6 Conclusions and Further Work
How resources are allocated to their most efficient use is an important concern in economics. Almost all of the empirical research on this topic concentrates on entry and exit of firms. This paper suggests that looking at the relative importance of added and dropped products is also important. Using a unique dataset that covers products at 10 digit codes for all exporting firms in New Zealand from 1997 to 2007, the patterns of their product mix, how it changes over time and how this relates to firm characteristics are analyzed.
The main results can be summarized as follows:
Multi-product exporters make up of 72% of all firms, and 99% of total value of exports. On average, they export 22 products, i.e., they are quite diverse. Multi-product firms have higher sales, employment and wages than single product ones. They are also more productive than single-product ones, supporting the theory that suggests that high productivity firms are able to cover fixed costs of producing a larger number of products.
It is also shown that for multi-product exporters, the share of exports is not evenly distributed across products. Exports are mainly skewed towards the main product, which is possibly the product they have a comparative advantage in.
An analysis of extensive and intensive margins shows that 16-28% of variation in sales is due to variations in extensive margins. Furthermore, there is a positive correlation between extensive and intensive margins. Bigger New Zealand exporters produce on average more products and firms that produce more products on average have larger sales per product.
Alteration of product mix is frequent and widespread across different sectors. This is a result of shocks to consumer tastes and firm productivity leading to adding and dropping of products. Dropping a product is more likely than adding a product.
Patterns of product switching are correlated with contemporaneous changes in firm outcomes. Adding products are associated with higher wages, value added, sales and exports, whereas dropping products are associated with lower values for these variables.
There is a positive correlation between product add and drop rates. This suggests that some firms receiving positive demand shocks add products, whereas firms receiving negative shocks drop them and patterns are not solely a result of reallocation of resources across products. Product-switching behaviour is not explained solely by firm or product characteristics, but a combination of firm-product factors. The probability of dropping a product decreases with firm age and size as well as product scale and tenure.
The results show that New Zealand exporters are dominated by multi-product firms, which tend to be higher performing and dynamic in response to changing economic conditions, i.e., they add and drop products. This suggests an economy that is not restricted by tight regulations or bound by high sunk costs. This is similar to US firms and in contrast to Indian firms where regulations are much tighter. This might be an indication that the structural reforms of the 1980s were successful in creating a business environment that allows firms to be dynamic.
The analysis also shows that dropped products tend to have smaller size and age, suggesting that there is a systemic reallocation of resources towards high productivity sectors, and across products within firms and across firms. In this sense, dropped products are not an indication of firms in trouble, but their move towards more productive activities.
In light of these results, the most important question to consider is why exporters choose to span several products and sectors. If this is due to their need to hedge and diversify in order to better handle volatilities in overseas markets and in the domestic economy, such as movements in the exchange rate or capital markets that are not functioning well, then there is room for improvement in the economic environment that exporters operate in. On the other hand, if firms choose to be multi-product exporters because they are able to cover fixed costs for a large range of products, this could be a sign that firms find operating across products and sectors more profitable, i.e. diversifying product mix not only reduces variability of revenue but also increases total revenue. The results show that multi-product exporters “select” themselves into being multi-product exporters and are better performing ex ante as well as ex post, giving support to the second explanation. However, further work that controls for changes in the macroeconomic variables is necessary to draw stronger conclusions.
It would be worth further investigating questions such as: Do different product mixes affect the profitability of firms? Does the product mix of a firm affect the volatility of its earnings? If so, does this affect the volatility of the economy as a whole? Do firms choose to export multiple products to hedge against exchange rate volatility?
