Levels and changes in the value of exports and imports divided by aggregate GDP (the trade/GDP ratio) are occasionally used as measures of trade "openness." The oft-quoted work of Dollar and Kraay (2001) and the World Bank (2002) uses changes in the trade/GDP as a basis for classifying countries as "globalizers" or "non-globalizers". We argue that neither the level nor the change in a country's trade/GDP ratio can be taken as an indication of the "openness" of a country's trade policy. In particular, we examine the ways in which terms of trade shifts have affected trade/GDP ratio over the past two decades. While commodity prices were high in the early 1980s, commodity producing countries financed large trade deficits with expected export revenue. When the prices collapsed, their capacity to import fell precipitously and they were forced to close their trade deficits in order to balance the current account. Since the numerator of the trade/GDP ratio includes the sum of exports and imports, and the denominator includes the trade balance, this adjustment resulted in a decline and then stagnation in the trade/GDP ratio. Therefore, using stagnant or declining trade/GDP ratios to identify countries that are less "open" systematically picks out those countries that are highly dependent on commodities for their export revenue. Because these same countries have experienced stagnant or negative economic growth over the past two decades, the empirical evidence offered by Dollar and Kraay overstates the importance of trade policy in economic growth. Adding a "commodity dependence" dummy variable to their growth regressions reduces the magnitude of the apparent "growth effect" of their "openness" variable at least by half. We briefly review the literature on the relationships between commodity dependence and slow growth, highlighting that the whole question of "openness" vs. "closedness" is orthogonal to the problems of poor, slow growing, commodity producing countries.