IntlTrade_2024_Exerc_1_solution_Pdf

L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 Class exercise Set 1 (for Week 2): 1) RCA Index: Against the backdrop of the Ricardo model introduced in Lecture 1, explain what the indicator of revealed comparative advantage (RCA) is supposed to achieve....

L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 Class exercise Set 1 (for Week 2): 1) RCA Index: Against the backdrop of the Ricardo model introduced in Lecture 1, explain what the indicator of revealed comparative advantage (RCA) is supposed to achieve. How is it calculated, and what values can the index take on? There is a steer to get you started on the last page of Lecture 1 slides, but for the three questions on RCA in this exercise sheet you are encouraged to do some independent research. Classical models of trade predict some sort of cross-country specialisation along the lines of comparative advantage. It is natural to look for this prediction to be borne out in the data. Using the notation in the slide on Canvas, RCA = (share of product k in country i's total exports) / (share of product k in total world exports). This juxtaposes an individual country’s export share against the world average, which is a useful benchmark and accounts for the degree of tradability of product k. RCA ranges from [0, ∞). The threshold value is 1, at which a country’s export share is exactly the same as the world average. So where the RCA is > 1 then the country is said to have a revealed comparative advantage in that product, and where the RCA < 1, then conversely the country is said not to have a revealed comparative advantage. There is a variant of the above indicator which is the normalised RCA. The normalised RCA is: RCAnormalised = (RCA-1) / (RCA+1) with a symmetric range of [−1, +1). The normalised RCA is better suited to comparing RCA’s either across time for a given country; or across countries for a given time period. For instance, an RCA = 0.5 (i.e. below 1) translates into a normalised RCA = -1/3. Level of definition for product “k”: if very aggregate, such as ‘manufacturing’, will likely lump together both CA and non-CA goods exports, and thus become an increasingly noisy measure the more aggregated products are. Too disaggregated might also not be ideal since every product becomes a distinct variety at a certain level of extreme disaggregation. 2) Explain the intuition behind the RCA indicator in your own words: why can it be seen as a good measure of a given country’s comparative advantage or disadvantage in a given set of industries? If a given country had a comparative advantage eg. in textiles, you would expect the country to be specialised in textile production and exports. You can see this both in the Ricardian model where you get complete specialisation; and also later in the HOS model as the country moves along its PPF when it opens up to trade. In any case, if the country is specialised in textiles, you would expect the share of textiles in its exports to be higher on average than for those countries that do not have a 1 L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 comparative advantage. Hence you would expect the RCA (which compares those shares) to be greater than 1. Notice that comparative advantage is inherently a bilateral concept whereas the RCA index is a country characteristic. As such, it makes an average statement and may be incorrect for particular trading partners or export products (esp. quality differentiation). 3) Under what circumstances might the RCA indicator be a poor measure of a country’s comparative advantage? The RCA is based on observed trade flow. Thus there may be factors impacting on observed trade flows other than considerations of comparative advantage. Examples of this might include: any kind of trade barriers including tariffs, export subsidies, institutions and infrastructure not reflecting accurately a country’s relative factor endowments. In bilateral RCAs, gravity forces such as distance would also affect the export shares. On demand side: difference in tastes and preferences etc. Value chain trade, whereby imported inputs are augmented with minor value added and then re-exported again, may also give the false impression of comparative advantage, as the RCA index does not reflect the plethora of imported inputs. The large trade deficit in HS product 8542 (integrated circuits), which nonetheless was China’s third most important export product in 2022 (see exercise 4.b. below), is a case in point. 2 L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 4) Stata exercise: First empirical work on specialisation and comparative advantage a. Use the Stata software and the Stata dataset provided along with this exercise sheet on Canvas (Lecture 1 tab) to identify the top-5 products, in terms of HS 4-digit categories1, exported by Brazil, China, the EU27, the UK, and a group called “Sub- Saharan Africa” (SSA), respectively, to the World for the most recent year in the dataset (which is 2022). Have Stata display these results in the output window. b. Create a descriptive output: create a table of the above results which you can export from Stata e.g. into Excel. In addition to containing the reporting country (or country group) as well as the HS product code and description, make sure that the table includes the following entries (ie. columns): value of exports, value of imports, the trade balance (defined as exports minus imports), the share of each product in each country’s total exports, and the rank in terms of export share that you had identified in part (a) above. c. Interpret your output: Looking at the resultant table, in what sense is there evidence of specialisation? We will return to this table and its interpretation in later seminars when we have covered more trade theory and thereby different, alternative sources of comparative advantage. Note: In case you were not able to complete parts (a) and (b) using Stata, below is a Table of top-5 HS 4-digit products from an earlier year 2016. You can interpret that table in terms of specialisation instead so as to be able to complete part (c) of Question 4 regardless of whether or not you got a Stata program to work. Incidentally, that table also offers some guidance of what your Stata generated excel table should look like, at least in terms of contents. The Stata generated table / excel output does not need to include country “Total” rows as in the example table below, as this is a functionality not easy to pull off in Stata. Simply five rows per country (one for each of the top-5 products) will be fully satisfactory. 1 In this dataset, products are identified as items of the Harmonised System (HS) classification of products. HS is a hierarchical system such that longer codes denote more disaggregated products within a shorter code. For instance, all entries at the HS 4-digit level (called “HS headings”) 8703 are part of the HS 2-digit chapter 87. 3 L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 Technical guidance on how to do complete parts (a) and (b)2: — The dataset is called L1070_Exerc1_data.dta. Open Stata and load that dataset into Stata’s memory. — In the raw data, exports and imports are stacked on top of each other in one variable called “flow.” Subsequent data manipulation is easier when you create separate variables for exports and imports, respectively. To do this, use the -reshape wide- command and specify -j(flow) string- in the command’s options. — For calculating the export share, use the -egen- command with the -total(.)- function. Note that total exports have to be calculated within a country and year, not overall. The -bysort varlist:…- command does that for you. — Once you have a country’s exports and total exports in two separate variables, you can directly construct the export share in percentage terms with the -generate- command. — To construct the rank of HS4 products, use again the -egen- command but this time together with the -rank(.), field- option. For instance, this command line will do the trick: bysort reporteriso3 year: egen exp_rank = rank(exp), field — To export the table to Excel, use the -export excel- command. To view it on Stata’s output window, use the -list- command. Coding solution: see do-file uploaded on Canvas that produces the table show below. From that table, it is apparent that developing countries tend to specialise in agricultural produce and raw minerals; this is most visible for the case of Sub-Saharan Africa (SSA) but also for Brazil, which is, which is one of the world’s largest suppliers of Soya Beans (HS 1201, #1), but also Maize (#5) or Iron ores (#3). China is known as ‘the factory of the world’ and hence the set of top-5 products is no surprise as they consist of manufactures that are likely intermediate inputs into final (consumer or capital) goods. For advanced economies, the specialisation on Motor Cars (HS 8703) for the EU27, and prior to Brexit also for the UK, reflects the presence of a large automotive industries in these countries. Similar for Medicaments (#3 for the EU27 and #4 for the UK) as there is a large pharmaceutical industry, which is intensive in both capital and skills. 2 Here I’m essentially giving you a list of Stata commands that form the backbone of a program that would solve this exercise. Stata has a very powerful help function, which allows you to read up on the correct syntax of every command. For instance, to learn how to use the -generate- command, simply type -help gen- into the command prompt and a window will pop up that describes this command. Some or all of these Stata commands may be new to you, but the idea is that by revealing which commands to use, you should be able— with some extra effort—to write a short workable program. 4 L1070 INT’L TRADE (2024/25) SEMINAR WEEK 2 +---------------------------------------------------------------------------------------------------------------------------+ | reportername year HS4_code HS4_desc exp imp trade_~l exp_sh~e exp_rank | |---------------------------------------------------------------------------------------------------------------------------| | Brazil 2022 1201 Soya beans, whether or not broken. 46664 203 46462 14.0 1 | | Brazil 2022 2709 Petroleum oils and oils obtained fr 42688 10145 32543 12.8 2 | | Brazil 2022 2601 Iron ores and concentrates, includi 28889 48 28841 8.6 3 | | Brazil 2022 2710 Petroleum oils and oils obtained fr 13036 24683 -11647 3.9 4 | | Brazil 2022 1005 Maize (corn). 12264 654 11610 3.7 5 | |---------------------------------------------------------------------------------------------------------------------------| | China 2022 8525 Transmission apparatus for radio-te 248409 47396 201012 7.0 1 | | China 2022 8471 Automatic data processing machines 210062 28263 181800 5.9 2 | | China 2022 8542 Electronic integrated circuits and 156305 365932 -209627 4.4 3 | | China 2022 8541 Diodes, transistors and similar sem 65483 21151 44332 1.8 4 | | China 2022 8507 Electric accumulators, including se 57228 3460 53768 1.6 5 | |---------------------------------------------------------------------------------------------------------------------------| | EU27 2022 8703 Motor cars and other motor vehicles 373905 306037 67868 5.4 1 | | EU27 2022 2710 Petroleum oils and oils obtained fr 293249 268063 25186 4.2 2 | | EU27 2022 3004 Medicaments (excluding goods of hea 292596 198762 93834 4.2 3 | | EU27 2022 3002 Human blood; animal blood prepared 212474 146240 66234 3.1 4 | | EU27 2022 8708 Parts and accessories of the motor 180955 173937 7018 2.6 5 | |---------------------------------------------------------------------------------------------------------------------------| | SSA 2022 2709 Petroleum oils and oils obtained fr 40325 4390 35935 19.1 1 | | SSA 2022 7110 Platinum, unwrought or in semi-manu 16872 32 16841 8.0 2 | | SSA 2022 2701 Coal; briquettes, ovoids and simila 15632 820 14812 7.4 3 | | SSA 2022 7102 Diamonds, whether or not worked, bu 13528 3348 10180 6.4 4 | | SSA 2022 7108 Gold (including gold plated with pl 11244 1014 10230 5.3 5 | |---------------------------------------------------------------------------------------------------------------------------| | United Kingdom 2022 7108 Gold (including gold plated with pl 72682 41809 30874 15.4 1 | | United Kingdom 2022 8411 Turbo-jets, turbo-propellers and ot 25163 18713 6450 5.3 2 | | United Kingdom 2022 2709 Petroleum oils and oils obtained fr 25105 39459 -14354 5.3 3 | | United Kingdom 2022 3004 Medicaments (excluding goods of hea 20331 19816 515 4.3 4 | | United Kingdom 2022 2710 Petroleum oils and oils obtained fr 18683 29819 -11136 4.0 5 | +---------------------------------------------------------------------------------------------------------------------------+ 5