Page 37 - GHES-3-2

Global Health Economics and
Sustainability
Random price generators in health policy

a recent development; other pricing methods, especially nature is that it is a useful process to achieve neutrality
revealed-preference methods (e.g., hedonic pricing), (Sunstein, 2019), which is important for policymakers,
have prevailed in valuation methodologies used for especially when using “nature nudging” as a policy-aiding tool.
macroeconomic models over contingent valuation Nature randomness may change at any time, so using
methods (e.g. Appendix A1; United Nation [UN], 2003). random variables in modeling with a process approach may
They proved to be statistically reliable in market dynamics fit more data on population health. Economists have run
of liberal economies and health systems primarily using analyses using random parameters for variables in models
price mechanisms. The concept of randomness is not for health or medical markets for a long time; however, this
new in medicine and life sciences. It has been used for paper shows the possible interest of random prices in an
modeling in population studies and their evolution (e.g., economic model for physicians. The methodological steps of
breeding populations are randomly combined in pairs this experimental user case are tested on a statistical model for
to produce offspring). Scientists such as epidemiologists the Type 2 diabetes market in the USA. This type of economic
and biostatisticians have used Markov chains models for model aims to strengthen the analysis of the demand side of
disease progression (e.g., Bernoulli’s principle for simple
cases or Monte Carlo models). a medical market to better represent the heterogeneity of
physicians’ preferences at the individual level.
Using random parameters in statistical models for
economic analysis of medical markets is also widespread; Figure 1 provides the different types of statistical models
however, generating random prices is quite recent for medical used to analyze this medical market; it lists predictive disease
products and services. This paper discusses the implication of econometrics, especially from Professor Huttin’s previous
the experimental research for an economic physicians’ choice studies and more recent projects for oligopoly markets.
model initiated by Huttin & Hausman (2021) on patients with The current milestone aims to provide a comparative
Type 2 diabetes. In their study, random prices were generated analysis of study results for the latest approaches: choice
for drug prices on an analytical dataset for Type 2 diabetes. models with random price generators and Hierarchical
Their paper also explained the main methodological steps. The Bayesian analysis with shrinkable estimators. According
project provided encouraging results on such an approach; it to the structural models for national economies, facing
implemented the recent econometric specification tests on a open boundaries of science and new players disrupting
mixed logit (ML) model at an individual level, proposed by traditional game rules, these approaches can contribute to
Burda et al. (2012) and Hahn, Hausman & Lustig (2020), it calibration methods for adjusting supply and demand for
complements the generalization of Hausman & McFadden medical markets.
tests for the Multinomial Logit Model (1984). Random
generators for drug prices were computed for this ML 2.2. Experimental research in other sectors: A useful
model. Such development may help to move beyond latent resource for the use of random generators in
class models, existing ML model statistics, and additional healthcare markets
applications to other models for oligopolistic markets. Experimental research using random parameters has
Limitations of this approach have been discussed, expanded greatly, especially in transportation research and
especially concerning the development of pricing methods finance. Studies using random parameters usually show
such as algorithmic pricing in the context of digitalization very good results in such fields, especially in transportation
and mobile economics (Huttin & Hausman, 2021). Further research, marine research, and energy (Glenk et al., 2020),
development of other choice models using random pricing comparing valuation methods such as random parameters
is also needed for a validated simulated tool, especially logit and latent class models. Systematic reviews may help
with comprehensive user cases on choice sets of medicines, further milestones for more comprehensive user cases in
devices, or procedures, and applied to analytical datasets healthcare markets with such an approach; however, they
from different national or regional contexts. 1 are out of the scope of this paper.
The sharing economy also leads to studying the
2. Why is it useful to include randomness in preferences of consumers and populations when adjusting
modeling for policy-aiding tools in health? to the new economy. These include workers’ preferences

2.1. Nature randomness and development of for co-working spaces and hours in labor markets, new
economic models for health alternative sharing models on bicycle or car sharing in
transportation research, preferences for quality and choice
One of the main arguments for using randomness to study
of genetically modified food in nutrition studies, and
1 Grimmett and David Stirzaker (2020) provide a good review of preferences between electric and hybrid cars in automobile
random processes for interested readers. markets.

Volume 3 Issue 2 (2025) 29 https://doi.org/10.36922/ghes.3579