Merge pull request #27 from econ-ark/ReparameterizeBeqMotive

Reparameterize bequest motive

src/do_all.py

@@ -136,7 +136,9 @@ def run_replication():
     else:
         print("Invalid replication choice.")
         return
-
+
+    if subjective_markets == "":
+        subjective_markets = "1"
     if int(subjective_markets) > 1:
         agent_name += "Sub"
 

src/estimark/estimation.py

@@ -87,6 +87,9 @@ def make_agent(agent_name):
     if "Portfolio" in agent_name:
         track_vars += ["Share"]
     agent.track_vars = track_vars
+    if "WarmGlow" in agent_name:
+        agent.BeqMPC = 1.0 # dummy value
+        agent.BeqInt = 1.0
 
     agent.name = agent_name
 
@@ -248,9 +251,13 @@ def simulate_moments(params, agent=None, emp_moments=None):
         agent.TranShkStd = init_subjective_labor["TranShkStd"]
         agent.PermShkStd = init_subjective_labor["PermShkStd"]
         agent.update_income_process()
-
+
+    # Update parameters on the agent / construct them
     agent.update()
-
+    if "WarmGlow" in agent.name:
+        agent.BeqFac = agent.BeqMPC**(-agent.CRRA)
+        agent.BeqShift = agent.BeqInt / agent.BeqMPC
+
     # Solve the model for these parameters, then simulate wealth data
     agent.solve()  # Solve the microeconomic model
 
@@ -881,8 +888,8 @@ def estimate(
 if __name__ == "__main__":
     # Set booleans to determine which tasks should be done
     # Which agent type to estimate ("IndShock" or "Portfolio")
-    local_agent_name = "WealthPortfolio"
-    local_params_to_estimate = ["CRRA", "WealthShare", "WealthShift"]
+    local_agent_name = "WarmGlowPortfolio"
+    local_params_to_estimate = ["CRRA", "BeqMPC", "BeqInt"]
     local_estimate_model = True  # Whether to estimate the model
     # Whether to get standard errors via bootstrap
     local_compute_se_bootstrap = False

src/estimark/options.py

@@ -2,7 +2,7 @@
 # resource requirements:
 from __future__ import annotations
 
-params_to_estimate = ["CRRA", "BeqShift", "BeqFac", "WealthShare"]
+params_to_estimate = ["CRRA", "BeqMPC", "BeqInt", "WealthShare"]
 
 low_resource = {
     "estimate_model": True,

src/estimark/parameters.py

@@ -81,11 +81,11 @@
 init_WealthShift = 0.0  # Initial guess of the wealth shift parameter
 bounds_WealthShift = [0.0, 100.0]  # Bounds for the wealth shift parameter
 
-init_BeqFac = 1.0  # Initial guess of the bequest factor
-bounds_BeqFac = [0.0, 100.0]  # Bounds for the bequest factor
+init_BeqMPC = 0.1  # Initial guess of the pseudo MPC of the bequest motive
+bounds_BeqMPC = [0.0, 1.0]  # Bounds for the pseudo-MPC of the bequest motive
 
-init_BeqShift = 0.0  # Initial guess of the bequest shift parameter
-bounds_BeqShift = [0.0, 70.0]  # Bounds for the bequest shift parameter
+init_BeqInt = 1.0  # Initial guess of the pseudo intercept of the bequest motive
+bounds_BeqInt = [0.0, 10.0]  # Bounds for the pseudo intercept of the bequest motive
 
 ######################################################################
 # Constructed parameters
@@ -141,24 +141,24 @@
         "DiscFac": init_DiscFac,
         "WealthShare": init_WealthShare,
         "WealthShift": init_WealthShift,
-        "BeqFac": init_BeqFac,
-        "BeqShift": init_BeqShift,
+        "BeqMPC": init_BeqMPC,
+        "BeqInt": init_BeqInt,
     },
     "upper_bounds": {
         "CRRA": bounds_CRRA[1],
         "DiscFac": bounds_DiscFac[1],
         "WealthShare": bounds_WealthShare[1],
         "WealthShift": bounds_WealthShift[1],
-        "BeqFac": bounds_BeqFac[1],
-        "BeqShift": bounds_BeqShift[1],
+        "BeqMPC": bounds_BeqMPC[1],
+        "BeqInt": bounds_BeqInt[1],
     },
     "lower_bounds": {
         "CRRA": bounds_CRRA[0],
         "DiscFac": bounds_DiscFac[0],
         "WealthShare": bounds_WealthShare[0],
         "WealthShift": bounds_WealthShift[0],
-        "BeqFac": bounds_BeqFac[0],
-        "BeqShift": bounds_BeqShift[0],
+        "BeqMPC": bounds_BeqMPC[0],
+        "BeqInt": bounds_BeqInt[0],
     },
 }