Big hits, export concentration and volatility

Abstract

Recent empirical work has documented the high concentration of trade flows, and the large role played by few “big hits” in each country’s export. We propose a simple stochastic benchmark against which we assess each economy’s actual number of “big hits”. We show that most European countries underperform the benchmark, while China, the US and Germany do better. A low number of “big hits” (relative to our prediction) is associated with higher export volatility. Looking at possible determinants of “big hits”, we find they depend on the actual performance of each country, so that industrial policy needs to be country-specific.

Appendices

Appendix 1: List of variables and their description

• Initial income is GDP per capita at PPP (constant 2005 international $) in logarithm for the year 1995 which is the beginning of the period under consideration [source from World Development Indicators (WDI)].

• GDP represents the size of the economy. It is total GDP amount in logarithm for each country at constant US$.

• Country capabilities is calculated as in Hidalgo and Hausmann (2009) , $k_{c,0} = \sum _{p} M_{cp}$ i.e. the diversification of country c as the sum of $M_{cp}$ over all products p where $M_{cp}$ = 1 if country c exports product p with a revealed comparative advantage above 1, $M_{cp}$ = 0 otherwise. We use our own data and average the figure for each country.

• Product ubiquity is also taken from Hidalgo and Hausmann (2009) and represents the pervasiveness of the products of a country, $k_{p,0} = \sum _{c} M_{cp}$.

• Betweenness centrality, following Jackson (2008), is defined as $\sum \nolimits _{ij:i\ne j, k \not \in ij} \frac{P_{ij}(k)}{P_{ij}}$ where $P_{ij}$ denotes the number of shortest paths from node i to j and $P_{ij}(k)$ denotes the number of shortest paths from node i to j that node k lies on. It is a measure of the importance of a country based on its connections.

• Clustering, where the clustering coefficient of a node is the probability that two randomly selected neighbours are connected to each other. The clustering coefficient $C_{c}$ of node i, $C_{c} ({\mathbf {A}}) = \frac{\sum \nolimits _{j\ne i; k\ne j; k\ne i} A_{ji} A_{ik} A_{jk} }{\sum \nolimits _{j\ne i; k\ne j; k\ne i} A_{ji} A_{ik} }$.

• Private credit is credit from all financial institutions and sourced from the Financial Development and Structure Dataset, World Bank developed by Beck et al. (2000). The April 2013 version has been used. The variable is Private Credit by Deposit Money Banks and Other Financial Institutions to GDP (%) which is defined as claims on the private sector by deposit money banks and other financial institutions divided by GDP. As described in Beck and Demirguc-Kunt (2009), this asset side variable captures one of the most important function of financial intermediaries, that is, allocation of credit.

• Tariffs is average tariff rate on manufactured products (%) (source WDI).

• Start-up costs represents costs of business start-up procedures as a percentage of gross national income per capita (source WDI).

• Export docs is simply the number of documents required per shipment to export goods such as documents required for clearance by government ministries, customs authorities, port and container terminal authorities, health and technical control agencies and banks are taken into account. It is an ease of doing business indicator.

• Inflation is used to capture changes in the price level. It reflects annual percentage changes in consumer prices (source WDI).

• Export/GDP is a standard indicator of openness to trade (source WDI).

• Rule of law, percentile rank, is a World Governance Indicator, World Bank.

• Telephones captures technological development (source WDI).

• FDI is net inflows of Foreign Direct Investment in current US$ in logarithm (source WDI).

• ${D^{Oil}}$ is a dummy variable capturing whether a country is a significant oil-producer, that is, its oil rents is more than 40 % of its GDP (source WDI).

• ${D^{SSA}}$, ${D^{SA}}$, ${D^{NA}}$, ${D^{MENA}}$, ${D^{LAC}}$ and ${D^{EAP}}$ are regional dummy variables representing regions as per the WDI 7 regions classification, namely, Sub-Saharan Africa, South Asia, North America, Middle East and North Africa, Latin Ame and Caribbeans, East Asia Pacific.

• Capital stock, investment and savings are capital stocks in billions constant 2005 $US, investment rate as a percentage of GDP, savings rate as a percentage of GDP respectively (source CEPII Econmap baseline database 2.1).

• Pop. is population density (people per sq. km of land area) (source WDI).

• Ki is the number of product–destinations in logarithm.

• Agricultural and industrial empl. represent employment in the agricultural sector and employment in industry (% of total employment) respectively (source WDI).

• Primary school and secondary school are years of primary schooling and secondary schooling for population aged 15 and over respectively (source Barro and Lee 2012).

• Export volatility is the standard deviation in export growth from 1995 to 2005 (calculated using data from BACI database). All the variables are for the year 2005 except where specified above. To maximise the number of observations, where data was missing for the year 2005, averages of previous and successive years were used.

Appendix 2

See Table 8.

Table 8 List of all countries with simulated and empirical BHs, and the actual deviations between simulated and empirical BHs