Appendix I

Appendix I
Description of Empirical Methodology

Introduction

This appendix describes the methodology used to estimate the level of U.S. exports to Cuba in selected sectors in the event that U.S. restrictions are removed and Cuban barriers are lowered. The main approach used in the empirical analysis is an enhanced gravity model.[1]

The basic gravity model is a widely used economic model that relates trade between two countries to various country characteristics that are accepted to be common determinants of trade, such as distance, having a common language or border, and size (output) of the economies. For example, the basic gravity model predicts, correctly, that larger countries trade more than smaller ones and that countries located closer together trade with each other more than those further apart.[2] The enhanced gravity model implemented here adds important general equilibrium effects of trade policy to the basic structural gravity model.[3] This enhanced gravity model is better suited than the basic model to estimate how the trade restrictions investigated in this study affect U.S. trade flows.

Estimating U.S. exports to Cuba in the event that U.S. restrictions are removed and Cuban barriers are lowered presents a number of challenges that preclude the use of some common types of economic models, including computable general equilibrium (CGE) and partial equilibrium models.[4] First, existing U.S. exports to Cuba are zero in many industries. Second, a tariff equivalent for current U.S. restrictions is not known. Third, it is important to account for the competition that U.S. producers will face from other countries, such as China, when entering the Cuban market.[5] The enhanced gravity model used in this study is able to address these challenges: it allows for increased competition and lower prices in the Cuban market and allows productivity and wages to play a role in determining trade. The key determinants of trade in the model are trade cost, productivity (which determines comparative advantage), and wages.

The model uses trade costs to estimate the value of U.S. exports to Cuba in the event that U.S. restrictions are removed and also in the event that Cuban tariff and nontariff measures are lowered. Trade cost is defined as the difference between the cost of a good at a production source and in its destination market. It is estimated as the equivalent of an ad valorem tariff (a tariff calculated as a percentage of a good’s value) that is measured relative to domestic trade cost.[6]

Total trade costs have been divided into two components: bilateral and importer-specific. Bilateral trade costs include freight, insurance, translation, preferential tariffs, and trade restrictions, such as the U.S. restrictions on trade with and travel to Cuba (table I.1). Importer-specific trade costs are costs specific to the importing country that apply equally to all exporters, including normal trade relations (NTR) tariffs[7] and nontariff barriers such as costs due to poor infrastructure, corruption, customs procedures, and sanitary and phytosanitary measures. There are no data to directly measure all of these trade costs. However, aggregate bilateral and importer-specific trade costs can be estimated based on observed trade flows.

The other two determinants of trade in the model are productivities and wages. Relative productivities across industries determine comparative advantages. For example, the United States has comparative advantages in medical equipment, wheat, and poultry (among other industries). Wages affect trade because lower wages benefit the production of labor-intensive goods. These are held constant in this model.

Aggregate bilateral and importer-specific trade costs for each pair of countries in each industry are estimated using data on current trade between all countries as well as data on domestic trade (i.e., purchases of domestically made goods). Bilateral trade costs are related to observable country characteristics, making it possible to estimate what trade costs would be in the absence of U.S. restrictions.[8]

Table I.1: Relationships between selected unobservable trade costs and observable country characteristics

Cost (unobservable)	Observable characteristic
Freight	Distance, shared border
Insurance	Distance, shared border
Translation	Commonality of language
Preferential tariffs	Trade agreements
U.S. restrictions	Residual

Source: U.S. International Trade Commission.

To respond to the original request letter, the Commission’s analysis first calculates U.S.-Cuba trade costs in the absence of U.S. restrictions using current trade flows. It then uses an extended gravity model to estimate trade between all countries in the absence of U.S. restrictions. To answer the expanded request letter, the Commission’s analysis lowers Cuba-specific trade costs to the calculated average of the developing countries. It then uses an extended gravity model to estimate trade between all pairs of countries in the absence of U.S. restrictions and with lower Cuban tariff and nontariff barriers. An alternative scenario estimates the effects of reducing Cuban barriers to the level of the developing country with the lowest barrier in each sector.

Methodology

Estimation of U.S. exports to Cuba in the selected sectors in the absence of U.S. restrictions proceeds in several steps. In the first step, parameters of the gravity model, including trade costs, are estimated. In the second step, the potential value of U.S. exports to Cuba in each selected sector is estimated. Both steps use the extended gravity model $—$ a basic gravity model supplemented with elements of a general equilibrium model. This model is based on the model developed by Eaton and Kortum (2002), extended to the industry dimension in Shikher (2012).[9]

The industry-level Eaton-Kortum model has been used in a large number of studies, including publications in leading academic journals. The model has been used to study the effects of various determinants of trade and to estimate the effects of past and future trade policies. For example, Chor (2010) and Shikher (2011) use the model to study the effects of various determinants of trade, such as capital and institutions.[10] Caliendo and Parro (2015) use this model to estimate the effects of the North American Free Trade Agreement (NAFTA) on the U.S. economy; Shikher (2012), to evaluate how accurately this model could predict the effects of NAFTA from the point of view of 1989; and Levchenko and Zhang (2012), to estimate the effects of European integration.[11] Major papers that use this model have been recently reviewed in Eaton and Kortum (2014).[12]

Estimation of the Parameters of the Gravity Model

The approach used in estimating the parameters of the gravity model recognizes that bilateral trade costs are not necessarily symmetric. Rather, the cost of exporting from country A to country B can be different from the cost of exporting goods from country B to country A. Following the standard methodology of the gravity literature, trade costs are given by the trade cost function (equation 1).

International trade costs are denoted by $d_{n i j}$ where $j$ is the industry, $i$ is the country of origin, and $n$ is the country of destination.
$\log d_{n i j} = d_{k n i j}^{p h y s} + b_{n i j} + l_{n i j} + f_{n i j} + m_{n j} + δ_{n i j}$ ( 1)

The right-hand side of equation (1) consists of the variables that affect international trade costs. The first variable is the physical distance, $d_{k n i j}^{p h y s}$ . Given evidence in the literature that the effects of distance are nonlinear, the cost of moving goods an additional kilometer depends on the total distance traveled. Each $d_{k n i j}^{p h y s}$ is the effect of the physical distance lying in the kth interval. Here, distance is divided into 6 $(k = 1,…,6)$ intervals (in km): [0,599], [600,1199], [1200,2399], [2400,4799], [4800,9599], and [9600,maximum]. The distance between the U.S. and Cuba is in the second interval.[13]

The second, third, and fourth variables on the right-hand side of (1) are the effects of having a shared border, common language, and common free trade agreement (FTA). Sharing any of these things reduces trade costs. Subscript $j$ on these variables indicates that they are industry-specific because the effects on trade costs of a shared border, common language, or FTA can be different across industries.

The fifth variable on the right-hand side of (1), $m_{n j}$ , is the international trade cost that is specific to the importer $n$ and common to all sources $i$ . This trade cost includes non-preferential tariffs, nontariff measures, and costs due to corruption, security, and poor infrastructure in country $n$ . All exporting countries selling in destination $n$ face these costs equally.

The last variable on the right-hand side of (1) includes all the bilateral determinants of trade costs not accounted for by the other variables on the right-hand side and differs according to country pair and direction of trade.

As is common in the gravity literature, trade costs are measured relative to domestic trade costs. Therefore, domestic trade costs $d_{n n j}$ are equal to 1 and $\log d_{n n j} = 0$ . If country $n$ has high domestic trade costs (as many developing countries do), then international trade costs for that country may be relatively low.

A gravity equation is used to estimate international trade costs. This gravity equation is derived from theory, with the full derivation presented in Eaton and Kortum (2002) and Shikher (2012). The basic form of the gravity equation is
$X_{n i j} = \frac{S_{n j}^{i m p} S_{i j}^{e x p}}{d_{n i j}^{θ}}$ (2)

where $X_{n i j}$ is the spending by country n on goods of industry j from country i (i.e., the volume of trade in industry j from country i to country n). $S_{n j}^{i m p}$ includes all determinants of trade that are specific to the importer, while $S_{i j}^{e x p}$ includes all determinants of trade that are specific to the exporter. In older gravity literature, $S_{n j}^{i m p}$ and $S_{i j}^{e x p}$ were often importer and exporter GDPs. However, recent literature, such as Anderson and van Wincoop (2003) and Head and Mayer (2014), explains that exporter- and importer-specific variables also include other determinants of trade, such as output and spending.[14]

The denominator of (2) includes the trade cost $d_{n i j}$ raised to the power $- θ$ . The parameter $θ$ is usually called the elasticity of trade with respect to trade cost.[15] Taking logs of equation (2) gives the log-linear form of the gravity equation
$\log X_{n i j} = \log S_{n j}^{i m p} + \log S_{i j}^{e x p} - θ \log d_{n i j}$ (3)

The above equation can also be used to describe the domestic trade in country $n$ , i.e., the amount of goods that $n$ buys from itself. The domestic trade is noted by $X_{n n j}$ and calculated from output and export data as $X_{n n j} = Q_{n j} - E X_{n j}$ , where $Q_{n j}$ is the total output in industry $j$ of country $n$ and $E X_{n j}$ are total exports of industry $j$ in country $n$ .

Some manipulation of equation (3) yields the estimating equation:[16]
$\log X_{n i j} = φ_{n j}^{i m p} D_{n j}^{i m p} + φ_{i j}^{e x p} D_{i j}^{e x p} + γ_{d k j} d_{k j}^{p h y s} + γ_{b j} b_{j} + γ_{l j} l_{j} + γ_{f j} f_{j} + ε_{n i j}$ (4)

using the definitions $φ_{i j}^{e x p} D_{i j}^{e x p} = \log S_{i j}^{e x p}$ , $φ_{n j}^{i m p} D_{n j}^{i m p} = - \log S_{n j}^{e x p} + \log X_{n n j} - θ m_{n j}$ , and $ε_{n i j} = - θ δ_{n i j}$ . In equation (4), $φ$ ’s and $γ$ ’s are the coefficients to be estimated. All the variables on the right-hand side of (4) are fixed effects (dummy variables). $D_{n j}^{i m p}$ is equal to 1 if the importer is $n$ and zero otherwise. $D_{i j}^{e x p}$ is equal to 1 if the exporter is $i$ and zero otherwise. Country fixed effects $D_{n j}^{i m p}$ and $D_{i j}^{e x p}$ are measured relative to the United States. In other words, $D_{us, j}^{i m p} = D_{us, j}^{e x p} = 0$ .

Variable $d_{k j}^{p h y s}$ is equal to 1 if the physical distance between $n$ and $i$ lies in the interval k. Similarly, $b_{j}, l_{j}, and f_{j}$ indicate if $n$ and $i$ share a common border, language, or FTA, as discussed above. Equation (4) is estimated for each industry $j$ using data on bilateral trade, distance, shared borders, language, and FTAs. Estimating (4) will produce estimated coefficients $γ$ and $φ$ and error terms $ε_{n i j}$ .[17]

Equation (4) is estimated using ordinary least squares. When estimating (4), missing and zero trade values are excluded, since the log of zero is not defined.[18]

With equation (4) estimated, the next step is to calculate importer-specific trade costs $m_{n j}$ . They are calculated as follows. Rearranging the definition of $D_{n j}^{i m p}$ yields $m_{n j} = (1 / θ) (\log X_{n n j} - D_{n j}^{i m p} - \log S_{n j}^{e x p})$ . Using the definition of $D_{i j}^{e x p}$ and adding coefficients gives the result
$m_{n j} = (1 / θ) (\log X_{n n j} - φ_{n j}^{i m p} D_{n j}^{i m p} - φ_{n j}^{e x p} D_{n j}^{e x p})$ (5)

Variable $X_{n n j}$ is obtained from data as explained above. Coefficients $φ_{n j}^{i m p}$ and $φ_{n j}^{e x p}$ are estimated using equation (4). The value of parameter $θ$ is taken from Eaton and Kortum (2002), where it is estimated to be 8.28. Sensitivity analysis for a range of plausible values for this parameter shows that this choice has only a small effect on the results.[19]

Now it is possible to calculate international trade costs $d_{n i j}$ for each pair of countries as follows:
$\log d_{n i j} = - (1 / θ) (γ_{d k j} d_{k j}^{p h y s} + γ_{b j} b_{j} + γ_{l j} l_{j} + γ_{f j} f_{j}) + m_{n j} + δ_{n i j}$ (6)

where $δ_{n i j} = - (1 / θ) ε_{n i j}$ . Current trade costs $d_{n i j}$ can be calculated using (6) for all $n$ , $i$ , and $j$ for which there are current trade data. In the absence of current trade, the term $δ_{n i j}$ cannot be estimated. This term includes all trade costs not accounted for by physical distance, shared border, language, and FTA. For most pairs of countries, this term is small.[20] For current U.S.-Cuba trade, in the sectors where trade occurs, this term is high because it includes the effects of U.S. restrictions. The next section will lay out the steps for estimating trade costs in the absence of U.S. restrictions.

Producing Counterfactuals

The second step produces estimates for U.S. exports to Cuba under two counterfactual scenarios: (1) no U.S. restrictions and (2) no U.S. restrictions and lower Cuban import barriers. To obtain trade values under the counterfactuals, it is first necessary to estimate trade costs between the United States and Cuba under the two scenarios. These trade costs can then be used to estimate trade between the United States and Cuba under the two scenarios, assuming trade will be determined by normal market forces.

Estimating Trade Costs in the Absence of U.S. Restrictions

The estimation of trade costs between the United States and Cuba in the absence of U.S. restrictions is based on equation (6). It is assumed that in the absence of U.S. restrictions, the cost of international trade between the United States and Cuba is equal to the typical cost of international trade between countries whose relation to Cuba in terms of physical distance, commonality of language, and border is similar to that of the United States to Cuba, and that do not have an FTA with Cuba. In addition, it is assumed that importer-specific barriers $m_{n j}$ remain the same with or without U.S. restrictions. As noted above, these barriers include tariffs, nontariff measures, costs due to corruption, weak institutions, security issues, and poor infrastructure in country $n$ , which for convenience are referred to here as country-specific “import barriers.” The international trade cost for U.S. exports to Cuba includes the Cuba-specific trade cost $m_{Cuba, j}$ . This cost is imposed on exports from all countries to Cuba, and it is assumed that they are constant and do not change when U.S. restrictions are removed.

Trade costs between the United States and Cuba in the absence of U.S. restrictions, $d_{Cuba, US, j}$ and $d_{US, Cuba, j}$ , are calculated using the following equation:
$\log d_{n i j} = - (1 / θ) (γ_{d k j} d_{k j}^{p h y s} + γ_{b j} b_{j} + γ_{l j} l_{j} + γ_{f j} f_{j}) + m_{n j}$ (7)

Estimating Trade Costs in the Absence of U.S. Restrictions and with Lower Cuban Import Barriers

Cuban import barriers are measured by $m_{Cuba, j}$ . These barriers may be different in different industries $j$ . Any country exporting to Cuba has to pay this cost to get its product to the Cuban market. These trade costs include Cuban tariffs, nontariff measures, and other country-specific trade impediments.

Cuban barriers to imports are relatively high, as shown below (tables I.8 and I.9). The first step here is to reduce Cuban trade barriers in each industry to an average of all developing countries in the dataset. That is, $m_{Cuba, j}$ is set equal to ${\bar{m}}_{Dev, j} = \frac{1}{N_{Dev}} \sum_{n = 1}^{N_{Dev}} m_{n j}$ , where the summation is over the countries classified as low-income or middle-income by the World Bank and $N_{Dev}$ is the number of such countries in the dataset.[21]

At this point, $\log d_{Cuba, i, j}$ is recalculated for all sources of Cuban imports $i$ . For all exporters except the United States, equation (6) is used. For the United States, equation (7) is used.

In addition, a sensitivity analysis is performed to examine the effects of reducing Cuban import barriers to other levels. As an alternative to setting $m_{Cuba, j}$ equal to the average of developing countries ${\bar{m}}_{Dev, j}$ , Cuba’s importer-specific costs are set equal to the minimum across developing countries in each industry, $m_{Dev, j}^{min} = {min}_{n \in Dev} m_{n j}$ . The results of this simulation, presented later in this appendix, should be taken with caution, because these results set Cuba-specific trade costs $m_{Cuba, j}$ to very low values. It would take many years for Cuba to achieve such low trade costs, at which point many parameters that are held constant in the model, especially Cuban productivity, would probably be different from their current values.

U.S. Exports to Cuba with New Trade Costs

Having obtained trade costs in the absence of U.S. restrictions, or in the absence of U.S. restrictions and with lower Cuban import barriers, U.S. exports to Cuba could have been estimated under these trade costs using equation (3). However, doing so would assume that country-specific variables $S$ will remain constant when U.S. restrictions are removed. The gravity literature indicates that these variables include information on prices, which will change when U.S. restrictions are removed.[22] Therefore, to more accurately estimate U.S.-Cuba trade, the model needs to be extended in order to explain how variables $S$ are determined.[23] To do that, the multicountry Ricardian model of Eaton and Kortum (2002) is used and extended to the industry dimension, as in Shikher (2012).

Starting with equation (3), the following expression is derived by setting $i = n$ and subtracting the resulting equation from (3):
$\log X_{n i j} = \log S_{i j}^{e x p} - \log S_{n j}^{e x p} + \log X_{n n j} - θ \log d_{n i j}$ (8)

Eaton and Kortum (2002) derive the expression for $S_{i j}^{e x p}$ as a function of the costs of production in industry $j$ of country $i$ , $c_{i j}$ , and the productivity of industry $j$ of country $i$ , $A_{i j}$ : $S_{i j}^{e x p} = {(c_{i j} / A_{i j})}^{- θ}$ .[24] Plugging the expression for $S_{i j}^{e x p}$ into equation (8) gives
$\log X_{n i j} = - θ \log d_{n i j} + \log {(\frac{c_{i j}}{A_{i j}})}^{- θ} - \log {(\frac{c_{n j}}{A_{n j}})}^{- θ} + \log X_{n n j}$ (9)

The cost of production is given by the Cobb-Douglas cost function: $c_{i j} = w_{i}^{β} ρ_{i j}^{1 - β}$ , where $w_{i}$ is the wage in country $i$ , $ρ_{i j}$ is the cost of intermediate goods in industry $j$ of country $i$ , and $β$ is the share of labor in output. The cost of intermediate goods is a Cobb-Douglas composite of goods from all industries: $ρ_{i j} = \prod_{m} p_{i m}^{η_{j m}}$ , where $η_{j m}$ is the share of industry m in costs of industry $j$ .

Costs of production are obtained by solving the following equation:
$\log c_{i j} = β_{j} \log w_{i} - \frac{1 - β_{j}}{θ} \sum_{m = 1}^{J - 1} (η_{j m} \log (\sum_{n = 1}^{N} d_{i n m}^{- θ} {(\frac{c_{i j}}{A_{i j}})}^{- θ}))$ (10)

where J is the number of industries and N is the number of countries. This expression is derived in Eaton and Kortum (2002) and extended to multiple industries in Shikher (2012). Solving for costs requires simultaneously solving N x J equations. Note that costs of production depend on trade costs d.

Once costs of production in the absence of U.S. restrictions are obtained, the next step is to calculate the new $\log X_{n n j}$ . First, the following must be calculated:
$\log X_{n i j} - \log X_{n n j} = - θ \log d_{n i j} + \log {(\frac{c_{i j}}{A_{i j}})}^{- θ} - \log {(\frac{c_{n j}}{A_{n j}})}^{- θ}$ (11)

From $X_{n i j} / X_{n n j}$ , the term ${IM}_{n j} / X_{n n j}$ can be calculated, where IM are imports. From ${IM}_{n j} / X_{n n j}$ , the term $X_{n j} / X_{n n j}$ can be calculated, where $X_{n j}$ is total spending on good $j$ in $n$ . $X_{n j}$ is calculated as labor income plus spending on intermediate goods. Once new $\log X_{n n j}$ is calculated, new trade is calculated using (9).

Data

The model requires data on trade and production in each industry for a broad set of countries. The base year for the analysis uses the average of data from 2010 $-$ 13 in order to maximize data availability and minimize data measurement errors. One challenge that needed to be overcome was obtaining reliable Cuban production data by industry.

Data on bilateral trade come from the UN’s Comtrade database. Data on agricultural production come from the FAO database, while data on manufacturing production come from the IndStat2 database maintained by UNIDO. All data use the same industrial classification and currency units for consistency. The data sources are summarized in table I.2.

Table I.2: Data sources

Data source	Description of data
Comtrade	Bilateral trade flows by industry
IndStat2	Manufacturing output by industry
FAOSTAT	Value of agricultural production by industry
Cuban statistical series - industry	Index of Cuban manufacturing output
CEPII gravity database	Gravity variables: distance, language, border

The analysis of the agricultural data was at the level of FAO’s industrial classification. The analysis of the manufacturing sector was done at the ISIC rev.3 two-digit industry level. Table I.3 describes ISIC rev. 3 industries. The Comtrade data on agricultural goods were concorded to the FAO industrial classification, while FAO’s production data are already reported using this classification.[25] The Comtrade data on manufactured goods were concorded to the ISIC rev.3 classification, while IndStat’s production data are already reported in the database using this classification.[26]

Table I.3: ISIC sector descriptions

ISIC	Sector name	Contains
15	Food and beverages	Production, processing and preservation of meat, fish, fruit, vegetables, oils and fats, Manufacture of dairy products, Manufacture of grain mill products, starches and starch products, and prepared animal feeds, Manufacture of other food products, Manufacture of beverages
16	Tobacco products	Manufacture of tobacco products
17	Textiles	Spinning, weaving and finishing of textiles; Manufacture of other textiles; Manufacture of knitted and crocheted fabrics and articles
18	Wearing apparel, fur	Manufacture of wearing apparel, except fur apparel; Dressing and dyeing of fur, manufacture of articles of fur
19	Leather products	Tanning and dressing of leather; manufacture of luggage, handbags, saddlery and harness; Manufacture of footwear
20	Wood products (excl. furniture)	Sawmilling and planing of wood; Manufacture of products of wood, cork, straw and plaiting materials
21	Paper and paper products	Manufacture of paper and paper products
22	Printing and publishing	Publishing; Printing and service activities related to printing; Reproduction of recorded media
23	Petroleum products	Manufacture of coke oven products; Manufacture of refined petroleum products; Processing of nuclear fuel
24	Chemicals and chemical products	Manufacture of basic chemicals; Manufacture of other chemical products; Manufacture of man-made fibres
25	Rubber and plastics products	Manufacture of rubber products; Manufacture of plastics products
26	Non-metallic mineral products	Manufacture of glass and glass products; Manufacture of non-metallic mineral products n.e.c.
27	Basic metals	Manufacture of basic iron and steel; Manufacture of basic precious and non-ferrous metals; Casting of metals
28	Fabricated metal products	Manufacture of structural metal products, tanks, reservoirs and steam generators; Manufacture of other fabricated metal products, metal working service activities
29	Other machinery and equipment	Manufacture of general purpose machinery; Manufacture of special purpose machinery; Manufacture of domestic appliances n.e.c.
30	Office and computing machinery	Manufacture of office, accounting and computing machinery
31	Electrical machinery and apparatus	Manufacture of electric motors, generators and transformers; Manufacture of electricity distribution and control apparatus; Manufacture of insulated wire and cable; Manufacture of accumulators, primary cells and primary batteries; Manufacture of electric lamps and lighting equipment; Manufacture of other electrical equipment n.e.c.
32	Communication equipment	Manufacture of electronic valves and tubes and other electronic components; Manufacture of television and radio transmitters and apparatus for line telephony and line telegraphy; Manufacture of television and radio receivers, sound or video recording or reproducing apparatus, and associated goods
33	Medical and precision instruments	Manufacture of medical appliances and instruments and appliances for measuring, checking, testing, navigating and other purposes, except optical instruments; Manufacture of optical instruments and photographic equipment; Manufacture of watches and clocks
34	Motor vehicles, trailers	Manufacture of motor vehicles; Manufacture of bodies (coachwork) for motor vehicles; manufacture of trailers and semi-trailers; Manufacture of parts and accessories for motor vehicles and their engines
35	Other transport equipment	Building and repairing of ships and boats; Manufacture of railway and tramway locomotives and rolling stock; Manufacture of aircraft and spacecraft; Manufacture of transport equipment n.e.c.
36	Furniture; other manufacturing	Manufacture of furniture; Manufacturing n.e.c.

Source: United Nations Statistics Division, Detailed structure and explanatory notes (accessed February 9, 2016).http://unstats.un.org/unsd/cr/registry/regcst.asp?Cl=2.

Note: "N.e.c." stands for "not elsewhere classified."

The IndStat database does not contain recent production data for Cuba. For this study, Cuban production values for all manufacturing industries were obtained by combining 1989 production values from IndStat2 with the 1989 $-$ 2012 index of industrial production from the Oficina Nacional de Estadísticas e Información (ONEI), Cuba’s statistical agency.[27]

ONEI does not report current values of output for all Cuban manufacturing industries. However, it reports output for several industries, which was used to cross-check the output values obtained from the production index. Cuban production values were also checked by industry analysts.

Cuban data can be reported in one of the two currencies: the Cuban national peso or the Cuban convertible peso. Great care was taken to use the correct exchange rate to USD for each Cuban currency.[28] Cuban output data in agriculture were taken from FAO (table I.4). Industry experts verified the value of Cuban output in agriculture by industry. Cuban manufacturing production data used in the Commission’s analysis are reported in table I.5.

Table I.4: Cuban output in selected agricultural sectors, base year (million U.S. dollars)

Sector name	Output
Wheat	0.0
Rice	171.9
Corn	51.6
Pulses	54.6
Other oilseeds	11.9
Soybeans	0.0
Beef	192.5
Poultry	53.7
Pork	297.1
Total	833.0

Source: FAO, Value of Agricultural Production database; USITC estimates.

Table I.5: Cuban output in manufacturing sectors, base year (million dollars)

ISIC		Sector name	Output
15		Food and beverages	3,910.0
16		Tobacco products	542.3
17		Textiles	37.1
18		Wearing apparel, fur	76.1
19		Leather products	29.9
20		Wood products (excl. furniture)	15.3
21		Paper and paper products	16.5
22		Printing and publishing	46.9
23		Petroleum products	442.1
24		Chemicals and chemical products	1,327.8
25		Rubber and plastics products	107.7
26		Non-metallic mineral products	99.1
27		Basic metals	722.0
28		Fabricated metal products	61.5
29		Other machinery and equipment	327.6
30		Office and computing machinery	12.7
31		Electrical machinery and apparatus	63.4
32		Communication equipment	21.3
33		Medical and precision instruments	70.0
34		Motor vehicles, trailers	48.2
35		Other transport equipment	153.1
36		Furniture; other manufacturing	1,469.5
	Total		9,600.0

Sources: UNIDO; ONEI, Anuario Estadístico de Cuba 2013 [Statistical Yearbook of Cuba 2013], 2014; USITC estimates.

Output values in some industries and countries were missing in IndStat2 and FAO databases. Missing output values were estimated.[29] Note that the values of output are not used in estimating gravity equation (4). In fact, only U.S. and Cuban output values affect estimated U.S. exports to Cuba in the event that U.S. restrictions are removed.[30] Values of output of countries other than the U.S. and Cuba are used when calculating lower Cuban trade barriers $m_{Cuba, j}$ . Two different estimates of lower Cuban trade barriers are provided for robustness, as previously described.

In addition to trade and output, the model required standard gravity variables, such as distance between countries, commonality of border and language, etc. These variables were obtained from the database maintained by CEPII. However, because this dataset only contains information through 2006, information was incorporated on more recent trade agreements to update it through 2010.[31] This update was carried out using the World Trade Organization’s list of regional trade agreements.[32]

Twenty-two manufacturing industries and 9 agricultural industries are analyzed in the model. The level of industry detail and the number of industries used in the analysis are determined by data availability. As mentioned before, data required to analyze the services and mining sectors are unavailable, so those sectors are omitted from the analysis. The list of manufacturing industries covers the whole manufacturing sector, while the nine agricultural industries represent a subset of all agricultural products. The agricultural sectors include all agricultural products that the United States is expected to export to Cuba in significant quantities, based on overall U.S. export patterns and analyst projections.

Note that the manufactured (processed at a plant) food products are classified in the ISIC 15 industry (“food products”). For example, powdered milk is included in that industry. ISIC 15 also includes meats, since they are processed at plants. However, it proved possible to obtain data on trade and production of meat products separately, so meat industries are also analyzed separately as agricultural industries. However, given that there is some overlap between agricultural and manufacturing industries (specifically ISIC 15), total agricultural exports should not be added to total manufacturing exports because that would result in double-counting.[33] There are 47 countries in the dataset, shown in table I.6.

Table I.6: Countries included in the model

Country Name
Algeria	Germany	Portugal
Argentina	Greece	Romania
Australia	Hungary	Russia
Austria	India	Slovakia
Belgium	Indonesia	Spain
Brazil	Ireland	Sweden
Bulgaria	Italy	Switzerland
Canada	Japan	Taiwan
China	Korea, South	Thailand
Cuba	Lithuania	Turkey
Czech Republic	Malaysia	Ukraine
Denmark	Mexico	United Kingdom
Dominican Republic	Netherlands	United States
Egypt	Norway	Venezuela
Finland	Philippines	Vietnam
France	Poland

Results

First, the gravity equation (4) is estimated. The results are presented in tables I.7 and I.8. It is clear that most of the coefficients are statistically significant. Distance has a negative effect on trade, as anticipated. Greater distance has a greater negative effect on trade. The effects of distance are measured relative to the first distance interval (0 $-$ 600 km) and are nonlinear, as the cost of transporting goods an additional kilometer declines with distance.

The effects of distance are measured in ad valorem terms (i.e., as a percentage of value), so goods that are most expensive relative to their size and weight have lower ad valorem costs related to distance. For example, medical equipment is very expensive and has low ad valorem costs related to distance.

The effects of sharing a common border, sharing a common language, and belonging to the same FTA are generally positive, as expected, although having a common language is not a significant determinant of trade in agricultural sectors.

Table I.7: Gravity regression results for selected agricultural sectors

	Distance (600 $-$ 1,199 km)	Distance (1,200 $-$ 2,399 km)	Distance (2,400 $-$ 4,799 km)	Distance (4,800 $-$ 9,599 km)	Distance (9,600+ km)	Contiguity	Common official or primary language	Trade agreement in force	Observations	R-squared
Wheat	-2.273*** (0.377)	-3.663*** (0.422)	-3.749*** (0.533)	-4.664*** (0.551)	-4.631*** (0.619)	1.042*** (0.319)	-0.186 (0.333)	1.548*** (0.316)	852	0.640
Rice	-1.506*** (0.317)	-2.411*** (0.349)	-3.979*** (0.408)	-4.782*** (0.421)	-5.455*** (0.462)	0.896*** (0.260)	0.141 (0.250)	0.380* (0.225)	1,074	0.611
Corn	-1.927*** (0.347)	-3.165*** (0.386)	-3.307*** (0.464)	-5.042*** (0.479)	-5.757*** (0.518)	0.911*** (0.278)	-0.0855 (0.289)	0.837*** (0.261)	959	0.654
Beans and pulses	-1.079*** (0.266)	-2.260*** (0.287)	-2.677*** (0.329	-4.128*** (0.330)	-4.280*** (0.362)	0.905*** (0.214)	0.870*** (0.198)	0.0728 (0.169)	1,315	0.646
Other oilseeds	-1.512*** (0.286)	-2.713*** (0.303)	-3.466*** (0.345)	-5.346*** (0.342)	-5.631*** (0.370)	0.864*** (0.226)	0.180 (0.204)	0.00116 (0.164)	1,623	0.669
Soybeans	-1.899*** (0.393)	-3.545*** (0.447)	-4.743*** (0.561)	-4.763*** (0.557)	-4.216*** (0.614)	1.092*** (0.328)	-0.179 (0.333)	-0.192 (0.326)	751	0.625
Beef	-1.204*** (0.350)	-1.958*** (0.384)	-3.044*** (0.466)	-3.762*** (0.522)	-3.496*** (0.575)	1.314*** (0.302)	0.209 (0.309)	2.464*** (0.321)	900	0.631
Poultry	-1.539*** (0.319)	-2.384*** (0.347)	-3.688*** (0.409)	-4.098*** (0.434)	-3.755*** (0.475)	1.276*** (0.262)	0.154 (0.271)	0.751*** (0.257)	1,107	0.592
Pork	-0.940*** (0.326)	-1.492*** (0.357)	-2.244*** 0.440)	-2.211*** (0.478)	-1.751*** (0.544)	1.854*** (0.278)	-0.459 (0.286)	1.794*** (0.306)	922	0.674

Distance (600 $-$
1,199 km)

Distance (1,200 $-$ 2,399 km)

Distance (2,400 $-$ 4,799 km)

Distance (4,800 $-$ 9,599 km)

Distance (9,600+ km)

Contiguity

Common official or primary language

Trade agreement in force

Observations

R-squared

Wheat

-2.273***

(0.377)

-3.663***

(0.422)

-3.749***

(0.533)

-4.664***

(0.551)

-4.631***

(0.619)

1.042***

(0.319)

-0.186

(0.333)

1.548***

(0.316)

852

0.640

Rice

-1.506***

(0.317)

-2.411***

(0.349)

-3.979***

(0.408)

-4.782***

(0.421)

-5.455***

(0.462)

0.896***

(0.260)

0.141

(0.250)

0.380*

(0.225)

1,074

0.611

Corn

-1.927***

(0.347)

-3.165***

(0.386)

-3.307***

(0.464)

-5.042***

(0.479)

-5.757***

(0.518)

0.911***

(0.278)

-0.0855

(0.289)

0.837***

(0.261)

959

0.654

Beans and pulses

-1.079***

(0.266)

-2.260***

(0.287)

-2.677***

(0.329

-4.128***

(0.330)

-4.280***

(0.362)

0.905***

(0.214)

0.870***

(0.198)

0.0728

(0.169)

1,315

0.646

Other oilseeds

-1.512***

(0.286)

-2.713***

(0.303)

-3.466***

(0.345)

-5.346***

(0.342)

-5.631***

(0.370)

0.864***

(0.226)

0.180

(0.204)

0.00116

(0.164)

1,623

0.669

Soybeans

-1.899***

(0.393)

-3.545***

(0.447)

-4.743***

(0.561)

-4.763***

(0.557)

-4.216***

(0.614)

1.092***

(0.328)

-0.179

(0.333)

-0.192

(0.326)

751

0.625

Beef

-1.204***

(0.350)

-1.958***

(0.384)

-3.044***

(0.466)

-3.762***

(0.522)

-3.496***

(0.575)

1.314***

(0.302)

0.209

(0.309)

2.464***

(0.321)

900

0.631

Poultry

-1.539***

(0.319)

-2.384***

(0.347)

-3.688***

(0.409)

-4.098***

(0.434)

-3.755***

(0.475)

1.276***

(0.262)

0.154

(0.271)

0.751***

(0.257)

1,107

0.592

Pork

-0.940***

(0.326)

-1.492***

(0.357)

-2.244***

0.440)

-2.211***

(0.478)

-1.751***

(0.544)

1.854***

(0.278)

-0.459

(0.286)

1.794***

(0.306)

922

0.674

Source: USITC estimates.

Note: *** p<0.01, ** p<0.05, * p<0.1; standard errors in parentheses.

Table I.8: Gravity regression results for manufactured goods sectors

	Distance (600 $-$ 1,199 km)	Distance (1,200 $-$ 2,399 km)	Distance (2,400 $-$ 4,799 km)	Distance (4,800 $-$ 9,599 km)	Distance (9,600+ km)	Contiguity	Common official or primary language	Trade agreement in force	Observations	R-squared
Food and beverages	-1.147*** (0.194)	-2.068*** (0.202)	-2.662*** (0.223)	-3.916*** (0.221)	-4.355*** (0.236)	0.619*** (0.151)	0.636*** (0.125)	0.290*** (0.0953)	2,132	0.810
Tobacco products	-1.389*** (0.343)	-2.111*** (0.363	-2.769*** (0.411)	-4.136*** (0.406)	-4.553*** (0.440)	0.545** (0.272)	0.0347 (0.237)	1.314*** (0.197)	1,499	0.539
Textiles	-0.640*** (0.178)	-1.405*** (0.186)	-2.024*** (0.206)	-3.822*** (0.204)	-4.021*** (0.219)	0.704*** (0.139	0.636*** (0.117)	0.127 (0.0896)	2,060	0.854
Wearing apparel, fur	-0.871*** (0.201)	-1.636*** (0.210)	-2.277*** (0.232	-3.839*** (0.230)	-3.822*** (0.246)	0.761*** (0.157)	0.527*** (0.131)	0.113 (0.100)	2,077	0.856
Leather products	-1.048*** (0.227)	-1.957*** (0.237)	-2.710*** (0.263	-3.942*** (0.261)	-3.955*** (0.279)	0.866*** (0.177)	0.197 (0.150)	0.198* (0.115)	2,036	0.800
Wood products (excl. furniture)	-0.855*** (0.223)	-2.036*** (0.232)	-2.864*** (0.258)	-4.237*** (0.256)	-4.766*** (0.274)	1.014*** (0.174)	0.694*** (0.146)	0.0466 (0.113)	2,018	0.782
Paper and paper products	-0.878*** (0.220)	-2.185*** (0.229)	-3.122*** (0.254)	-4.702*** (0.252)	-5.003*** (0.270)	0.648*** (0.171)	0.608*** (0.145)	0.475*** (0.111)	2,017	0.814
Printing and publishing	-0.804*** (0.203)	-1.658*** (0.212)	-2.401*** (0.234)	-3.639*** (0.232)	-4.054*** (0.249)	0.827*** (0.158)	0.931*** (0.132)	0.314*** (0.102)	2,057	0.828
Petroleum products	-2.101*** (0.355)	-3.899*** (0.375)	-5.001*** (0.425)	-6.548*** (0.422)	-7.491*** (0.455)	1.487*** (0.277)	-0.0964 (0.240)	0.531*** (0.198)	1,660	0.627
Chemicals and chemical products	-0.625*** (0.170)	-1.430*** (0.177)	-2.213*** (0.195)	-3.332*** (0.194)	-3.867*** (0.207)	0.530*** (0.132)	0.674*** (0.109)	0.243*** (0.0836)	2,135	0.852
Rubber and plastics products	-0.510*** (0.170)	-1.216*** (0.177)	-2.116*** (0.196)	-3.117*** (0.194)	-3.392*** (0.208)	0.911*** (0.132)	0.610*** (0.110)	0.394*** (0.0845)	2,101	0.862
Non-metallic mineral products	-0.631*** (0.184)	-1.603*** (0.191)	-2.537*** (0.212)	-3.896*** (0.210)	-4.543*** (0.225)	0.905*** (0.143)	0.543*** (0.119)	-0.0178 (0.0923)	2,071	0.842
Basic metals	-0.891*** (0.228)	-2.062*** (0.237)	-2.977*** (0.263)	-4.421*** (0.261)	-4.975*** (0.279	0.577*** (0.177)	0.462*** (0.148)	0.485*** (0.114)	2,071	0.783
Fabricated metal products	-0.524*** (0.177)	-1.322*** (0.184)	-2.185*** (0.203)	-3.292*** (0.201)	-3.637*** (0.215)	0.722*** (0.137)	0.797*** (0.114)	0.202** (0.0874)	2,112	0.860
Other machinery and equipment	-0.295** (0.149)	-0.912*** (0.154)	-1.468*** (0.171)	-2.419*** (0.169)	-2.883*** (0.181)	0.566*** (0.116)	0.601*** (0.0955)	0.295*** (0.0731)	2,135	0.906
Office and computing machinery	-0.487** (0.193)	-1.312*** (0.201)	-1.862*** (0.222)	-2.981*** (0.221)	-3.255*** (0.236)	0.601*** (0.150)	0.502*** (0.126)	0.287*** (.0965)	2,067	0.870
Electrical machinery and apparatus	-0.112 (0.175)	-0.758*** (0.182)	-1.223*** (0.201)	-2.597*** (0.200)	-2.813*** (0.214)	0.742*** (0.136)	0.686*** (0.113)	0.235*** (0.0864)	2,121	0.868
Communication equipment	-0.549** (0.213)	-1.068*** 0.222)	-1.634*** (0.246)	-2.964*** (0.244)	-3.035*** (0.261)	0.507*** (0.166)	0.549*** (0.138)	0.363*** (0.106)	2,084	0.850
Medical and precision instruments	-0.192 (0.164)	-0.708*** (0.170)	-1.128*** (0.188)	-1.995*** (0.186)	-2.272*** (0.199)	0.516*** (0.127)	0.557*** (0.106)	0.279*** (0.0808)	2,115	0.880
Motor vehicles, trailers, and parts	-0.384* (0.219)	-0.941*** (0.227)	-1.780*** (0.251)	-3.102*** (0.249)	-3.433*** (0.267)	0.712*** (0.170)	0.622*** (0.141)	0.688*** (0.109)	2,088	0.852
Other transport equipment	-0.249 (0.254)	-0.936*** (0.265)	-1.510*** (0.294)	-2.153*** 0.292)	-2.723*** (0.312)	0.796*** (0.198)	0.696*** (0.166)	0.264** (0.129)	2,011	0.753
Furniture; other manufacturing	-0.601*** (0.200)	-1.315*** (0.208)	-1.679*** 0.230)	-2.959*** (0.228)	-3.426*** (0.244)	0.629*** (0.156)	0.619*** (0.129)	0.249** (0.0987)	2,123	0.841

Distance (600 $-$
1,199 km)

Distance (1,200 $-$
2,399 km)

Distance (2,400 $-$
4,799 km)

Distance (4,800 $-$ 9,599 km)

Distance (9,600+ km)

Contiguity

Common official or primary language

Trade agreement in force

Observations

R-squared

Food and beverages

-1.147***

(0.194)

-2.068***

(0.202)

-2.662***

(0.223)

-3.916***

(0.221)

-4.355***

(0.236)

0.619***

(0.151)

0.636***

(0.125)

0.290***

(0.0953)

2,132

0.810

Tobacco products

-1.389***

(0.343)

-2.111***

(0.363

-2.769***

(0.411)

-4.136***

(0.406)

-4.553***

(0.440)

0.545**

(0.272)

0.0347

(0.237)

1.314***

(0.197)

1,499

0.539

Textiles

-0.640***

(0.178)

-1.405***

(0.186)

-2.024***

(0.206)

-3.822***

(0.204)

-4.021***

(0.219)

0.704***

(0.139

0.636***

(0.117)

0.127

(0.0896)

2,060

0.854

Wearing apparel, fur

-0.871***

(0.201)

-1.636***

(0.210)

-2.277***

(0.232

-3.839***

(0.230)

-3.822***

(0.246)

0.761***

(0.157)

0.527***

(0.131)

0.113

(0.100)

2,077

0.856

Leather products

-1.048***

(0.227)

-1.957***

(0.237)

-2.710***

(0.263

-3.942***

(0.261)

-3.955***

(0.279)

0.866***

(0.177)

0.197

(0.150)

0.198*

(0.115)

2,036

0.800

Wood products (excl. furniture)

-0.855***

(0.223)

-2.036***

(0.232)

-2.864***

(0.258)

-4.237***

(0.256)

-4.766***

(0.274)

1.014***

(0.174)

0.694***

(0.146)

0.0466

(0.113)

2,018

0.782

Paper and paper products

-0.878***

(0.220)

-2.185***

(0.229)

-3.122***

(0.254)

-4.702***

(0.252)

-5.003***

(0.270)

0.648***

(0.171)

0.608***

(0.145)

0.475***

(0.111)

2,017

0.814

Printing and publishing

-0.804***

(0.203)

-1.658***

(0.212)

-2.401***

(0.234)

-3.639***

(0.232)

-4.054***

(0.249)

0.827***

(0.158)

0.931***

(0.132)

0.314***

(0.102)

2,057

0.828

Petroleum products

-2.101***

(0.355)

-3.899***

(0.375)

-5.001***

(0.425)

-6.548***

(0.422)

-7.491***

(0.455)

1.487***

(0.277)

-0.0964

(0.240)

0.531***

(0.198)

1,660

0.627

Chemicals and chemical products

-0.625***

(0.170)

-1.430***

(0.177)

-2.213***

(0.195)

-3.332***

(0.194)

-3.867***

(0.207)

0.530***

(0.132)

0.674***

(0.109)

0.243***

(0.0836)

2,135

0.852

Rubber and plastics products

-0.510***

(0.170)

-1.216***

(0.177)

-2.116***

(0.196)

-3.117***

(0.194)

-3.392***

(0.208)

0.911***

(0.132)

0.610***

(0.110)

0.394***

(0.0845)

2,101

0.862

Non-metallic mineral products

-0.631***

(0.184)

-1.603***

(0.191)

-2.537***

(0.212)

-3.896***

(0.210)

-4.543***

(0.225)

0.905***

(0.143)

0.543***

(0.119)

-0.0178

(0.0923)

2,071

0.842

Basic metals

-0.891***

(0.228)

-2.062***

(0.237)

-2.977***

(0.263)

-4.421***

(0.261)

-4.975***

(0.279

0.577***

(0.177)

0.462***

(0.148)

0.485***

(0.114)

2,071

0.783

Fabricated metal products

-0.524***

(0.177)

-1.322***

(0.184)

-2.185***

(0.203)

-3.292***

(0.201)

-3.637***

(0.215)

0.722***

(0.137)

0.797***

(0.114)

0.202**

(0.0874)

2,112

0.860

Other machinery and equipment

-0.295**

(0.149)

-0.912***

(0.154)

-1.468***

(0.171)

-2.419***

(0.169)

-2.883***

(0.181)

0.566***

(0.116)

0.601***

(0.0955)

0.295***

(0.0731)

2,135

0.906

Office and computing machinery

-0.487**

(0.193)

-1.312***

(0.201)

-1.862***

(0.222)

-2.981***

(0.221)

-3.255***

(0.236)

0.601***

(0.150)

0.502***

(0.126)

0.287***

(.0965)

2,067

0.870

Electrical machinery and apparatus

-0.112

(0.175)

-0.758***

(0.182)

-1.223***

(0.201)

-2.597***

(0.200)

-2.813***

(0.214)

0.742***

(0.136)

0.686***

(0.113)

0.235***

(0.0864)

2,121

0.868

Communication equipment

-0.549**

(0.213)

-1.068***

0.222)

-1.634***

(0.246)

-2.964***

(0.244)

-3.035***

(0.261)

0.507***

(0.166)

0.549***

(0.138)

0.363***

(0.106)

2,084

0.850

Medical and precision instruments

-0.192

(0.164)

-0.708***

(0.170)

-1.128***

(0.188)

-1.995***

(0.186)

-2.272***

(0.199)

0.516***

(0.127)

0.557***

(0.106)

0.279***

(0.0808)

2,115

0.880

Motor vehicles, trailers, and parts

-0.384*

(0.219)

-0.941***

(0.227)

-1.780***

(0.251)

-3.102***

(0.249)

-3.433***

(0.267)

0.712***

(0.170)

0.622***

(0.141)

0.688***

(0.109)

2,088

0.852

Other transport equipment

-0.249

(0.254)

-0.936***

(0.265)

-1.510***

(0.294)

-2.153***

0.292)

-2.723***

(0.312)

0.796***

(0.198)

0.696***

(0.166)

0.264**

(0.129)

2,011

0.753

Furniture; other manufacturing

-0.601***

(0.200)

-1.315***

(0.208)

-1.679***

0.230)

-2.959***

(0.228)

-3.426***

(0.244)

0.629***

(0.156)

0.619***

(0.129)

0.249**

(0.0987)

2,123

0.841

Source: USITC estimates.

Note: *** p<0.01, ** p<0.05, * p<0.1; standard errors in parentheses.

The next step is to calculate importer-specific trade costs $m_{n j}$ for each industry and country using equation (10). The United States has low trade barriers, as do most of the developed countries. Developing countries have higher barriers. Cuba has some of the highest trade barriers of all the countries in the dataset. Tables I.9 and I.10 show the rankings of the countries in the dataset according to their $m_{n j}$ . Lower-numbered rankings mean lower trade barriers. The first column of numbers shows the average (across industries) rank of each country. Note that $m_{n j}$ is measured relative to domestic costs.

Table I.9: Ranking of countries according to their importer-specific trade cost $m_{n j}$ for selected agricultural sectors

	Average	Wheat	Rice	Corn	Beans and pulses	Other oilseeds	Soybeans	Beef	Poultry	Pork
Canada	5.11	3	18	8	1	5	3	3	4	1
United States	6.00	6	2	4	3	2	4	2	21	10
Australia	11.44	5	1	9	16	11	7	12	34	8
Italy	12.44	9	4	22	9	29	8	13	1	17
Spain	12.44	8	6	14	13	7	40	6	12	6
Germany	13.33	11	19	10	15	9	36	11	7	2
Russia	13.89	21	13	3	8	18	6	15	28	13
United Kingdom	13.89	16	12	31	4	15	28	8	6	5
France	16.56	29	23	5	21	12	17	22	5	15
Venezuela	18.00	1	17	6	22	27	1	21	45	22
Romania	18.22	10	9	13	40	13	10	39	9	21
Brazil	18.89	2	11	12	35	23	25	9	14	39
Argentina	19.11	38	8	2	2	8	38	1	30	45
Portugal	19.56	4	10	24	5	44	32	20	8	29
Denmark	19.89	27	39	38	10	16	39	4	3	3
Netherlands	20.00	7	31	18	33	1	16	28	26	20
Hungary	22.11	35	35	17	26	36	13	18	10	9
Bulgaria	22.22	13	3	16	39	19	43	32	11	24
Thailand	22.33	36	7	30	7	31	11	23	16	40
Lithuania	23.11	25	16	27	28	30	34	14	15	19
Malaysia	23.33	41	29	1	18	6	30	10	37	38
South Korea	23.67	39	33	7	29	32	31	19	19	4
Sweden	24.00	24	25	23	34	22	26	27	17	18
Mexico	24.56	14	14	45	25	10	33	5	40	35
Ireland	25.33	22	30	37	43	45	35	7	2	7
India	25.44	43	21	26	11	3	42	33	18	32
Ukraine	25.78	15	36	11	24	21	23	29	36	37
Belgium	25.89	17	28	34	23	28	14	43	35	11
Japan	26.11	12	44	43	41	4	24	24	29	14
Turkey	26.11	26	15	25	20	25	5	44	42	33
Cuba	26.22	18	32	21	17	41	18	37	24	28
Egypt	26.44	33	20	41	12	26	12	40	38	16
Slovakia	26.56	30	24	15	19	39	19	42	20	31
Austria	26.67	28	40	20	37	24	9	25	31	26
Poland	26.67	32	26	19	32	33	46	17	23	12
Philippines	26.78	23	37	36	30	40	2	26	22	25
Vietnam	28.89	34	22	29	27	42	20	31	13	42
China	29.67	42	42	44	6	14	22	30	33	34
Greece	29.67	20	5	35	36	43	45	35	25	23
Finland	30.22	19	27	39	31	37	27	38	27	27
Indonesia	30.56	31	43	28	14	17	15	41	43	43
Dominican Republic	35.56	45	45	32	38	20	21	34	44	41
Czech Republic	37.11	40	46	42	45	34	29	36	32	30
Taiwan	37.11	37	38	33	46	46	41	16	41	36
Switzerland	41.67	46	41	40	42	38	37	46	39	46
Norway	42.44	44	34	46	44	35	44	45	46	44

Source: USITC estimates.

Table I.10: Ranking of countries according to their importer-specific trade cost $m_{n j}$ for manufactured goods sectors

	Average	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	37
United States	2.14	2	6	4	1	3	1	1	2	1	2	1	1	1	1	1	1	3	6	4	2	2	1
China	7.00	16	37	6	2	2	2	2	4	6	9	3	2	6	2	6	3	5	5	6	15	9	6
Germany	7.36	11	7	3	16	21	7	7	8	24	1	5	3	9	4	3	13	4	1	3	5	4	3
Netherlands	8.95	1	1	2	13	11	13	5	10	10	33	2	18	4	10	2	4	2	7	7	4	8	30
France	11.27	12	11	17	5	17	15	15	12	11	6	8	7	20	12	15	18	9	13	9	13	1	2
Canada	11.32	8	20	35	10	8	8	10	3	7	5	20	21	5	19	8	10	10	16	10	6	6	4
United Kingdom	11.55	14	30	12	3	25	22	11	6	16	4	10	11	7	17	11	14	6	3	1	9	10	12
Thailand	12.27	3	24	13	19	9	5	13	15	23	11	4	15	8	3	12	9	14	14	24	7	12	13
Malaysia	12.73	7	3	10	22	14	6	19	9	4	15	7	16	11	7	22	2	20	4	23	30	20	9
Taiwan	12.95	18	27	19	28	16	11	6	5	26	31	11	8	2	9	5	7	1	11	15	19	5	5
Belgium	14.14	15	10	5	17	31	12	8	23	32	21	6	13	12	25	4	22	13	17	12	3	3	7
Spain	14.86	10	17	22	4	6	17	16	22	22	7	14	9	18	16	16	25	19	19	8	12	13	15
South Korea	15.18	22	16	9	26	18	26	20	11	5	8	15	17	3	6	7	24	12	10	21	8	21	29
Italy	16.27	13	32	11	9	4	18	18	21	27	12	17	12	13	20	13	5	24	21	17	16	18	17
Japan	17.41	33	23	14	23	26	34	25	16	8	16	9	5	16	13	18	26	17	23	5	10	15	8
Indonesia	17.50	5	26	23	27	1	4	3	20	17	19	30	4	14	18	30	19	16	12	36	28	23	10
Mexico	18.18	28	39	27	7	28	20	30	7	30	25	16	10	21	5	19	12	8	9	14	1	26	18
India	18.50	27	28	18	14	7	19	17	14	18	10	21	14	15	8	25	15	28	20	22	21	19	27
Australia	19.32	6	5	25	31	15	21	12	13	3	18	27	22	25	28	26	17	23	34	16	20	24	14
Vietnam	19.41	4	8	8	20	5	3	28	30	13	35	12	20	22	11	29	8	27	18	31	34	37	24
Brazil	22.00	9	35	37	41	24	9	4	34	12	17	18	6	10	14	21	35	30	36	25	25	7	35
Czech Republic	22.36	39	21	7	6	30	36	26	18	37	3	26	23	32	23	14	16	11	24	30	24	25	21
Poland	23.00	23	18	30	15	20	29	27	25	33	24	24	24	17	24	20	23	22	15	29	22	22	20
Sweden	23.09	21	46	15	33	38	24	14	17	20	23	13	26	29	15	17	31	15	25	18	17	32	19
Philippines	23.68	19	13	28	18	34	10	22	26	34	36	19	33	41	22	36	6	18	8	19	38	30	11
Turkey	24.95	32	29	16	11	29	33	24	35	28	29	22	19	24	21	9	29	32	26	27	14	27	33
Russia	26.45	24	22	41	38	32	14	9	32	9	20	34	36	19	29	35	21	38	38	33	31	11	16
Portugal	26.55	26	14	26	12	10	16	23	39	38	32	25	25	35	36	31	39	35	2	26	26	34	34
Austria	26.77	30	45	24	32	12	23	29	19	46	26	29	28	30	26	24	28	26	27	28	18	16	23
Denmark	28.23	17	34	1	25	33	30	35	27	35	30	32	30	38	27	33	30	36	29	20	36	17	26
Hungary	29.18	41	38	40	24	23	35	33	29	43	22	23	29	42	32	27	20	7	22	13	23	38	38
Switzerland	29.64	31	15	32	8	37	43	34	28	39	13	35	32	37	35	23	43	34	35	11	41	14	32
Finland	30.14	42	42	34	21	27	27	21	24	41	14	36	31	26	38	34	40	25	37	2	35	35	31
Norway	30.77	20	47	39	34	13	40	31	38	21	34	37	40	23	31	40	34	31	31	32	11	28	22
Romania	31.41	44	4	21	30	35	31	46	40	36	38	28	42	33	30	32	27	21	33	34	27	31	28
Bulgaria	34.27	38	40	20	36	19	39	39	44	44	28	42	37	34	44	10	37	37	40	35	37	29	25
Argentina	34.82	34	9	42	40	41	38	41	31	2	41	39	27	36	33	38	32	40	41	37	39	42	43
Ireland	35.45	25	41	36	43	44	37	38	1	45	42	33	44	44	39	28	11	39	28	41	40	39	42
Ukraine	36.09	36	36	38	37	22	28	42	43	29	27	43	45	31	41	39	38	33	43	39	32	36	36
Slovakia	36.18	43	44	31	29	36	44	36	33	40	44	31	34	28	37	41	42	29	32	43	29	33	37
Egypt	36.64	37	19	29	42	39	25	40	36	14	39	38	35	39	34	37	44	44	46	45	43	41	40
Lithuania	38.18	35	12	33	39	42	32	43	37	47	37	40	38	46	40	44	33	43	30	38	44	46	41
Greece	39.27	29	25	43	35	40	46	37	41	42	40	41	39	40	45	43	45	42	39	40	33	40	39
Venezuela	40.23	46	2	46	45	46	45	47	42	25	43	45	43	27	43	42	36	41	42	42	47	44	46
Cuba	42.64	40	31	44	46	47	41	32	45	31	46	46	46	43	46	46	41	45	44	46	42	43	47
Dominican Republic	43.36	45	33	45	44	45	42	45	46	15	47	44	41	47	47	45	46	46	47	47	45	47	45
Algeria	44.32	47	43	47	47	43	47	44	47	19	45	47	47	45	42	47	47	47	45	44	46	45	44

Source: USITC estimates.

Note: See table I.3 for industry names corresponding to ISIC codes 15-36.

Table I.11: Trade costs for Cuba’s imports from various countries, $d_{Cuba, i j}$ , in tariff equivalents in selected agricultural sectors

	Wheat	Rice	Corn	Beans and pulses	Other oilseeds	Soybeans	Beef	Poultry	Pork
Argentina	-44.2	150.6	72.8	86.5	632.6	-41.1	277.3	146.3	393.0
Australia	115.5	567.6	280.6	500.4	381.5	36.1	727.7	512.1	557.4
Austria	38.0	237.8	282.4	225.8	341.0	17.0	363.3	378.0	536.1
Belgium	31.7	344.5	176.7	273.6	387.4	17.6	388.0	531.2	705.3
Brazil	112.9	72.3	72.6	348.6	445.1	-18.4	241.2	151.0	270.7
Bulgaria	68.4	279.2	311.9	214.8	515.7	-18.1	232.8	399.9	365.5
Canada	-36.3	184.5	46.2	100.3	303.1	89.5	119.0	104.3	220.8
China	32.5	497.7	227.2	95.4	695.4	58.0	402.1	531.4	580.3
Cuba	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Czech Republic	-55.7	238.4	211.3	262.9	352.1	1.7	256.5	383.3	406.3
Denmark	48.4	236.7	191.3	264.2	389.6	-4.6	390.8	497.3	771.4
Dominican Republic	-22.3	162.0	97.0	232.6	314.6	-0.8	295.8	297.7	307.8
Egypt	-7.8	474.3	154.9	339.5	374.4	-3.3	249.7	388.5	391.7
Finland	43.8	225.8	127.9	185.3	256.6	-3.7	234.6	382.9	500.2
France	-59.8	357.0	443.3	218.4	199.5	13.1	397.0	622.3	674.7
Germany	-49.3	343.6	299.3	324.4	487.7	14.2	481.6	592.2	830.3
Greece	33.9	402.9	195.7	218.6	348.1	-8.7	223.1	362.0	395.3
Hungary	58.9	226.8	342.5	308.5	398.7	13.9	302.4	554.7	587.5
India	36.3	337.2	333.5	404.5	240.3	50.6	548.1	375.1	457.9
Indonesia	21.2	348.4	241.4	310.8	440.4	22.7	264.5	426.8	338.4
Ireland	18.7	247.8	181.0	176.8	265.8	-12.1	459.7	465.7	626.7
Italy	41.6	238.7	272.4	196.0	166.0	29.9	405.0	518.9	562.1
Japan	30.5	370.4	135.7	233.3	410.6	6.3	335.0	416.3	429.3
Korea, South	-10.8	366.2	152.5	229.2	309.7	4.3	361.3	406.2	485.3
Lithuania	57.8	246.5	171.8	242.5	319.3	0.1	320.8	409.8	384.0
Malaysia	-17.0	300.7	163.4	306.7	382.2	7.2	294.2	363.0	335.8
Mexico	-46.9	91.9	81.9	109.8	289.4	-6.5	557.4	430.5	449.9
Netherlands	33.6	374.6	242.4	303.8	255.4	34.3	451.5	321.9	800.9
Norway	-25.7	221.8	101.6	189.0	191.1	4.0	205.7	287.5	344.1
Philippines	19.5	359.9	176.2	219.2	429.4	8.2	356.6	470.0	488.4
Poland	-52.0	272.1	259.3	250.6	358.6	-14.1	440.5	604.9	635.3
Portugal	14.1	329.4	174.4	277.6	282.7	1.4	301.5	513.4	523.4
Romania	91.2	317.9	394.3	225.9	449.0	28.6	276.3	488.6	463.3
Russia	-43.2	382.5	401.8	394.4	466.8	64.6	349.4	397.7	454.8
Slovakia	30.8	229.4	277.9	229.3	344.8	8.9	218.6	379.3	362.7
Spain	48.7	169.8	273.3	120.7	169.5	10.7	157.2	137.3	265.5
Sweden	32.1	232.2	167.5	244.8	359.2	-15.9	260.2	413.1	472.1
Switzerland	-15.5	217.9	166.5	192.7	261.3	-4.9	176.1	267.0	318.0
Taiwan	7.2	402.0	149.6	252.4	371.7	13.8	296.1	456.6	430.2
Thailand	4.0	318.8	242.8	368.6	509.4	12.3	311.8	535.8	391.1
Turkey	38.0	327.2	256.5	373.7	439.7	38.7	252.5	427.1	363.0
Ukraine	109.9	221.1	81.6	340.3	521.7	88.0	229.9	312.3	312.1
United Kingdom	51.4	353.0	192.2	357.3	421.7	6.4	401.4	563.9	602.8
United States	-28.9	455.2	62.1	126.5	(a)	-37.3	267.5	84.9	160.9
Venezuela	41.9	360.7	296.1	300.8	425.7	41.9	367.0	366.8	392.1
Vietnam	6.8	61.5	205.4	302.4	405.5	4.9	243.7	421.2	358.4

Source: USITC estimates.

a Absence of trade precludes estimating trade costs for this sector in the base year.

Table I.12: Trade costs for Cuba’s imports from various countries, $d_{Cuba, i j}$ , in tariff equivalents in manufactured goods sectors

	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36
Algeria	267.3	145.6	228.6	262.9	338.7	282.5	227.9	241.9	40.2	351.2	295.5	326.1	306.8	293.2	418.7	209.3	301.1	205.6	270.7	241.7	313.7	300.3
Argentina	187.3	256.0	247.2	424.4	457.7	238.4	71.8	249.9	222.8	181.9	166.4	192.7	199.9	172.6	246.4	256.7	166.9	129.3	173.3	194.7	108.2	365.3
Australia	232.0	249.1	395.5	636.2	564.8	532.1	532.3	468.0	412.5	621.1	483.7	199.6	567.5	336.3	306.4	576.9	267.2	227.3	356.2	684.6	264.9	599.8
Austria	212.3	212.2	395.2	691.2	548.7	619.6	355.9	417.9	179.2	254.5	291.2	410.3	359.1	381.7	341.0	352.9	389.9	362.0	307.3	431.3	203.4	362.8
Belgium	166.0	121.7	325.9	606.0	479.4	532.4	173.3	301.0	93.8	244.8	314.9	266.3	257.9	271.3	307.3	203.6	306.5	257.3	232.9	298.6	465.8	343.1
Brazil	154.0	101.0	199.5	278.0	190.5	153.9	159.2	163.3	166.9	203.9	199.1	229.4	242.0	177.6	209.9	182.8	184.1	257.3	188.7	194.4	363.2	254.4
Bulgaria	556.5	215.7	265.5	615.3	370.1	183.8	237.8	346.0	151.5	139.9	312.7	478.5	316.5	296.1	293.1	275.4	327.1	349.6	321.4	188.3	312.3	513.5
Canada	155.3	253.0	192.9	240.4	191.0	199.2	162.5	251.4	106.4	156.2	207.4	193.7	149.1	158.1	217.6	137.3	176.1	194.1	204.3	158.4	290.4	271.0
China	223.0	335.2	243.2	416.9	243.0	240.7	194.5	300.7	144.2	205.9	268.5	291.1	176.7	280.6	300.7	334.7	277.5	280.4	242.2	197.9	197.2	363.6
Cuba	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Czech Republic	111.4	307.3	401.9	690.3	538.7	265.2	391.1	484.3	149.6	308.9	270.2	288.4	178.8	229.4	281.3	281.4	198.2	335.3	211.4	313.0	176.3	668.3
Denmark	277.3	227.3	360.9	695.1	386.4	458.7	216.2	606.6	157.2	249.3	359.5	344.1	186.4	253.1	326.5	261.1	372.6	245.0	246.3	147.8	200.1	454.2
Dominican Republic	63.7	276.7	120.5	250.0	203.1	73.9	82.8	106.4	148.7	74.0	187.6	145.5	110.3	79.0	125.8	103.2	173.0	152.8	237.5	50.8	136.0	183.3
Egypt	278.0	221.0	336.4	454.4	427.4	305.4	279.2	443.2	409.7	171.9	235.0	411.4	480.3	546.3	324.4	132.7	174.9	273.3	403.4	186.1	347.2	646.2
Finland	245.7	126.6	549.9	615.9	440.3	402.5	201.9	552.2	197.9	787.3	464.1	536.9	262.8	423.0	399.1	389.7	233.8	653.8	321.4	371.8	587.7	812.1
France	197.5	334.8	360.1	477.1	390.9	207.4	259.3	346.1	241.3	218.1	295.2	252.4	261.1	290.8	317.3	212.3	278.3	261.7	260.5	191.3	320.3	430.6
Germany	192.0	173.4	364.7	372.5	331.7	357.7	180.9	360.9	72.0	223.0	271.2	273.5	260.2	254.1	309.3	281.9	295.6	323.8	241.0	267.9	549.4	393.1
Greece	249.4	222.1	378.3	685.7	381.7	355.7	360.8	270.2	210.6	372.1	620.2	616.2	567.3	597.3	734.0	356.0	408.5	373.5	468.0	273.7	528.0	536.5
Hungary	455.9	231.2	540.2	677.2	492.1	279.1	410.6	396.6	45.8	311.7	356.0	668.3	186.1	489.4	370.2	327.9	497.6	737.2	272.0	391.0	541.7	495.5
India	415.3	297.2	393.6	631.0	445.1	392.4	310.4	346.5	333.8	241.2	264.1	418.6	240.0	331.1	381.0	610.7	419.0	230.2	238.0	285.4	403.8	537.5
Indonesia	462.2	363.5	428.4	589.5	593.4	775.5	140.2	354.6	340.0	281.3	488.9	454.6	698.1	569.3	1025.4	664.6	365.1	798.9	375.7	716.2	728.8	651.6
Ireland	261.1	226.3	474.3	329.1	403.9	122.0	329.2	782.5	162.6	458.5	470.8	446.7	366.8	604.7	389.4	437.2	430.6	568.8	318.6	284.7	651.0	521.4
Italy	198.5	210.5	278.5	363.9	286.0	154.6	140.2	265.4	146.2	179.3	235.1	213.8	169.5	229.6	257.3	209.2	197.2	179.3	201.2	229.0	268.6	335.2
Japan	677.3	277.1	378.5	770.7	257.5	494.6	253.3	405.9	370.7	362.2	469.5	608.0	451.3	359.0	417.5	336.4	333.5	408.1	278.2	464.9	527.6	463.4
Korea, South	420.5	322.0	313.4	866.4	373.2	510.8	348.9	369.7	359.9	317.7	346.1	565.8	372.6	344.4	394.1	581.8	374.6	412.2	314.5	234.5	478.5	608.6
Lithuania	103.5	317.0	514.9	516.0	323.3	498.2	278.7	332.6	154.4	169.7	242.1	447.7	349.3	290.2	302.8	342.9	263.4	413.8	512.6	141.4	383.1	723.0
Malaysia	273.8	328.1	354.7	407.0	189.0	718.4	461.9	469.0	407.6	265.4	531.3	414.5	351.0	440.1	457.3	390.3	442.8	309.1	544.0	259.8	744.3	604.1
Mexico	121.3	245.1	137.4	252.0	130.2	72.7	84.6	174.2	14.9	128.5	146.2	161.6	132.3	117.4	225.1	216.7	201.8	232.8	199.9	271.2	178.7	265.3
Netherlands	217.3	148.7	241.1	350.5	210.3	227.2	174.9	419.0	33.6	201.4	247.9	282.8	170.9	205.0	284.8	352.5	258.9	359.1	250.5	185.3	274.2	371.0
Norway	738.1	121.5	196.5	466.9	330.8	165.0	191.9	376.9	411.2	400.2	250.1	431.8	639.2	219.2	306.4	265.0	319.6	287.6	313.4	283.3	431.7	341.4
Philippines	677.7	306.9	579.1	452.4	482.2	500.1	406.2	494.3	206.6	691.3	787.7	618.7	442.7	502.6	831.4	377.2	824.2	463.4	224.8	304.2	600.2	345.6
Poland	173.0	386.3	323.3	780.4	381.0	580.2	232.7	470.7	173.5	255.4	330.8	302.6	439.9	469.4	328.5	491.7	399.7	509.4	331.1	296.4	263.0	609.5
Portugal	201.3	266.7	286.0	657.7	462.5	264.0	167.4	165.7	218.6	127.9	241.2	215.7	110.7	220.5	234.0	279.2	218.1	342.8	225.1	247.4	289.6	372.9
Romania	498.2	323.5	559.2	839.8	570.3	228.4	288.8	367.8	213.3	639.7	323.9	563.0	570.0	330.3	500.7	484.0	750.2	554.8	574.0	392.8	174.9	641.6
Russia	348.7	292.9	174.4	322.9	252.6	334.6	112.9	248.3	105.0	235.4	146.2	211.0	236.4	165.2	153.3	150.8	120.7	93.3	135.1	72.6	167.2	305.0
Slovakia	425.4	142.5	275.3	644.1	504.9	188.8	396.9	173.9	188.4	326.0	262.3	187.9	172.5	167.4	271.2	164.9	199.0	352.0	196.5	683.4	470.6	482.4
Spain	146.9	207.6	149.9	268.7	165.0	118.3	98.7	157.1	136.8	147.5	181.4	176.9	110.5	143.8	173.5	129.6	142.7	128.6	135.1	158.6	184.0	235.3
Sweden	370.7	208.8	356.8	586.5	429.7	344.0	214.6	325.0	212.8	221.8	343.3	607.1	374.3	341.4	330.6	229.7	283.3	257.8	262.0	227.2	643.7	550.0
Switzerland	271.6	283.7	391.8	598.7	533.0	399.0	307.9	401.8	145.2	249.4	340.7	315.6	262.2	296.4	320.1	300.8	327.6	327.9	265.2	212.1	440.1	444.3
Taiwan	279.7	234.8	390.5	604.7	693.3	534.6	571.0	593.9	309.0	429.2	374.9	539.4	378.1	521.5	466.8	457.0	425.7	508.0	472.4	441.6	539.6	547.4
Thailand	325.9	258.3	716.2	757.4	561.3	624.4	314.3	698.8	278.1	484.5	429.5	774.7	396.8	550.0	411.2	889.6	652.6	885.0	527.3	574.9	783.7	619.9
Turkey	187.0	244.9	394.7	697.9	526.0	316.0	157.1	356.1	88.8	210.3	341.1	387.0	193.3	321.0	422.5	405.8	261.6	371.0	392.3	303.1	582.1	535.6
Ukraine	175.2	201.6	161.8	562.9	133.2	237.2	257.7	196.7	341.9	151.4	154.1	238.7	165.7	161.9	198.3	202.4	185.4	141.2	154.8	82.3	94.9	289.5
United Kingdom	245.9	210.9	405.5	522.4	525.7	294.9	199.9	378.3	402.0	269.7	456.8	330.6	194.0	277.6	328.6	348.4	368.3	386.1	281.9	334.3	482.0	424.6
United States	157.4	(a)	485.5	583.3	(a)	279.4	(a)	702.8	(a)	407.3	718.8	(a)	(a)	695.7	560.7	(a)	751.5	439.1	470.5	510.3	594.8	501.0
Venezuela	109.0	392.9	133.7	197.4	306.9	190.4	186.9	140.4	41.1	99.1	58.5	107.3	82.7	94.2	147.1	109.4	89.6	84.6	167.0	169.9	140.9	239.1
Vietnam	129.5	308.8	294.9	474.1	294.4	480.5	110.7	565.1	211.3	135.2	285.8	195.7	302.7	432.2	429.5	741.1	315.9	823.6	256.1	556.9	453.2	352.6

Source: USITC estimates.

Note: See table I.3 for industry names corresponding to ISIC codes 15-36

a Absence of trade precludes estimating trade costs for this sector in the base year.

Table I.13: Trade costs for U.S. imports from Canada and Germany, d_us_ij, in tariff equivalents, current

Exporter	Food	Apparel	Metals	Machinery, office	Medical	Vehicles
Canada	22.6	5.2	10.4	-3.2	8.9	-17.1
Germany	73.3	67.2	46.7	23.7	25.2	32.5

Source: USITC estimates.

Table I.14: Trade costs for Brazilian imports from the United States and Argentina, d_br_ij, in tariff equivalents

Exporter	Food	Apparel	Metals	Machinery, office	Medical	Vehicles
United States	91.9	242.7	101.9	176.5	94.1	132.1
Argentina	63.5	100.7	35.4	102.0	36.4	14.1

Source: USITC estimates.

Trade costs for the U.S. exports to Cuba in the absence of U.S. restrictions are calculated using equation (7). The resulting trade costs are presented in the second column of table 8.1 in chapter 8.

To calculate trade costs for the U.S. exports when U.S. restrictions are removed and Cuban import barriers are lowered, it is necessary to calculate reduced Cuban import barriers. Cuban import barriers $m_{Cuba, j}$ in each industry are set equal to the average of all developing countries in the dataset. Cuba’s ranking among the countries in the dataset according to the country-specific import barrier is shown in table I.15. The first column shows the base year values (also shown in tables I.8 and I.9). The second column shows the values of Cuban import barriers when they are set equal to ${\bar{m}}_{Dev, j}$ . As an alternative, the third column of table I.15 shows the values of Cuban import barriers when they are set equal to $m_{Dev, j}^{min}$ , the minimum across developing countries in each industry.

Table I.15: Ranking of Cuba among all U.S. export destinations, for alternative scenarios of reduced Cuban trade barriers

	Base year	Average of developing countries	Lowest of developing countries
Agricultural goods average	26.22	21.01	2.67
Wheat	18	17	1
Rice	32	23	3
Corn	21	18	1
Beans and pulses	17	15	2
Other oilseeds	41	35	3
Soybeans	18	17	1
Beef	37	25	1
Poultry	24	18	3
Pork	28	21	9
Manufactured goods average	42.64	27.86	3.77
Food products and beverages	40	28	3
Tobacco products	31	24	2
Textiles	44	29	6
Wearing apparel; dressing and dyeing of fur	46	30	2
Leather, handbags, footwear	47	27	1
Wood and wood products	41	24	2
Paper and paper products	32	28	2
Publishing, printing and reproduction of recorded media	45	30	4
Coke, refined petroleum products, and nuclear fuel	31	22	2
Chemicals and chemical products	46	29	9
Rubber and plastics products	46	28	3
Other non-metallic mineral products	46	28	2
Basic metals	43	28	6
Fabricated metal products	46	26	2
Other machinery and equipment	46	30	6
Office, accounting and computing machinery	41	25	2
Electrical machinery and apparatus	45	29	5
Radio, television, and communications equipment	44	28	4
Medical, precision, and optical instruments, watches	46	31	6
Motor vehicles, trailers, and parts	42	30	1
Other transport equipment	43	30	7
Furniture; other manufacturing	47	29	6

Source: USITC estimates.

Using the new values of $m_{Cuba, j}$ , new trade costs between U.S. and Cuba can be calculated. The trade costs for the U.S. export to Cuba without U.S. restrictions and with Cuban trade barriers set equal to the average of developing countries are presented in third column of table 8.1 in chapter 8. The trade costs for the U.S. exports to Cuba without U.S. restrictions and with Cuban trade barriers set equal to the minimum (lowest) of developing countries are presented in table I.16 below.

Table I.16: Estimated trade costs for U.S. exports to Cuba without U.S. restrictions and with Cuban import barriers set equal to the lowest developing-country level

Sector	Trade cost
Agriculture
Wheat	-57.8
Rice	11.2
Corn	29.1
Pulses	49.0
Other oilseeds	48.6
Soybeans	-49.2
Beef	88.5
Poultry	43.3
Pork	102.1
Manufacturing
Food products and beverages	24.8
Tobacco products	86.6
Textiles	44.5
Wearing apparel; dressing and dyeing of fur	71.5
Leather, handbags, footwear	1.6
Wood and Wood products	50.1
Paper and paper products	52.6
Publishing, printing and reproduction of recorded media	95.8
Coke, refined petroleum products and nuclear fuel	83.2
Chemicals and chemical products	52.4
Rubber and plastics products	61.9
Other non-metallic mineral products	84.7
Basic metals	60.2
Fabricated metal products	66.2
Other machinery and equipment	55.4
Office, accounting and computing machinery	53.9
Electrical machinery and apparatus	52.5
Radio, television and communication equipment	37.8
Medical, precision and optical instruments, watches	53.8
Motor vehicles, trailers, and parts	41.0
Other transport equipment	68.2
Furniture; other manufacturing	32.8

Source: USITC estimates.

Note: International trade costs in this table are measured relative to domestic trade costs in Cuba.

The effects of the two potential sets of policy changes on U.S. exports to Cuba are presented in chapter 8. That chapter shows the estimated U.S. exports to Cuba in the event that U.S. restrictions are removed. It also shows the estimated U.S. exports to Cuba if U.S. restrictions are removed and Cuba-specific trade costs, $m_{Cuba, j}$ , are lowered to the average across developing countries. As an alternative, tables I.17 and I.18 below show estimates for U.S. exports to Cuba if Cuba-specific trade costs are lowered to the lowest of developing countries in the dataset. In this scenario, Cuba’s proximity, combined with very low barriers, would make it a very attractive export destination for U.S. products. As noted above, estimates for this scenario should be taken with considerable caution, since the model assumes that no adjustments to Cuban productivity and wages would occur as barriers were lowered.

Table I.17: Estimated U.S. exports to Cuba in selected agricultural sectors, with U.S. restrictions removed and Cuban import barriers lowered to the lowest developing-country level

	U.S. exports to Cuba		Shares in Cuban spending		Shares in Cuban imports
Industry name	Base year	Estimated	Base year	Estimated	Base year	Estimated
Wheat	18.0	187.8	6.1	58.2	6.1	58.2
Rice	(a)	182.5	(b)	43.8	(b)	44.9
Corn	100.6	165.5	36.8	57.2	45.3	61.1
Beans, cowpeas, pulses	6.6	29.9	5.7	25.9	10.6	29.1
Other oilseeds	0.0	6.0	0.0	52.1	0.0	53.4
Soybeans	50.6	61.6	74.0	82.5	74.0	82.5
Beef	0.4	59.4	0.2	30.3	5.8	81.6
Poultry	125.4	189.3	56.4	81.6	74.3	87.4
Pork	11.3	69.9	3.6	21.9	78.7	89.8
Total	312.8	952.0	16.3	47.7	28.7	68.2

Source: USITC estimates.

a Less than $50,000.

b Less than 0.05 percent.

Table I.18: Estimated U.S. exports to Cuba in manufacturing sectors, with U.S. restrictions removed and Cuban import barriers lowered to the lowest developing-country level

		U.S. exports to Cuba		Shares in Cuban spending		Shares in Cuban imports
ISIC	Industry name	Base year	Estimated	Base year	Estimated	Base year	Estimated
15	Food and beverages	212.4	1,333.1	5.0	33.2	22.8	34.5
16	Tobacco products	0.0	2.5	0.0	0.7	0.0	25.6
17	Textiles	0.1	8.7	0.0	6.7	0.1	6.7
18	Wearing apparel, fur	(a)	1.6	(b)	1.6	0.1	1.6
19	Leather products	0.0	1.1	0.0	0.9	0.0	0.9
20	Wood products (excl. furniture)	1.3	10.8	2.1	18.1	2.6	18.2
21	Paper and paper products	0.0	40.4	0.0	40.8	0.0	41.0
22	Printing and publishing	(a)	16.3	(b)	28.8	0.1	29.4
23	Petroleum products	0.0	74.5	0.0	13.0	0.0	18.4
24	Chemicals and chemical products	3.2	378.2	0.2	22.9	0.4	23.2
25	Rubber and plastics products	(a)	33.8	(b)	10.8	(b)	10.8
26	Non-metallic mineral products	0.0	25.9	0.0	12.7	0.0	12.8
27	Basic metals	0.0	44.7	0.0	9.3	0.0	9.5
28	Fabricated metal products	(a)	26.4	0.0	8.3	0.0	8.3
29	Machinery and equipment n.e.c.	1.5	209.6	0.2	21.6	0.2	21.6
30	Office and computing machinery	0.0	17.3	0.0	36.2	0.0	36.3
31	Electrical machinery and apparatus	(a)	46.8	(b)	14.0	(b)	14.0
32	Communication equipment	0.3	14.6	0.4	20.9	0.5	20.9
33	Medical and precision instruments	0.7	57.2	0.4	29.3	0.5	29.4
34	Motor vehicles, trailers	0.3	72.9	0.1	17.4	0.1	17.4
35	Other transport equipment	0.1	39.8	0.0	14.3	0.1	14.4
36	Furniture; manufacturing n.e.c.	5.0	464.2	0.5	30.8	4.0	30.8
	Total	224.8	2,920.3	1.8	19.7	4.3	19.7

Source: USITC estimates.

Note: "N.e.c." stands for "not elsewhere classified."

a Less than $50,000.

b Less than 0.05 percent.

Bibliography

Anderson, James E., and Eric van Wincoop. “Gravity with Gravitas: A Solution to the Border Puzzle.”American Economic Review, 93(2003): 170 $-$ 192.

$—$ $—$ $—$ .“Trade Costs.” Journal of Economic Literature, 42 no. 3 (2004): 691 $-$ 751.

Caliendo, Lorenzo, and Fernando Parro. “Estimates of the Trade and Welfare Effects of NAFTA.”The Review of Economic Studies, 82(2015): 1 $-$ 44.

CEPII. CEPII Gravity database. http://www.cepii.fr/CEPII/en/bdd_modele/presentation.asp?id=8 (accessed April 17, 2015).

Chor, Davin. “Unpacking Sources of Comparative Advantage: AQuantitative Approach.”Journal of International Economics, 82 (2010): 152 $-$ 67.

Eaton, Jonathan, and Samuel Kortum. “Putting Ricardo to Work.”Journal of Economic Perspectives 26(2012): 65 $-$ 90.

$—$ $—$ $—$ .“Technology, Geography, and Trade.”Econometrica, 70(2002): 1741 $-$ 79.

Food and Agriculture Organization of the United Nations. Value of Agricultural Production database. http://faostat3.fao.org/download/Q/QV/E (accessed August 13, 2015).

Global Trade Information Systems, Inc. (GTIS). Global Trade Atlas database. http://www.gtis.com(accessed various dates).

Head, Keith and Thierry Mayer.“Gravity Equations: Workhorse, Toolkit, and Cookbook.” Chapter 3 inHandbook of International Economics, 2014.

Oficina Nacional de Estadísticas e Información (ONEI) [Cuban National Office of Statistics and Information]. Anuario Estadístico de Cuba 2014 [Statistical Yearbook of Cuba 2014], 2015. http://www.one.cu/aec2014/00%20Anuario%20Estadistico%202014.pdf and http://www.one.cu/series2013.htm.

Levchenko, Andrei A. and Jing Zhang.“Comparative Advantage and the Welfare Impact of European Integration.”Economic Policy, CEPR & CES & MSH. 27(2012): 567 $-$ 602.

Shikher, Serge. “Capital, Technology, and Specialization in the Neoclassical Model.”Journal of International Economics 83(2011): 229 $-$ 42.

$—$ $—$ $—$ .“Putting Industries into the Eaton $-$ Kortum Model.”Journal of International Trade and Economic Development 21 (2012): 807 $-$ 37.

$—$ $—$ $—$ .“Predicting the Effects of NAFTA: Now We Can Do It Better!”Journal of International and Global Economic Studies 5(2012): 32 $-$ 59.

United Nations (UN) Comtrade database. Via the World Integrated Trade Solution (WITS).http://unstats.un.org/unsd/tradekb/Knowledgebase/Use-UN-Comtrade-via-World-Integrated-Trade-Solution-WITS (accessed February 27, 2015).

United Nations Industrial Development Organization (UNIDO). IndStat2. 2014 release.

WTO. Regional Trade Agreements Information System (RTA $-$ IS). http://rtais.wto.org/UI/PublicAllRTAList.aspx(accessed May 21, 2015).

[1] Head and Mayer, “Gravity Equations:Workhorse,Toolkit, and Cookbook,” 2014.

[2] There are different types of gravity models, some of which are grounded in theory and called structural gravity models. While the majority of the gravity models in use before the mid2000s were ad hoc models not based on theory, most of those are now considered misspecified.

[3] Gravity models without general equilibrium features ignore the effects of increased competition in Cuba on U.S. exports. They hold country fixed effects (or multilateral resistance indices) constant even though trade policy would change them. When U.S. sanctions are removed, competition in Cuba becomes fiercer. When faced with the entry of U.S. firms into the Cuban market, some foreign firms would exit the Cuban market, but other, more efficient, foreign firms would drop their prices to remain competitive. This decrease in prices offered by competitors (captured by changing multilateral resistance terms) would decrease the potential U.S. exports to Cuba.

[4] The Commission regularly uses the Global Trade Analysis Project (GTAP) CGE model and partial equilibrium analysis to estimate the effects of trade policy. These models cannot be used here because, as noted here and in chapter 8, they require positive trade flows and tariff equivalents. Moreover, GTAP does not include Cuba, and incorporating Cuba into the model would require a Cuban input-output table, which is not available.

[5] The removal of U.S. restrictions would reverse the trade diversion that occurred when restrictions were imposed. Without the U.S. restriction, U.S. firms would be able to compete in Cuba on a level playing field with firms from other countries. Therefore, Cuban consumers would be able to buy from the most efficient suppliers. The entry of U.S. firms into Cuban market would make competition there more fierce and result in lower product prices in the Cuban market.

[6] Anderson and van Wincoop, “Trade Costs,”2004, 691751, explains the nature and typical magnitudes of trade costs.

[7] NTR tariffs (called Most-Favored Nation (MFN)tariffs outside the U.S.) are what countries promise to impose on imports from other members of the WTO, unless the country is part of a preferential trade agreement.

[8] Anderson and van Wincoop, “Trade Costs,”2004; Head and Mayer, “Gravity Equations:Workhorse,Toolkit, and Cookbook,” 2014.

[9] Eaton and Kortum, “Technology, Geography, and Trade,” 2002, 174179; Shikher, “Putting Industries into the Eaton-Kortum Model,” 2012, 80737.

[10] Chor, “Unpacking Sources of Comparative Advantage,”2010, 15267; Shikher, “Capital, Technology, and Specialization in the Neoclassical Model,” 2011, 22942.

[11] Shikher, “Predicting the Effects of NAFTA,”2012, 3259; Caliendo and Parro, “Estimates of the Trade and Welfare Effects of NAFTA,” 2015, 144; Levchenko and Zhang, “Comparative Advantage and the Welfare Impact,”2012, 567602.

[12] Eaton and Kortum, “Putting Ricardo to Work,” 2014, 6590.

[13] Physical distance between countries is based on bilateral distances between the biggest cities of the

two countries, those intercity distances being weighted by the share of the city in the overall

country’s population. This measure, distw, is provided by the Centre d’Études Prospectives et d’Informations Internationales (CEPII). The measure accounts for the fact that goods have to be sent from various parts of the country in order to be exported. Therefore, the distance between the United States and Cuba is greater than the distance between Miami and Cuba.

[14]Andersonand van Wincoop, “Gravity with Gravitas,”2003, 17092; Head and Mayer, “Gravity Equations:Workhorse,Toolkit, and Cookbook,” 2014.

[15] The model presented by Eaton and Kortum in “Technology, Geography, and Trade,” 2002, has a more micro-founded interpretation of . In their model it is a parameter of the distribution of productivities of producers within an industry.

[16] Set in equation (3) and subtract the resulting equation from (3); then move to the right-hand side, plug in the expression for trade costs (1), and add coefficients. The details of this derivation can be seen in Eaton and Kortum, “Technology, Geography, and Trade,” 2002, and Shikher, “Putting Industries into the Eaton-Kortum Model,” 2012.

[17] For notational simplicity, “hat” symbols are omitted from the estimated coefficients.

[18] This is a common approach in gravity literature. It produces consistent estimates if missing and zero trade values are randomly distributed across observations. See UNCTAD and WTO, “A Practical Guide to Trade Policy Analysis,” 2012, and Head and Mayer, “Gravity Equations:Workhorse,Toolkit, and Cookbook,” 2014, for more discussion.

[19] Counterfactuals are estimated and simulated using a range of parameters between 3 and 13. The choice of has only a small effect on the results presented in chapter 8. Changing affects the value of trade costs shown in tables I.10 and I.11. Using a lower results in higher estimates of for all country pairs, including Cuba-U.S. trade costs. However, trade costs between the U.S. and Cuba relative to the trade costs between other country pairs are not affected.

[20] When estimating (9), the 's are high (see tables I.6 and I.7).

[21] These are countries with GDP per capita of less than 23 percent of the U.S.GDP per capita.

[22] See Anderson and van Wincoop, “Gravity with Gravitas,”2003; Head and Mayer “Gravity Equations:Workhorse,Toolkit, and Cookbook,” 2014.

[23] It is also necessary to assume that Cuban sourcing decisions are driven by prices and product characteristics.

[24]Eaton and Kortum in “Technology, Geography, and Trade,” 2002, denote the productivity parameter by T. The relationship between T and A is .

[25] The concordance used for the agricultural trade data was taken from the FAO.

[26] The concordance used for the manufacturing trade data was created by the Commission’s economists and verified by the Commission’s industry experts.

[27] Using the 1989 values of production, the 19892012 index time series were converted into values. The resulting output values were in 1989 USD, which were converted to 201012 USD using a U.S. GDP deflator.

[28] ONEI usually does not specify which peso is being used in the reported data. However, as a rule, trade data are reported in convertible pesos, while production data are reported in national pesos. Note that the Commission’s procedure for obtaining Cuban data does not use any data reported in pesos, except as an external source for cross-checking.

[29] About 15 percent of industry output values were missing in each of the IndStat2 and FAO databases. Missing values were estimated using output values in years neighboring 201013 (using linear interpolation), or values of trade and total output (in which case the share of industry output in total output is assumed to be equal to the share of industry exports in total exports). None of the U.S. or Cuban industry output values were obtained using this procedure. Also note that output values are not needed to estimate the gravity equation (4).

[30] The values of outputs of other countries do not affect estimates of in (9) because values of , which are calculated using data on output, are cancelled out by values of , which are part of and which can be seen in equations (5) and (6).

[33] In addition to meats, some oil products may be double-counted as well, but that amount is small relative to the total.

[34] Negative values for trade costs are common for imports of wheat and soybeans to Cuba.

[35]Mexico and Canada also face low costs in several sectors, such as tobacco, leather, and petroleum.