SOCIAL SPHERE MODELING
BASED ON SYSTEM DYNAMICS METHODS
Natalia N. Lychkina
State University of Management
99 Ryazansky Avenue, Moscow, Russia, 109542
Tel: +7 (495) 749-71-77 / Fax: +7 (495) 371-05-28
E-mail: lychkina@ guu.ru _http://kafis.quu.ru
Dmitriy L. Andrianov
Perm State University
15 Bukireva Str., Perm, Russia, 614990
Tel: +7 (342) 239-62-58, +7 (342) 240 36 63 / Fax: +7 (342) 240-37-70
E-mail: andrianov@econ.psu.ru _http://ecd.psu.ru/
Yulia A. Morozova
State University of Management
99 Ryazansky Avenue, Moscow, Russia, 109542
Tel.: +7 (916) 828-76-86
E-mail: limnoria@ gmail.com
Abstract
This article describes a complex of social sphere simulation models aimed to
support decision making in the social sphere, address such issues as a housing
reform, public health services and social security. The complex is implemented on the
basis of system dynamics methods and using modern simulation modeling techniques.
Key words: social service, decision support system, socio-economic
development of a region, simulation modeling, system dynamics methods, description
of the model complex, system flow charts.
Introduction
The social sphere comprises such industry systems as health care, physical
culture and sport, education, culture, social services and social security, housing,
social insurance and pensions.
The analysis of the current situation in the social sphere shows that its
governance requires changes. The existing governance system is inadequate under
current conditions: the current economic situation in Russia and rate of a new
govemance system development in Russia at various levels of the industrial and
economic sector.
Only elaborate reforms of the social sphere governance will be able to improve
the current situation in Russia. An essential component of the social sphere reform is
the development of innovative social technologies based on decision support systems
to provide information and analytical support to the government authorities in
addressing social issues. This will also help optimize regional costs, forecast main
socio-economic development indicators for a long-run period, and ensure living
standards and regional economic growth.
The economic analysis of the social sphere is primarily aimed at analyzing the
allocative efficiency of limited financial, material, labor and other resources of
altemative ways of their use to find a solution that is commensurate with the costs of
these resources. The economic analysis should cover characteristics of processes and
relationships associated with activities involving comparison and selection of efficient
solutions in industry systems, integrated forecasting of living standards considering
rates of social service provision. During the review of the economic system of the
social sphere industries it is necessary to keep in mind that that these industries have
specific economic features that distinguish them from other parts of the economy.
They manifest in obvious imperfections in the exchange relations, extensive role of
the government in the social service provision, forms of the public regulation and
financing, their blending with market mechanisms, and prevailing non-profit
organizations.
The analysis of socio-economic processes at the regional level, including its
social sphere, should be conducted using a system approach that allows combining a
set processes taking place in the economic, social, environmental and other sectors of
a region. Making adequate and balance management decisions requires the
consideration of regional features, dynamics of socio-economic processes, evolution
of the system and its components, actual demographic trends, medical and
demographic situation, etc. The socio-economic system under study is viewed as a
complex semistructured system, which modeling involves the identification of a large
number of complex interrelated causal relationships among factors reviewed within
the complex system description and which effects are not always obvious in making
decisions; the description of the modeling object structuring contains a significant
portion of expert knowledge; the study requires taking into account a large number of
stochastic factors under the condition of uncertain initial information. Management
decision analysis tools should allow reviewing multiple alternatives and development
strategies and be efficient in solving industry restructuring tasks.
The authors of this paper made an attempt to develop a set of models to
adequately accommodate the current socio-economic situation in Russia, including
such major social sphere subsystems as health care, housing, social welfare,
education, culture; to describe demographic, economic and financial processes and
solutions to meet contemporary challenges of the social sphere reforming.
Social Sphere M odel C omplex
The Social Sphere Model Complex is a system of models implemented based
on the system dynamics methods that describe relationship of socio-economic
development indicators of a region and its separate social sphere industries and are
oriented on integrated regional social development forecasting based on national
economic management scenarios considering the territorial aspect. Models within the
complex are designed to perform scenario calculations based on expert formulated
strategies with an explicit set of control actions, i.e. those socio-economic indicators
of the social sphere which adjustment falls within the competence of the municipal,
regional and federal authorities. The main objective of the regional management is to
make such management decisions and select such economic management methods to
arrive to such proportions of social reproduction that to the maximum extent help
meet the public needs in a region and raise its living standards.
The Social Sphere model complex allows solving such tasks as developing an
efficient social policy, planning and managing the social sphere (reforming of the
housing and utilities, health care sector, financial planning of the social sphere),
forecasting and comprehensive analysis of living standards across regions. Models
also analyze funding and commissioning of individual capacities within the
educational system and cultural activities in a region at the expense of the budget. The
complex includes an aggregate model of the social sphere, Housing and Utilities,
Health Care and Pension System simulation models.
Social Sphere Aggregated M odel
This model allows forecasting general development trends and situation in the
social sphere in a region as a whole, main socio-economic development indicators of
a region, conducting a comprehensive analysis of living standards and quality of life
by territories in the short and long term.
As regional socio-economic development objectives are used such objectives
as increase of income, improvement of education, nutrition and health care, poverty
reduction, environmental enhancement, provision of equal opportunities, enlargement
of personal liberty, and cultural life enrichment. In line with regional development
objectives is defined a set of criteria and their corresponding regional socio-economic
development indicators:
= Gross regional product (absolute value and per capita)
« Financial indicators of situation in a region, budget income and spending, ratio
of consolidated regional budget income in its full expenditures, relative
income of the federal center, regional budget deficit
= Average household income and its differentiation
= Social support for needy
= Regional unemployment
= Life expectancy, health in a region
= Consumption of goods and services
= Level of health care (provision of outpatient clinics, hospitals, quality of
medical services)
= Provision of housing stock
= Level of education of the population of the region and provision of educational
institutions across all levels of education (primary, secondary, tertiary)
= Cultural life enrichment and provision of culture institutions of the public
= Overall standard of living
In accordance with objectives and tasks specified a number of subsystems
were singled out.
An economic subsystem details regional economic processes, forecasts GRP,
requirement in labor forces, household income and payments to the budget. The
subsystem is based on the Solow macroeconomic model.
A demographic subsystem describes the structure of population with its
division into three age groups, population movement inside a region, birth and death
rates, composition of social groups, unemployment in a region.
A financial subsystem describes the budgetary processes, tax collection,
financial flows, itemized budget income and spending.
The Housing Stock subsystem divides the housing stock into several types
depending on construction factors, housing stock depreciation and evolution,
transition of the municipal housing stock to private ownership to citizens and
businesses through its sale and privatization. Regional budget income from the
municipal housing stock sale is also considered during modeling. A key socio-
economic development indicator generated by this subsystem is public satisfaction
with living conditions. Figure 1 illustrates a fragment of the Housing Stock subsystem
flow chart.
Municipal and private
housing stock
Fig. 1. Fragment of Housing Stock system flow chart
The Health Care subsystem describes the overall dynamics of morbidity rate
(patient population), which depends on such factors as life pattern, environmental
conditions, genetic risk, level of health care, professional risk, quality of medical
services and others. The morbidity rate serves a basis for determining needs in
medical care and demand for medical services. The subsystem describes general
structure of outpatient and inpatient clinics and their staffing. A key indicator
generated by the subsystem is public satisfaction with health care services in a region.
Figure 2 illustrates a fragment of the Health Care system flow chart.
Fig. 2. Fragment of Health Care system flow chart
The Education and Culture subsystem gives an insight into the main
educational institutions in a region by level of education, including primary,
secondary, tertiary education. The subsystem measures the need in education services,
provision of the educational sector with skilled personnel, dynamics of highly
educated population and quality of education in a region. This subsystem generalizes
public satisfaction with education and culture services.
An integrated living standards index in the model is implemented on the basis
of rolled up indicators of public provision with basic social services.
Housing and Utilities Simulation Model
A computer model Housing and Utilities enables describing dynamics of urban
development taking into account various factors such as evolution of the housing
stock and planning of housing and utilities activities, budgetary process and business
activity in a town or city, activities of construction companies, financial relations,
actual demographic and migration processes.
The modeling is aimed at:
= Assessing and forecasting situation in the housing and utilities sector of a
region provided that the current conditions stay unchanged (assessing the
current management strategy defined by a set of controls)
= Analyzing performance of the housing and utilities sector — finding possible
ways of influencing the situation (finding potential controls)
= Comparing and selecting different options of the regional housing and utilities
sector development driven by alternative management decisions
= Planning budget spending for the housing stock to increase housing provision
of population in accordance with social norms at stable regional development
judging by main socio-economic indicators
During the problem area analysis the following factors and processes were
identified and taken into account during modeling:
= Preparation of income and spending budget, local budget spending on the
housing stock and infrastructure (maintenance, overhaul and construction,
other expenditures)
= Evolution of the housing stock and infrastructure (new construction projects,
depreciation, demolition of dilapidated buildings and structures)
= Segregation of elite housing class within the total housing stock of a town or
city
= Housing stock sale to citizens and businesses
= Economic activity of developers and maintenance companies
= Investment processes in the housing stock and infrastructure construction
= Dynamics of household cash income
" Tariff policy in terms of housing and public utility services costs for the
population
« Housing affordability for a family
= Living conditions, including a number of factors the main of which is
depreciation of the housing stock and infrastructure, as well as provision of
resources
= Specific subsidies to socially disadvantaged segments of the population for the
purchase of housing
= Limited free lands for the development
" Differentiation of the housing stock by the form of ownership: municipal and
non-municipal property
In accordance with objectives and tasks specified in the simulated territorial
system were identified the below subsystems.
The Budget subsystem models local budget income and spending with an
emphasis on the housing stock.
The Enterprises subsystem describes commercial legal entities, private
entrepreneurs, municipal and public enterprises and organizations. The model
identifies construction companies that are directly related to the new housing
construction, as well as investment processes in construction. Investments can come
from the government (from budget and off-budget funds), from enterprises and banks,
as well as from the public.
The Housing Stock and Infrastructure subsystem covers all buildings suitable
for permanent residence of citizens. The model distinguishes several housing types
depending on factors of aging and the transition of the municipal housing to private
ownership of citizens and businesses through its sale and privatization.
The Population subsystem. The population is the main consumer of the
housing stock for whom such stock is created. The population can act as housing
construction investors. The population also covers a certain part of public utility
service and rental costs. The model takes into account various benefits, compensation,
rents, subsidies for housing construction and purchasing to socially disadvantaged
segments of the population.
The Housing and Utilities simulation models are designed to address current
issues discussed in the course of the housing and utilities reform implemented at the
federal and regional levels.
Health Care Simulation Model
The Health Care model is aimed at solving a set of tasks associated with the
development and assessment of strategic options for the development of social
infrastructure in a region and health care system management to increase efficiency of
industry resource potential use, improve medical service quality and living standards,
and to improve population health in a region.
The health care process modeling is primarily aimed at conducting
comprehensive analysis and forecasting living standards and health of population in a
region taking into account impact of internal and external factors reflecting actual
medical, demographic and economic situation in a region, assessing and comparing
different health care system management strategies.
The Health Care simulation model is divided into four subsystems.
The Medical and Demographic subsystem describes population in a region,
which dynamics is defined by fertility and mortality rates. Labor resources are
separated from the total population. The model features population morbidity (patient
population), which depends on such factors as life pattern, living standards,
environmental conditions, genetic risk, level of health care, professional risk, quality
of medical services and other. The population morbidity rate serves as a basis for
defining needs in and demand for medical services.
The Medical Service subsystem describes outpatient clinics of a given
capacity, inpatient facilities, hospital beds, material and technical basis, population
medical service costs (labor costs, material and technical base maintenance costs,
operating costs of outpatient and inpatient facilities), private medical and preventive
treatment institutions (characterized by number of facilities, hospital beds and medical
staff), absolute number of patients treated and service rate, medical personnel
(itemized by medical and nursing staff), medical personnel employment.
The Financial subsystem models health care funding taking into account
receipts from the federal budget, receipts from the regional budget, receipts from the
compulsory medical insurance fund, receipts from commercial activities of medical
and preventive treatment institutions and from other sources. Financial flows cover
such expense items as financing of free medical care guaranteed to the population,
operating expenses, medical personnel salaries, capital investments to new medical
and preventive treatment facilities, R&D and new technology introduction expenses,
pharmaceutical provision, cost of preventive actions, retraining and professional
development and other expenses.
The Socio-economic subsystem describes regional economic situation (gross
regional product) taking into account budgetary processes and employment.
In terms of health care industry management a health care simulation model
developed should enable solving the following tasks:
« Analyzing and forecasting living standards in a region
« Analyzing and forecasting population health in a region
= Analyzing, assessing and forecasting provision of population with health care
services
= Analyzing, assessing and forecasting health care sector needs in material,
financial and labor resources
= Strategic planning and restructuring the network of medical care and
preventive treatment facilities
« Financial planning for target areas within the health care sector
= Analyzing decisions made from standpoint of their impact on living standards
in a region and quality and efficiency of the health care system
The model focuses on the overall assessment of the medical and demographic
situation in a region and enables financial planning of the industry.
Pension System Simulation Model
The pension provision is one of the most acute social issues requiring
government solution. A reasoned management decision should be preceded by
statistical analysis of monitored data and “what-if ...” scenario calculations that allow
assessing implications of management decisions made using simulation and
mathematical models.
A pension provision system simulation model developed allows solving the
following tasks:
= Analyzing financial sustainability of the Pension Fund
= Analyzing and forecasting average labor pension size
= Scenario modeling of allocation of pension savings between financial market
segments and forecasting changes in profitability of the total investment
portfolio as the result of changes in its structure
= Analyzing changes in amount of pension savings
= Analyzing financial market impact on investment portfolio profitability and
amount of pension savings
The simulation model is implemented based on system dynamics methods and
includes the following subsystems: insured persons, insurants, pension funds,
management companies, pension legislation, and financial market.
Insured persons: The subsystem models the natural movement of population,
employment, process of choosing a pension scheme by an insured, process of
acquiring pension rights, social characteristics that entitle benefits. Basic social
behavior processes are simulated by the subsystem using the agent-based computer
simulation method with cognitive modeling elements. Expert methods and methods of
multidimensional statistical analysis of pension system data coming from a data
warehouse are used to parameterize a system dynamic model.
Insurants: The subsystem describes dynamics of performance indicators of
employers by industry.
Pension funds: The subsystem generates revenues and plans expenditures of
the pension system.
Pension legislation: The subsystem describes a pension scheme mechanism
depending on wage rates, different pension formulas, pensioner social characteristics.
The subsystem has production model features where a specific pension scheme is
assigned for each insured person depending on his/her characteristics (age, wage,
period of service, health, marital status).
Managing company: The subsystem models activities of management
companies on investing pension savings, describes the formation of the aggregate
investment portfolio consisting of assets in which pension savings are invested,
including deposits, the funds in the accounts of credit institutions, the Russian
government securities, securities of the Russian Federation constituents, municipal
bonds, bonds of Russian business companies, unit investment funds, mortgage
securities, shares of Russian emitters.
Financial market: The subsystem based on multi-agent financial market model
describes dynamics of financial assets in which pension savings are invested.
tot ti
Na
Insurants
) $
\ , WT y [
Insured Persons
Pension Funds
Pension Legislation Management Companies
Fig. 3 - Subsystems of the Pension System simulation model
The model takes into account the differentiated rate of return on each asset and
its fluctuations over time.
The simulation model output statistics is represented by socio-economic
indicators and indicators of the pension system financial sustainability.
Socio-economic indicators:
= Average pension
= Pension adjusted for inflation
= Real wage
= Pension to real wage ratio
Indicators of the pension system financial sustainability:
= Current state of the Pension Fund
"Current position of a management company
= Average yield of an investment portfolio
PROGNOZ experience in modeling solutions
The PROGNOZ company (www.prognoz.com) was founded on base of the
Perm State University during Perestroika by professors of economics and
mathematics. The main clients of young company were government and local
authorities of the Soviet Union and then the Russian Federation. And the basic tasks
the company had to solve were analysis, forecasting, planning of the national and
regional economy and sectors.
There were different applications on the software market but all of them were
created for solving very partial tasks, not so complicated and sophisticated ones.
Large government analytical centers accumulate statistical data on economies
of different countries, regions and industries. These data arrays include information
from many different data-providers and are calculated using different methodologies
and algorithms.
In decision making process specialists implement a wide range of
econometrical methods including smoothing, extrapolation, regression and factor
analysis, multidimensional statistical analysis and so on. And the main requirement of
modelers is simple interface (without coding).
Beside econometrical tools handling a single equations it is required to make
systems of equations, dynamic system of equations (Vector autoregression and Error-
correction model), to sort equations and to analyze cycles in them. Designing of large-
scale models also requires users to specialize in determined sectors of economy and
corresponding blocks of model.
Besides simple scenarios of forecasting and simulations, practical tasks need
to have instruments for optimization and dynamic programming.
Summarizing all the above mentioned, we can conclude that application
oriented to such tasks solving should:
1. be integrated with industrial databases and include tools for data validation;
2. include many statistical and mathematical methods with user-friendly
interface;
. include tools for handling large-scale mathematical models;
. have tools for forecasting and optimization;
. include multiuser working with modeling and administration tools;
. advanced decision making tools;
. include results of visualization tools.
Large-scale model can include some hundreds and even thousands of
equations and modeling variable. This feature allows users to take into account not
only direct linkages between variables but also indirect ones and feedbacks.
Sensitivity analysis allows users to estimate influence of different shocks on key
macroeconomic variables and elasticities in dynamics.
NOP w
‘Consumer Price Index
Exchange Rate RMB/100 USD alia ili
Money Supply M2, 100 mn Yuan
100 mn Yuan
[eraware taonny [eat @ wan
Fig. 4. Sensitivity analysis examples
Scenarios approach in these models allows experts to determine threshold
values of exogenous variables. Such kind of analysis is used to determine limit of oil
price that can be offered by oil exporter countries.
Budget deficit Trade balance
200,00
150.00
400,00 | | |
50.00
oo om oe |
0 0 3 4 50 60) MD ee ee a |)
GDP index
1.00
0.00
4,00
2.00
300
ce a a a a
Fig. 5. Determination of oil price threshold value
Conclusion
The elaborated Social Sphere model complex allows forecasting regional
socio-economic development and living standards, performing comprehensive
analysis of altemative management decision on future regional socio-economic
development by selecting financial, economic and other controls to ensure growth of
living standards and sustainable development of a region as a whole. The model
complex enables solving resource management and financial planning tasks in the
overall social sphere and its industries, including housing, health care and others, as
well as developing a social policy taking into account actual demographic trends,
environmental, medical and demographic situation; financial, labor and other
resources in a region. The use of the model complex in practice by regional and local
govemments will enable solving specified tasks using a computer model taking into
account short-term and long-term outlook.
The simulation models are implemented in high-tech simulation environment
powered by AnyLogic software based on regulatory schemes of system dynamics
models with advanced ideographic capabilities for building and visualizing system
flow charts, tools for scenario calculations based on simulation models.
References
1. N. N. Lychkina, D.N. Shults Simulation Modeling of Regions’ Social
and Economic Development in Decision Support Systems - 27" ‘International
Conference of the System Dynamics Society, Albuquerque, New Mexico, USA, 2009.
2. N. N. Lychkina, Y.A. Morozova, D.N. Shults Stratification of Socio-
Economic Systems Based on the Principles of the Multi-Modeling in a Heterogeneous
Information-Analytical Environment // 2" Intemational Multi-Conference on
Complexity, Informatics and Cybemetics, Orlando, Florida, USA: Intemational
Institute of Informatics and Cybemetics, March 27-30, 2011.
