Configuration file

The configuration file is written in YAML syntax and has a .yml extension. By default, AMES assumes the configuration file will be called AMES_params.yml, but other names are possible, and in fact encouraged, because each configuration file corresponds to a different scenario.

Online YAML checkers

YAML is a widely-used language for representing data. As explained on the official YAML website, it is a "human-friendly data serialization language for all programming languages." Because YAML is so widely used, many online tools are available. While AMES will report an error if it finds a problem with your YAML script, the online syntax checkers are more user-friendly, including the YAML checker and the YAML validator.

The configuration file used as an example on this page is the AMES_params.yml file that is distributed with the demonstration files (see the Quick start page).

Configuration file

General settings

The configuration file is made up of several blocks. The first block names a subfolder for storing outputs. It will be created inside an outputs folder (see Output files). Different configuration files should specify different output folders so that scenarios can be distinguished.

#---------------------------------------------------------------------------
# Folder inside the "outputs" folder to store calibration, results, and diagnostics
#---------------------------------------------------------------------------
output_folder: Baseline

The next block sets the start and end years for the simulation. When running with LEAP, these should normally be the same as the base year and final year in LEAP, and the start year should be appropriate for the supply-use table. However, during model development, the start year might be set to an earlier value to calibrate against historical data.

#---------------------------------------------------------------------------
# Start and end years for the simulation
#   Note: the supply-use table above must be appropriate for the start year
#---------------------------------------------------------------------------
years:
    start:  2010
    end:    2040

Required input files

The next block specifies the required external parameter files. Other, optional input files are described below. In many cases, the input files will be the same across a set of scenarios. However, it is possible that they might differ. For example, the supply-use table, given by the SUT parameter, could be drawn from different years for calibration purposes, and different time_series might distinguish different scenarios.

This block also contains a flag saying whether the exchange rate time series, in the time_series file, is for the nominal or real exchange rate. This line can be omitted, in which case the default is for a nominal exchange rate. To specify a real exchange rate, set this equal to true.

#---------------------------------------------------------------------------
# Supply-use table and supplementary tables as CSV files
#---------------------------------------------------------------------------
# AMES model data files
files:
    SUT: Freedonia_SUT.csv
    sector_info: sector_parameters.csv
    product_info: product_parameters.csv
    time_series: time_series.csv
    xr-is-real: false

Output folders

The following block says how to manage the output folders.

Ordinarily it is useful to clear the folders with each run, since files will in any case be overwritten. But sometimes during model development it is useful to set clear-folders to false temporarily – for example, if only a few files need to be compared between one run to the next. To compare results from different runs, make a copy of output files of interest and set clear-folders to false, or copy the entire output file folder.

Reporting diagnostics is optional. It is highly recommended while developing a model, but report-diagnostics can be set to false for a model in active use.

# Say whether to clear the contents of the results, calibration, and diagnostic folders before the run
clear-folders:
    results: true
    calibration: true
    diagnostics: true
# Set to "true" to send results to the diagnostics folder, including dumps of the linear goal program
report-diagnostics: true

LEAP run settings

The next block is optional, as are each of the individual entries. It has settings for running LEAP.

To develop the AMES model independently of LEAP, set run_leap to false. This is the default value, and if it is set to false then the other parameters are unused.

To reduce the time spent while running LEAP and AMES together, set hide_leap to true (the default is false).

The max_runs parameter can normally be set to 5 (the default value). In most models, convergence is reached after 2-3 iterations. However, it is better to be safe, so a higher value is used in the Freedonia model.

The max_tolerance is the maximum allowable percentage difference in AMES model outputs between runs. The tolerance can be tightened (a smaller value) or loosened (a larger value) depending on the needs of the analysis. The default value is 5.0%. That is easily achieved in the Freedonia example, so a tighter value is entered into the Freedonia configuration file.

model:
    # Set run_leap to "false" to do a single run of the AMES model without calling LEAP
    run_leap: false
    # Hide LEAP while running to (possibly) improve performance
    hide_leap: false
    # Maximum number of iterations before stopping (ignored if run_leap = false)
    max_runs: 7
    # Tolerance in percentage difference between values for indices between runs
    max_tolerance: 1.0 # percent

Optional input files

The next block is optional, and can be entirely omitted. This block allows for additional exogenous time series that might be of interest in some studies. Any of them can be added or excluded individually. The allowed input files include:

investment: A time series of exogenous investment demand beyond that simulated by the model (e.g., public investment)
pot_output: Potential output, which will override the value simulated by the model, where the entered values are converted to an index (e.g., agricultural production might be determined by an independent crop production model)
max_utilization: Maximum capacity utilization (e.g., if an exogenous constraint prevents operation at full capacity)
real_price: Real prices for tradeables; the entered values are converted to an index by AMES and multiplied by an index of inflation at the user-specified world inflation rate

The format and purpose of these files are explained in more detail under external parameter files, and examples of each type of file are included in the sample Freedonia model.

Because this block is optional, it can be deleted entirely or commented out (as in the Freedonia configuration file). Alternatively, individual files can be commented out or set to "~", which is the way to signal a missing value in YAML.

#---------------------------------------------------------------------------
# Optional input files for exogenous time series (CSV files)
#---------------------------------------------------------------------------
# Uncomment the lines below to use the files included in the Freedonia sample model
# For pot_output and max_utilization, include only those sectors where values are constrained -- the others will be unconstrained
# For real_price:
#   * Include only those products where values are specified -- for others the real price will be held constant
#   * Prices for non-tradeables will be ignored; they are calculated internally by AMES
exog-files:
    # investment: exog_invest.csv # Time series of exogenous investment demand, additional to that simulated by the model
    # pot_output: exog_pot_output.csv # Potential output (any units -- it is applied as an index): sectors label columns; years label rows
    # max_utilization: exog_max_util.csv # Maximum capacity utilization: sectors label columns; years label rows (must lie between 0 and 1)
    # real_price: exog_price.csv # Real prices for tradeables (any units -- it is applied as an index): products label columns; years label rows

Model parameters

The next several blocks contain some of the exogenous parameters for the AMES model that are not specified in the files identified in earlier blocks.

Initial value adjustments

The first block of model parameters is optional. It specifies adjustments to initial values. Each parameter defaults to 0.0.

These parameters can be used to make a correction if, for example, the economy was recovering from a recession in the first year of the simulation. These parameters should normally be set to zero when starting a calibration (the default value).

#---------------------------------------------------------------------------
# Adjustment parameters for initializing variables
#---------------------------------------------------------------------------
# These factors are expressed as a fractional addition/subtraction to the estimate: set to zero to accept the default estimate
calib:
    # Calibration factor for estimating the first-period profit rate and capital productivity
    nextper_inv_adj_factor: 0.00
    # Adjustment factor for maximum export demand in the initial year
    max_export_adj_factor: 0.00
    # Adjustment factor for maximum household demand in the initial year
    max_hh_dmd_adj_factor: 0.00
    # Potential output relative to actual output in the base year
    pot_output_adj_factor: 0.05

Default values for the global economy

The second block of model parameters provides default values for the global inflation rate and growth rate of global GDP (or gross world product, GWP). These are applied if they are not specified for some year in the external time-series file. Also, the default global inflation rate is applied to price indices when initializing the model.

The correspondence between the parameters and the model variables is:

infl_default : $\underline{\pi}_{w,k}$ (applied uniformly to all products $k$, modified by the optional real_price file in the exog-files block)
gr_default : $\underline{\gamma}^\text{world}$

#---------------------------------------------------------------------------
# Global economy parameters
#---------------------------------------------------------------------------
global-params:
    # Default world inflation rate
    infl_default: 0.02
    # Default world growth rate
    gr_default: 0.015

Taylor rule coefficients

The next block contains the parameters for implementing a "Taylor rule", which adjusts the central bank interest rate in response to inflation and economic growth. See the page on model dynamics for details.

The correspondence between the parameters and the model variables is:

neutral_growth_band : $[\hat{\underline{Y}}^*_\text{min},\hat{\underline{Y}}^*_\text{max}]$
target_intrate :
- init : $\underline{i}^\text{init}_{b0}$
- band : $[\underline{i}^\text{min}_{b0},\underline{i}^\text{max}_{b0}]$
- xr_sens : $\underline{b}_\text{xr}$
- adj_time : $\underline{T}_\text{xr}$
target_infl : $\underline{\pi}^*$
init_infl : $\underline{\pi}_d^\text{init}$
gr_resp : $\underline{\rho}_Y$
infl_resp : $\underline{\rho}_\pi$

#---------------------------------------------------------------------------
# Parameters for setting the central bank lending rate (Taylor rule)
#---------------------------------------------------------------------------
taylor-fcn:
    # Allowable range for neutral growth
    neutral_growth_band: [0.02, 0.06]
    # Target interest rate (as a fraction, e.g., 2%/year = 0.02)
    target_intrate:
        init: 0.04
        band: [0.01, 0.10]
        xr_sens: 1
        adj_time: 2 # years
    # Target inflation rate (if missing or if value set to "~", will use global inflation rate)
    target_infl: 0.02
    # Initial inflation rate for domestic prices (if missing, or if value set to "~", will use target inflation rate)
    init_infl: 0.04
    # Response of the central bank rate to a change in the GDP growth rate
    gr_resp: 0.50
    # Response of the central bank rate to a change in the inflation rate
    infl_resp: 0.50

Investment function coefficients

The next block contains the parameters for the investment function. As explained in the page on model dynamics, the investment function calculates the demand for investment goods as a function of capacity utilization, profitability, borrowing costs, and (optionally) the current account-to-GDP ratio.

The profit rate in the investment function

Prior to version 2.2.8, the profit rate used in the investment function was calculated at full capacity utilization. However, this can lead to unexpected behavior. The default is now to use the profit rate calculated in terms of realized profits, taking capacity utilization into account. Because this change produces different results, an optional parameter, use_profits_at_full_capacity, can be set to true to revert to the prior behavior. This parameter can be omitted, and its default value is false.

The correspondence between the parameters and the model variables is:

init_neutral_growth: Initial value of $\gamma_{i0}$ (with the same initial value for every sector $i$)
util_sens : $\underline{\alpha}_\text{util}$
profit_sens : $\underline{\alpha}_\text{profit}$
intrate_sens : $\underline{\alpha}_\text{bank}$
net_export : $\underline{\alpha}_\text{netx}$
growth_adj : $\underline{\xi}$

#---------------------------------------------------------------------------
# Parameters for the investment function
#---------------------------------------------------------------------------
investment-fcn:
    # Starting point for the autonomous investment growth rate: also the initial target GDP growth rate for the Taylor rule
    init_neutral_growth: 0.060
    # Change in induced investment with a change in utilization
    util_sens:  0.07
    # Change in induced investment with a change in the profit rate
    profit_sens: 0.05
    # The profit rate is based on realized profits by default; leave blank or set to true for profits at full capacity utilization
    use_profits_at_full_capacity: false
    # Change in induced investment with a change in the central bank lending rate
    intrate_sens: 0.20
    # Change in induced investment with a change in the net export-to-GDP ratio
    net_export: 0.00
    # Rate of adjustment of the autonomous investment rate towards the actual investment rate
    growth_adj: 0.10

Employment, labor productivity, and wages

The next block contains default parameters for labor productivity (either a constant growth rate or in the form of the Kaldor-Verdoorn law) and for the function that determines the growth rate of the wage. See the technical documentation on labor productivity growth and wage determination for details.

All parameters are optional. Their definitions are as follows:

use_KV_model (default = true if KV_coeff_default or KV_intercept_default is defined, otherwise false) : A flag to say whether to use the Kaldor-Verdoorn model rather than constant labor productivity growth rates;
use_sector_params_if_available (default = true) : A flag to say whether to use sector-specific parameters if they are available (useful mainly to compare different specifications, otherwise it can be omitted);
labor_prod_gr_default (default = 0.0) : Default economy-wide labor productivity growth if not specified for some year (or if the column is omitted) in the external time-series file;
KV_coeff_default (default = 0.5) : Default economy-wide labor productivity growth if not specified for some year (or if the column is omitted) in the external time-series file;
KV_intercept_default (default = 0.0) : Default economy-wide labor productivity growth if not specified for some year (or if the column is omitted) in the external time-series file;
infl_passthrough (default = 1.0) : Degree of inflation pass-through (the default is full pass-through);
lab_constr_coeff (default = 0.0) : Response of the wage level to labor constraints (the default is no response).

The correspondence between the parameters and the model variables is:

For labor productivity:
- KV_coeff_default : $\underline{\alpha}^\text{KV}$
- KV_intercept_default : $\underline{\beta}^\text{KV}$
- labor_prod_gr_default : $\underline{\beta}^\text{KV}$ (with $\underline{\alpha}^\text{KV} = 0.0$)
For the wage:
- infl_passthrough : $\underline{h}$
- lab_constr_coeff : $\underline{k}$

#---------------------------------------------------------------------------
# Parameters for labor productivity, labor force, and wages
#---------------------------------------------------------------------------
labor-prod-fcn:
    # Flag for whether to apply Kaldor-Verdoorn model
    use_KV_model: true
    # Flag for whether to use sector parameters
    use_sector_params_if_available: true
    # Default labor productivity growth
    labor_prod_gr_default: 0.025
    # Default Kaldor-Verdoorn coefficient
    KV_coeff_default: 0.500
    # Default Kaldor-Verdoorn intercept
    KV_intercept_default: 0.005
wage-fcn:
    # Inflation pass-through (wage indexing coefficient)
    infl_passthrough: 1.00
    # Labor supply constraint coefficient
    lab_constr_coeff: 0.50

To simply use a fixed labor productivity growth rate, the following block will suffice:

#---------------------------------------------------------------------------
# Parameters for labor productivity, labor force, and wages
#---------------------------------------------------------------------------
labor-prod-fcn:
    # Default labor productivity growth
    labor_prod_gr_default: 0.025
wage-fcn:
    ...

Sector values will be used if they are specified, otherwise the time series values will be used. If time series values are not specified, then the default value will be applied.

Endogenous change in intermediate demand coefficients

The next block is optional. If it is present, it sets parameters for endogenously determining intermediate demand coefficients. These parameters can optionally be set by sector in the sector parameters file

Note that the calculation can generate unreasonably large rates of change in technical coefficients and can even create model instabilities. It is good practice to check the diagnostics folder for annual files labeled demand_coefficients_[year].csv to see whether the values are reasonable. If they are not, then adjust the parameters, either here or in the sector parameters file.

#---------------------------------------------------------------------------
# Parameter for rate of change in technical parameters
#---------------------------------------------------------------------------
tech-param-change:
    # Flag to say whether to calculate changing technical parameters
    calculate: false
    # Flag for whether to use sector parameters
    use_sector_params_if_available: true
    # Rate constant: Note that if this is too large then it can create instabilities
    rate_constant_default: 1.5 # 1/year
    # Exponent in the technical change function
    exponent_default: 2.0

Long-run demand elasticities

Initial values for demand elasticities for products with respect to domestic vs. world prices, global GDP (for exports), and the wage bill (for domestic final demand excluding investment) are specified in the external product parameters file. The way that the elasticities enter into the model is described in the page on model dynamics.

For products that are not labeled as "Engel products"^[1] (given by the parameter engel-prods):

If the initial wage elasticity is less than one, then it remains at its starting level;
If the initial wage elasticity is greater than one, it asymptotically approaches a value of one over time;

For products labeled as Engel products:

The wage elasticity approaches the engel_asympt_elast over time.

The rate of convergence on the long-run values is given by the decay parameter, with the possibility of a different rate of convergence for exports and final demand.

#---------------------------------------------------------------------------
# Demand model parameters
#---------------------------------------------------------------------------
# For exports, with respect to world GDP
export_elast_demand:
    decay: 0.01
    
# For final demand, with respect to the wage bill
wage_elast_demand:
    decay: 0.01
    engel_prods: [p_agric, p_foodpr]
    engel_asympt_elast: 0.7

Linear goal program weights

The AMES model solves a linear goal program for each year of the simulation. As described in the documentation for the linear goal program, the objective function contains weights, which are specified in the next block. The default weights should be suitable for most AMES applications.

The correspondence between the parameters and the model variables is:

For category weights:
- utilization : $\underline{w}_u$
- final_demand_cov : $\underline{w}_F$
- exports_cov : $\underline{w}_X$
- imports_cov : $\underline{w}_M$
For product and sector weights:
- utilization : $\underline{\varphi}_u$
- final_demand_cov : $\underline{\varphi}_F$
- exports_cov : $\underline{\varphi}_X$

#---------------------------------------------------------------------------
# Paramters for implementing the (goal program) obective function
#---------------------------------------------------------------------------
objective-fcn:
    # Category weights on deviations from normal levels
    category_weights:
        utilization: 8.00
        final_demand_cov: 4.00
        exports_cov: 2.00
        imports_cov: 1.00
    # Product & sector weights are defined by: φ * value share + (1 - φ) * 1/number of sectors or products; this is φ
    product_sector_weight_factors:
        utilization: 0.5
        final_demand_cov: 0.5
        exports_cov: 0.5

Linking to the supply-use table

The next block is for specifying the structure of the supply-use table and how it relates to variables in AMES.

The first section of this block specifies sectors and products that are excluded from the simulation. The entire section or any item can be excluded. Alternatively, items can be set to the empty list [] or to YAML's "no value" symbol, ~.

There are three categories:

First, and most important, are energy sectors and products. Those are excluded from the AMES calculation because the energy sector analysis is handled on a physical basis within LEAP, although they can optionally be included when running the model in stand-alone mode, without LEAP.
Second are any territorial adjustments. AMES recalculates some parameters to take account of those entries. If none are present in the supply-use table, then an empty list [] can be entered for this parameter, as in the sample Freedonia model file shown below.
Finally are any other excluded sectors and products. For example, some tables may have a "fictitious" product or sector entry.

#---------------------------------------------------------------------------
# Structure of the supply-use table
#---------------------------------------------------------------------------
# NOTE: Both supply and use tables must have sectors labeling columns and products labeling rows
excluded_sectors:
    energy: [s_coal, s_petr, s_util]
    territorial_adjustment: []
    others: []

excluded_products:
    energy: [p_coal, p_petr, p_util]
    territorial_adjustment: []
    others: []

Following that is a list of non-tradeable products and a domestic production share threshold. Imports and exports of products declared non-tradeable are set to zero within the AMES model, if they are not already zero in the supply-use table, and are maintained at zero throughout the simulation.

Other products may be almost entirely imported. That can sometimes cause difficulties. If the domestic share of the total of imports and domestic production falls below the threshold specified in the configuration file, then the corresponding sector (but not the product) is excluded during the simulation.

non_tradeable_products: [p_constr, p_comm]

# Domestic production as a % share of the total of imports and domestic production must exceed this value
domestic_production_share_threshold: 1.0 # percent

The following section specifies where to find the data needed by AMES within the supply-use table (a CSV file). Ranges are specified in standard spreadsheet form.

SUT_ranges:
    # Matrices arranged product (rows) x sector (columns)
    supply_table: J3:W16
    use_table: J21:W34
    # Columns indexed by products -- groups of columns will be summed together
    tot_supply: I3:I16
    margins: D3:E16
    taxes: F3:H16
    imports: Y3:Y16
    exports: Y21:Y34
    final_demand: Z21:AB34
    investment: AC21:AC34
    stock_change: AD21:AD34
    tot_intermediate_supply: X21:X34
    # Rows indexed by sector -- groups of rows will be summed together
    tot_intermediate_demand: J35:W35
    wages: J37:W38

Mapping AMES to LEAP

The next, and final, block specifies how LEAP and AMES are linked. Each of these entries is optional.

Core LEAP model information

The first section in this block says which LEAP scenario to send and retrieve results to and from, and for which region. It also identifies the first historical year, if that is different from the start year. (The last_historical_year is the year just before LEAP's First Scenario Year.)

If this section is omitted, then results are sent to and retrieved from the scenario currently active in LEAP, and for the currently active region (if any regions are specified). The last_historical_year is set to the start year.

#---------------------------------------------------------------------------
# Parameters for running LEAP with the AMES model
#---------------------------------------------------------------------------
# Core information for the LEAP application (optional)
LEAP-info:
    # The last historical year (equal to LEAP's First Scenario Year - 1): if missing, it is set equal to the start year
    last_historical_year: 2010
    # This can be, e.g., a baseline scenario (alternatively, can specify input_scenario and result_scenario separately)
    scenario: Baseline
    # The region (if any -- can omit, or enter a "~", meaning no value)
    region: ~

Alternatively, separate scenarios can be specified if LEAP receives inputs from AMES for one scenario (input_scenario), but sends results back to AMES from another scenario (result_scenario). In this case, scenario should be omitted, or set to ~ (meaning "no value"). For example,

#---------------------------------------------------------------------------
# Parameters for running LEAP with the AMES model (AMES)
#---------------------------------------------------------------------------
# Core information for the LEAP application (optional)
LEAP-info:
    # The last historical year (equal to LEAP's First Scenario Year - 1): if missing, it is set equal to the the start year
    last_historical_year: 2010
    # Scenarios
    input_scenario: Baseline
    result_scenario: Capital Plan
    ...

Pulling investment information from LEAP into AMES

The next section specifies information regarding investment. By default, any investment reported in LEAP in a given year is totaled and passed to AMES with no adjustment: inv_costs_units is set to a blank (so LEAP applies the default currency unit), inv_costs_scale is set to 1.0, and inv_costs_apply_xr is set to false. However, some investment branches can be omitted, investment can be spread over several years, the currency unit can be set to a different value, the scaling factor can be specified, and the exchange rate can be applied.

As an example for setting the scaling factor, if entries in the AMES model input files are in millions of US dollars, and investment costs in LEAP are reported in US dollars, then the scaling factor is one million (1000000 or 1.0e+6).

If inv_costs_apply_xr is true, then the nominal exchange rate will be applied to the values from LEAP after scaling. In this calculation, the exchange rate will not be applied as an index: it is assumed that the value entered into the time series column exchange_rate is the actual nominal exchange rate. If that is not the case, then the scaling factor can be adjusted to ensure the correct units.

The excluded_branches entry is set to an empty list [] in the Freedonia example below. This entry is particularly useful if NEMO is run with a backstop technology. Any "investment" in the backstop technology represents a desired supply expansion that was not achieved, so the value should be ignored.

The Freedonia configuration file below shows different ways to specify how costs are spread over multiple years. If this entry is blank, then the default is that all expenditure occurs in one year. Otherwise, a default value can be set that is applied to all branches. Additionally, a value or pattern can be set for specific branches.

The distribution of costs over years can be set in one of two ways. If it is a single number (e.g., in example below, the default is 5 years), then total investment is divided equally across all of the years. If expenditure is set as a pattern, then it is specified as a list of values in brackets []. The values in the list are then rescaled by AMES so that they sum to one. The example below for New Oil Combustion Turbine sums to 100, so 10% of expenditure will be applied in the first year, 20% in the second, and so on.

LEAP-investment:
    # Currency units for investment costs
    inv_costs_unit: U.S. Dollar
    # Scaling factor for investment costs (values are divided by this number, e.g., for thousands use 1000 or 1.0e+3)
    inv_costs_scale: 1.0e+6
    # Say whether to apply the nominal exchange rate to investment costs (e.g., if investment costs are in USD, but the SUT is in domestic currency)
    inv_costs_apply_xr: false
    # Exclude any investment branches that contain any of the text in the list (case-insensitive)
    excluded_branches: []
    distribute_costs_over:
        default: 5 # years: This will be rounded to an integer
        by_branch:
            - {
                path: Transformation\Electricity Generation\Processes\Existing Hydro,
                value: 10 # years: This will be rounded to an integer
            }
            - {
                path: Transformation\Electricity Generation\Processes\New Oil Combustion Turbine,
                value: [10, 20, 20, 20, 10, 10, 5, 5] # percent by years: This will be re-scaled so it sums to 1.0
            }

Passing values from AMES to LEAP

The next sections specify where in LEAP to put indices as calculated by AMES. For each index, LEAP should contain at least one historical value, while the index supplied by AMES is applied to the last historical value in the specified result_scenario. Indices appear as columns in an indices_#.csv file in the results output folder, where # is the run number.

The first index is for GDP. It can be omitted, but if it is present, the entry in the configuration file gives the name for the index, the LEAP branch and variable where the index should be inserted and, to cover cases where the last historical year is after the base year, the last historical year.

# Association between AMES's GDP result and LEAP (optional)
GDP-branch:
    name: GDP
    branch: Key\GDP
    variable: Activity Level

The second index is also optional, and has the same structure as for GDP.

# Association between AMES's employment result and LEAP (optional)
Employment-branch:
    name: Employment
    branch: Key\Employment
    variable: Activity Level

Next, the association between AMES sectors and LEAP sectors is specified. These entries are also optional. They start with a specification of the economic driver for LEAP, either production (the default) or value added. Value added subtracts from the value of production the cost of intermediate goods and services, to avoid double-counting when calculating gross domestic product (GDP).

Production is a better variable to use when the corresponding variable in LEAP is a physical quantity. For example, in the Freedonia model, the activity level for the Iron and Steel sector is given as tonnes of steel. This driver can be set explicitly as PROD. However, in some LEAP models the activity level is value added. In that case, the default driver can be set to VA. The driving variable can be set by branch, as well, as shown in the LEAP-sectors example below.

# Association between AMES sectors and LEAP sectors (optional)
LEAP-drivers:
    options:
        PROD: production
        VA: value added
    default: PROD

the LEAP-sectors parameter, if present, contains a list of indices by sector. If no sectoral aggregation is desired, then this entry can be excluded entirely, set to ~ (meaning no value), or set to an empty list: LEAP-sectors: [].

For each index, AMES will sum up the driver (either production or value added) across all of the sector codes listed. It will then calculate an index starting in the base year, and insert the index into the specified branches. In some cases, the same index might be applied to different branches. For example, if the supply-use table has a "services" sector but no transport, while LEAP has both a services and a commercial transport sector, the same index could be used to drive both.

In the example below, note that Iron and Steel and Other Industry do not have a branch-specific driver, so they use the default, PROD (production). The Commercial entry does have a branch-specific driver, which is set to VA (value added).

LEAP-sectors:
 - {
    name: Iron and Steel,
    codes: [s_ironstl],
    branches: [
        {
         branch: Demand\Industry\Iron and Steel,
         variable: Activity Level
        }
    ]
   }
 - {
    name: Other Industry,
    codes: [s_foodpr, s_hvymach, s_othind],
    branches: [
        {
         branch: Demand\Industry\Other Industry,
         variable: Activity Level
        }
    ]
   }
   ...
 - {
    name: Commercial,
    codes: [s_comm, s_othsrv],
    driver: VA,
    branches: [
        {
         branch: Demand\Commercial,
         variable: Activity Level
        }
    ]
   }

Passing potential output and prices from LEAP to AMES

The final sections say how to pass results for potential output and prices from LEAP to AMES. These are in addition to investment expenditure, which is automatically collected from LEAP and passed to AMES. Any values for potential output and prices drawn from LEAP override those specified in external input files, if any. (Investment expenditure from LEAP is added to investment specified in external input files.)

The Freedonia configuration file does not specify potential output and prices from LEAP, so the entries are set equal to empty lists. Alternatively, they can be completely ommitted or set to the YAML "no value" symbol, ~.

LEAP-potential-output: []

LEAP-prices: []

To demonstrate how these sections might look, suppose that instead of excluding the coal sector as is done in the link to the supply-use table, above, it is included in the AMES model, but potential output from the sector is provided by LEAP. That can be accomplished by setting

excluded_sectors:
    energy: [s_petr, s_util]
excluded_products:
    energy: [p_petr, p_util]

and then adding an entry like

LEAP-potential-output:
 - {
    branches: [
        {
         branch: Resources\Primary\Coal Bituminous,
         variable: Indigenous Production
        }
    ],
    code: s_coal
   }

Potential output is associated with a single sector in the AMES model, but can include values from multiple LEAP branches. If more than one LEAP branch is listed, then the values are summed together.

If the price of coal is available from the LEAP model, then it can be brought into AMES using an entry like

LEAP-prices:
 - {
    branch: Resources\Primary\Coal Bituminous,
    variable: Indigenous Cost,
    codes: [p_coal]
   }

Prices are associated with a single LEAP branch, but can be applied to multiple products in AMES.

1 Engel's Law states that as income rises, the proportion of income spent on food declines. That means that the income elasticity of expenditure on food is less than one.