being too unlucky. We argue that miracles are all alike and every failure is a failure in its own way, suggesting that the few growth miracles hold more clues about good policies than the vast number of growth failures. We infer important implications for the study of economic development. [ABSTRACT FROM AUTHOR]