Empirical studies in trade, structural change and growth

There is a general consensus regarding the positive relationship between trade and productivity growth. Openness to trade encourages an efficient allocation of the factors of production within that trade. This thesis encompasses three studies that analyse the relationships between trade policy and openness and economy-wide productivity growth and its components through different channels. In doing so, we attempt to add to the existing literature on international trade, while accounting for some observed shortcomings in the existing literature. Firstly, empirical studies examining relationships between trade, resource allocation and economy-wide productivity tend to focus only on developing economies and as such our studies comprise a mix of developed and developing nations. Furthermore, in the case of productivity growth, attention is usually biased in favour of looking at aggregate productivity, potentially missing important details at a disaggregated level. We account for this by conducting studies using disaggregated data so that we can identify any patterns or trends that may be masked by aggregate data. In addition to this, the trade-growth literature faces criticisms regarding its inability to identify an exogenous measure of trade and as such we employ the use of an exogenous instrument for trade to conduct a study on trade and productivity. In our first study (Chapter 2), we examine the relationship between trade liberalisation events and structural adjustment in employment and output. To conduct this study, we employ the 3-digit level of the International Standard Industrial Classification (ISIC), Revision 2 data for the manufacturing sector, covering the period 1976 to 2004 for a sample of 35 countries. We also investigate the conditioning effects of complementary policies, in particular institutional quality, on the trade-adjustment relationship. We use data on institutions from the Economic Freedom of the World Index. We find that the use of aggregate data indicates the absence of a systematic relationship between trade liberalisation and structural adjustment. However, through disaggregation, we find that the occurrence of a trade liberalisation event reduces adjustment in intermediate goods employment and output and increases adjustment in capital goods output. In our second study (Chapter 3), we use a panel of 38 countries and employ the 10-sector productivity database derived from the Groningen Growth and Development Centre (GGDC) for the period 1990 to 2005, in order to explain labour productivity gaps across developing regions. Specifically, we analyse patterns of economy-wide productivity and its two components across countries within Latin America, Africa and High-Income regional groupings. The first component, structural change, captures changing sectoral shares of employment as labour reallocates across sectors. The second component, the within component, captures the reallocation of resources within sectors as well as technological improvements occurring within sectors. Our findings suggest that differences in economic performances across regions are accounted for by negative structural change occurring in individual countries within these regions. This means a reallocation of employment from high productivity activities in favour of lower productivity ones, thereby contributing negatively to overall productivity growth. Furthermore external shocks such as falling oil prices appear to drive this type of growth reducing structural change. Finally in Chapter 4, we investigate the relationship between trade openness and economy-wide productivity and its structural change and within components as defined in Chapter 3. We use a panel of 38 countries, again employing the GGDC 10-sector productivity database for the period 1965 to 2006, along with the complete gravity dataset provided by Head, Mayer and Ries (2010). Our findings of this study support theories that suggest a positive relationship between trade and economy-wide productivity. Our results also indicate that it is the within component of economy-wide productivity that is driving this results