KITE — Kiel Institute Trade Policy Evaluation

---

The KITE model provides a tool for simulating and estimating various types of (trade) policy changes. The underlying model uses a computable general equilibrium (CGE) framework of the type that is commonly described as "New Quantitative Trade Model". The model has originally been developed by Caliendo & Parro (2015). They built a multi-sector version of the Ricardian trade model of Eaton & Kortum (2002), where countries produce and sell domestically as well as internationally according to their relative comparative advantage. The model extends this framework by allowing for extensive intra- and international input-output linkages where goods and services may enter as both final and intermediate goods. Trade policy is conducted through the tightening or easing of trade barriers in form of tariffs and non-tariff measures.

By now the model has been extensively used to evaluate (free) trade agreements (e.g. NAFTA, TTIP) and trade disputes (i.e. US-China trade war, Airbus-Boeing case) and economic sanction regimes. It derives the economic consequences for production, value added and welfare based on pre-defined scenarios that specify the policies to be evaluated. It allows various types of data sources that can be used in constructing the underlying model variables and/or parameters.

The model is constantly updated to improve efficiency and/or extend the underlying framework. Current projects include the implementation of CO2 footage in production or the introduction of the factor land in the production function.

Updates

Version 24.01 allows the user to run different models within the same quantitative framework. Included models are:

• Caliendo, Lorenzo, and Fernando Parro. 2015. “Estimates of the Trade and Welfare Effects of NAFTA.” The Review of Economic Studies 82 (1): 1–44.
• Sonali Chowdhry, Julian Hinz, Katrin Kamin and Joschka Wanner. 2024. "Brothers in arms: The value of coalitions in sanctions regimes." Economic Policy, forthcoming.

Copyright

Copyright 2019-2024 Kiel Institute for the World Economy & Austrian Institute of Economic Research, KITE Development Team

Developers

• Julian Hinz, Kiel Institute, Mail: julian.hinz@ifw-kiel.de

• Hendrik Mahlkow, Austrian Institute of Economic Research, Mail: hendrik.mahlkow@wifo.ac.at

• Joschka Wanner, Kiel Institute, Mail: joschka.wanner@ifw-kiel.de

Installing KITE

KITE is a R package. To install this package, download the current beta version and install using

install.packages("~/path/to/KITE_24.01.tar.gz", repos = NULL, type = "source")

Tutorial

The package comes with a vignette describing different trade policy scenarios. You can load it with the following command (the package needs to be installed):

vignette("KITE") # Trade policy scenarios with KITE

Usage

Running the model requires the user to follow 4 basic steps:

1. Choose an input-output table.
2. Define scenarios that are supposed to be simulated by the model.
3. Feed the data from 1) and 2) into the model and run.
4. Process the results.

The script example.R provides a step-by-step algorithm in performing a scenario analysis with KITE. It sources from different scripts and functions that are embedded in the KITE package and will be explained further below. Please use the script for all your calculations.

Step 1: Choose Input-Output Table

Load the I-O tables from a folder "input/../initial_conditions.rds"

initial_conditions = read_rds("input/../initial_conditions.rds")

Step 2: Set Model Scenarios

Define and set the scenarios for the model run by specifying the origin and destination countries concerned, as well as the counterfactual policy that is supposed to represent the given policy scenario.

Counterfactual Tariff

tariff_war = copy(initial_conditions$tariff)

# US tariffs on Chinese goods and services
tariff_war[origin == "usa" & destination == "chn", value := 1.2]

# Chinese tariffs on US goods and services
tariff_war[destination == "usa" & origin == "chn", value := 1.2]

In this example KITE simulates the US-China trade war, assuming that both countries simultaneously set up a tariff of 20% across all sectors. "value" specifies the tariff as value = 1 + tariff/100.

The user may specify other policies that are supposed to enter the counterfactual situation, such as non tariff barriers or export subsidies. The procedure works exactly the same as in the case of a tariff change. For more detailed information on the options for policy changes please proceed to Step 3. For more examples see vignette("KITE").

Step 3: Run the Model

The function calculates a new equilibrium based on the policy changes the user feeds into the model. The corresponding results are expressed in exact hat algebra, i.e., in discrete changes relating the base-line equilibrium and the counterfactual equilibrium (scenario).

The function rests on four building blocks, all of which need to be specified.

1. model = my_new_model

Here the user needs to specify the model used, e.g., model = caliendo_parro_2015. The model is specified in a respective model_*.R file. The user may also want to specify a new model, in which case the user needs to specify the model in a model_my_new_model.R script and then source it manually.

2. initial conditions = initial_conditions

Specify the initial conditions.

1. model_scenario = list(tariff_new = tariff_war,...)

Here, the user needs to specify all the variables that represent the desired counterfactual/scenario situation. Please note that these variables need to be defined in "Step 2: Set Model Scenarios", above. Options to specify the counterfactual situation include tariffs (tariff_new) and non-tariff barriers (ntb_change). The user may even want to recalculate the baseline equilibrium to change the situation a given counterfactual/scenario is supposed to be compared to. Options include tariffs, non-tariff barriers, consumption share, factor share, intermediate share, trade elasticity, trade share, value-added, trade balance.

3. verbose = 2 and tolerance = 1e-4

Set verbosity levels between 0 and 2 to get more or less detailed output while the model converges. The tolerance level determines the stopping point, 1e-6 being the default and lower numbers leading to more precise results.

The full command should look something like this

results = update_equilibrium(model = chowdhry_hinz_kamin_wanner_2022,
                             initial_conditions = initial_conditions,
                             model_scenario = list(tariff_new = tariff_war,
                                                   coalition_member = c("usa", "can", "mex")),
                             settings = list(verbose = 2L,
                                             tolerance = 1e-4))

Citation

Please cite the package as follows:

@Manual{,
    title = {KITE: Kiel Institute Trade Policy Evaluation Model},
    author = {Julian Hinz and Hendrik Mahlkow and Joschka Wanner},
    year = {2024},
    note = {R package version 24.01},
  }

References

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

Eaton, Jonathan, and Samuel Kortum. 2002. “Technology, Geography, and Trade.” Econometrica 70 (5): 1741–79.

R Core Team. 2020. "R: A language and environment for statistical computing." R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/