 Hello, my presentation is about agro-processing and horticultural exports from Africa. This is a joint work with Will Martin. He is here to answer all the difficult questions and comments. This figure shows the structure of Africa's exports by broad economic sector. We observe that the share of agriculture is relatively small at around 10 percent and which decreased somewhat for the past two decades. It's a bit surprising because many believe that African countries is a large exporter of agriculture. Within agriculture, we find quite a different picture between the world agriculture and African agriculture. In terms of world agriculture, the share of bulk products decreased from around 25 percent to 17 percent. The share of agro-processed agriculture and horticultural exports expanded by 21 percent to around 71 percent and 12 percent respectively. When we look at the structure of agricultural exports in Africa, the picture is quite different. The share of bulk exports account for still a large amount at around 42 percent, but it decreased from 60 percent. The share of processed agriculture expanded to around 35 percent, but this figure is still just a half of world average. What is interesting is the sharp expansion of horticultural exports, which reached around 22 percent in 2014. Relatively low share of processed agriculture export in total agricultural export in African countries may reflect relatively low level of income in African countries. This figure shows the relationship between the share of processed agriculture in total agriculture in terms of value added. We find close to linear relationship between the per capita and the share of processed agriculture, confirming that as the economy grows, the share of processed agriculture grows. When we look at African countries, protein in green, Africa appears to be following relatively, following broadly the similar pattern. Similarly, we look at the share of processed agriculture export in total agricultural export. We also find the positive relationship, but at the decreasing rate between the per capita income and the share of processed agriculture export. African countries is following broadly similar pattern. Now we turn to harmonized system 6-digit level data, which is the finest data in which internationally comparable data are available. The first column shows the number of exports which are shipped from our 10 African countries. We observe that the structure of African agriculture may be relatively diversified because these countries are shipping many products. However, when we look at the major of concentration, we find quite a different picture. Second column shows the share of top item shipped by each country is relatively large. For instance, for Ghana, just one product. Cocoa beans represent as high as 59% of total agriculture export. The last column shows numbers equivalent for high fundal index, that is our major of diversification. For instance, the number for half index for HOPI7, that means the structure of HOPI and exports is characterized by only 7 items which are equally distributed. The diversification figure for South Africa is 44, that means African economy structure is more diversified, while the number of Ghana is 3, less diversified. Then we look at how much the changing structure of agriculture exports is driven by new products. We construct the major of new products as defined, which were the bottom two export items in the initial year, which are about one decade ago, depending on the country, but which made the top 20 in 2013. We find that the new product, the share of new products are relatively large for some countries. For instance, 47% for Ethiopia and 69% for Rwanda. As we have seen, a small number of country, a small number of items are very important for many African countries. Israel and Russia made the same point, arguing that exports from African countries are dominated by a small number of big hit. Osman and Roderick also argues policies that encourage entrepreneurs to discover successful exports of the future are worthwhile. The case of Ethiopia's cut flowers can be cited as an example of a big hit. Ethiopia's cut flower was just one of bottom 2% items a decade ago, but Ethiopia emerged as the second largest African cut flower export after Kenya. There are a number of key factors. Ethiopia has a comparative advantage with favorable agro-climatic conditions and abundant labor. The success is characterized by private entrepreneurship experimentation supported by government programs such as duty-free import of machinery and tax incentives and credit. Cut flower venture was initially initiated by two domestic entrepreneurs but later attracted large foreign direct investment. They secured market access through international auction market and the right weight of flowers made a transportation viable overcoming Ethiopia's constraint of landlockedness. Now I turn to the implication of policy reforms. The impact of policy reforms is heavily influenced by the destination of Africa's export and the initial protection structure. We see that the largest destination of Africa's export, agriculture export, is EU accounting for 40% followed by Africa accounting for 19%. But when we look at process agriculture, the proportion of export shipped to Africa of 35% is disproportionately large, perhaps reflecting lower fixed cost to ship to Africa market and the existence of local market which is tailored to local consumption pattern and taste. This table shows the Valoran equivalent protection. The first column show that African exports face about 7.0% of protection. Second column shows the protection that Africa imports on its import is at about 12.2%. The fourth column show the exports that Africa face towards against EU market is low at 0.8%, reflecting the preferential access of Africa to EU market. Finally, the protection of intra-Africa trade is relatively large at 10.1%. Especially the protection of process agriculture is large at around 12.6%, reflecting clear tariff escalation pattern within Africa. At the global level, when we look at the data at the more disaggregated level, we also find tariff escalation within the same value change. For instance, party rise of protection 1.2% is smaller than process rise of 5.7%, and the protection of live animals of 1.3% is much lower for the meat product of 33.7%. Finally, using global trade analysis setup model, we conduct a series of simulation analysis. Simulation one experiment what would happen if African trade partners would reduce, would eliminate tariff escalation. The results show that total agriculture would expand by 39%, and the expansion of process agriculture is especially large by 114%. Simulation two, investigate what would happen if Africa loses preferential access to EU market. Africa's total agriculture and process agriculture would decrease by 5.5% and 12%. The larger amount of the reduction of process agriculture appears to reflect larger preference margin in EU market for processed goods. Simulation three, investigate what happens if Africa removes protection within Africa. Africa's export and for total agriculture and process agriculture would expand by 5.1% and by 13% respectively. The larger amount of expansion of process agriculture appears to reflect the elimination of tariff escalation resulting from liberalization. Simulation four, investigate what happens if African countries increased the productivity of processed agriculture goods by 10%. The results show processed agriculture would expand by 30%, highlighting the importance of increasing productivity to tap into export market. Finally, simulation five, investigate what happens if Africa reduced its own protection for both agriculture and non-agriculture. The results shows expansion of export by 7.5%. The positive impact appears to reflect the reduction of intermediate costs resulting from this simulation. We turn to policy questions. Tariff escalation in the same value chains by African trading partners hard agro processing exports. Thus policies imposed by Africa's trading partners matter. Preferential access to EU markets is important. Regional integration can be used to promote further Africa's trade liberalization. In terms of domestic policies, it seems to be important to create an environment without these intensives for export. In particular, it seems to me important for firms to have access to intermediate at world prices. Investment in trade logistics is also important because unbundling of global value change involves the logistics for the transfer of materials and knowledge. Also, costs associated with high transportation costs and inputs may make processing activities un-economical. Horticulture exports are perishable and particularly vulnerable to delays in shipping. We view that good policies lead to more and efficient process agricultural exports. But we view adding values should not be an independent goal because if it is related to tax raw materials, this may lead to value-substracting restrictions on export of raw materials and it may hurt the owner of raw materials who tend to be poor. In conclusion, high-valued agricultural exports could play an important role in increasing overall exports. We observe that Africa has been moving already toward high-valued agricultural products. But to tap fully the potential of high-valued agricultural exports, it is important to encourage experimentation of entrepreneurs to find future big hit while minimizing these intensives to export. At the same time, it is important to invest on trade logistics such as transport and trade facilitation. Finally, we view that seeking policy reforms that also encourage exports of other goods and services is important. Relying on only high-valued agricultural exports is limited because the current share of agricultural exports in total Africa is relatively small at around 10%. Thank you.