Micro-macro links

Research articles · Dec 2, 2024

The limits of Policymaking - An analysis of the consequences of boundedly rational government using the Swedish Micro-to-Macro model (MOSES)

Gunnar Eliasson, Erol Taymaz

Industrial Institute for Economic and Social Research (IUI), Stockholm, Sweden; Middle East Technical University, Ankara, Turkey

Volume 17(2) Summer 2024

Cite as: G. Eliasson, E. Taymaz; 2024; The limits of Policymaking - An analysis of the consequences of boundedly rational government using the Swedish Micro-to-Macro model (MOSES); International Journal of Microsimulation; 17(2); 182–201. doi: 10.34196/ijm.00295

Abstract

The dynamic complexity of the economic system makes it difficult for policy makers to control the outcome of their policies. One reason is that an economy is populated by agents with autonomous price setting leverage and with individual agendas that may run counter to the ambitions of policymakers (Eliasson, 1985). This is particularly so when it comes to policies that influence relative prices and production structures, and therefore also the market self-coordination mechanisms of the economy that regulate short- and long-term intergenerational tradeoffs. Hence, boundedly rational Governments must attend not only to the intended outcomes of their policies, but also to unintended, and not controlled for, side effects in the wider economy. In this paper this wider economy is represented by the Swedish Micro – to – Macro (MM) model, the market self-coordinating mechanisms of which are sufficiently complex to make all standard macro forecasting and policy models biased predictors of MM model performance.

The Government uses the forecasting model of Antonov and Trofimov (1991) and simple rules of thumb to attempt to control certain objective variables generated by the Swedish MM model economy. We study goal satisfaction and side effects. The objective policy variable targeted is unemployment that Swedish Government attempts to keep at the low Swedish rate through public recruitment based on various forecasts and using different model specifications. One side effect that we study is the consequence on macro-economic growth over the very long term. We find that the Government generally fails to forecast and to attain its unemployment targets. In the short- and medium-term Keynesian demand policies to raise employment may temporarily raise macro-economic growth and employment. In the process of doing short term good, the Government may however reduce long-term economic growth, because it misunderstands the consequences of short-term feedbacks on, and market reactions of agents, a Stockholm School feature of the model. In the very long run (beyond 25 years) these consequences tend to cumulate and macro-economic growth is significantly affected. Thus, a seemingly benevolent Government may cause permanent long-term harm by slowly transforming a once wealthy economy into a relatively poor state by consistently, year after year aiming for very short-term targets. Hence the social costs of maintaining an unreasonably low open unemployment, reducing economically motivated reallocations of labor for a long time, will be carried by the next generation. Since these long-term effects depend on the inability of Government – using inadequate forecasting and policy tools – to understand long term economy wide market dynamics the conventional policy conclusion changes to: do less. We concludewith a general discussing how to interpret very long-term simulations by quarter for up to or more than 50 years.

This article contains 7 figures and 3 tables.

Figures

Figure 1. Annual change in manufacturing wages.
Figure 2. Labour productivity.
Figure 3. Unemployment rate.
Figure 4. Manufacturing output.
Figure 5. (a) Rate of return distribution, ZEROBASE. (b) Rate of return distribution, BASE. (c) Rate of return distribution, SIMP 2.
Figure 6. Interest rate.
Figure 7. Public deficit/income ratio.

Tables

ExperimentUnemployment TargetForecasting methodNet public hiring per q.
BASE (Calibrated)NoneNone0.21 percent
EQBASENoneNone0.17 percent
ZEROBASENoneNone0 percent
SIMP11 percentadaptiveDetermined by forecast
SIMP22 percentadaptiveSame
KEY22 percentKeynesian by SBSame
RAT22 percentRESame
Table 1. Experimental results (Years 1-15).
ZEROBASEEQBASEBASESIMP1SIMP2KEY2RAT2
QTOP*150281152394153153156675134675146823141347
TEC*11662116251191711981115361190212052
Labor prod.*389439385954394917797181520333659736547376
Average annual growth rates
GNP3.774.064.115.374.415.044.57
0.740.690.750.720.580.830.67
Output5.045.035.374.704.765.125.27
1.341.541.402.121.571.801.66
Sales11.2311.2611.5710.9611.0011.3611.49
1.151.401.421.871.511.681.65
Prices5.875.885.885.875.915.875.91
0.530.530.520.540.550.540.54
Wages6.577.857.9314.0410.2812.4610.56
0.351.011.362.260.992.100.95
Profit margin45.0642.7043.0937.2039.6637.3839.52
6.163.793.751.722.061.372.44
Unemployment rate8.045.195.281.062.241.852.14
0.951.922.070.190.270.900.32
Exchange rate*1.591.721.712.032.032.032.03
Interest rate12.6312.6812.5910.2812.0510.8311.89
0.870.570.561.260.691.740.95
Public deficit/income-28.33-29.96-29.77-42.48-36.71-40.56-36.62
6.183.934.244.340.841.900.80
Gov. share in GNP35.3336.2936.2244.7540.9944.0041.02
0.570.390.385.082.713.752.52
Trade def./GNP ratio4.315.195.1215.0010.6314.4310.85
2.231.271.386.022.624.572.63

Note: The standard deviation of a variable over the period is shown in the second row.

denotes the value at the end of the period.

Table 2. Experiment results (Years 16-30).
ZEROBASEEQBASEBASESIMP1SIMP2KEY2RAT2
QTOP*4546403919804084934139715904953993222153
TEC*17992191551884022463161431993417895
Labor prod.*7223517197297298797882048250311011372922739
Average growth rates
GNP3.774.064.115.374.415.044.57
0.740.690.750.720.580.830.67
Output3.113.292.83..1.802.801.16
3.092.282.40..1.36...2.33
Sales9.369.669.217.238.5211.997.79
2.252.052.0723.121.7916.262.13
Prices5.875.996.018.006.546.816.49
0.620.570.583.590.862.600.95
Wages5.147.257.7111.759.349.638.93
0.580.820.641.450.201.231.01
Profit margin70.6957.0756.548.4850.2931.0154.28
6.585.604.7667.574.0949.604.30
Unemployment rate13.464.604.130.982.081.912.02
1.310.700.540.430.201.070.24
Exchange rate*1.172.062.203.733.733.723.72
Interest rate8.5912.6813.3722.3321.6023.3321.95
2.520.400.933.724.182.793.94
Public deficit/-2.21-19.19-21.92-60.50-41.86-55.78-43.86
income7.153.060.975.182.215.302.21
Gov. share in GNP35.2839.0939.9166.0152.4158.9051.69
0.681.071.6513.854.074.923.57
Trade deficit/-4.032.433.3247.0523.9432.7421.43
GNP ratio2.290.871.0428.936.057.624.77
Public emp*44.7554.5256.6198.2478.0692.6180.82
Private emp.*41.0139.4538.441.0120.335.6917.75
N new firms*72545724523751
N nullified firms*4250521766915480

Note: The standard deviation of a variable over the period is shown in the second row.

Labor force at the end of simulation is about 2530000.

denotes the value at the end of the period.