Content uploaded by Martin Kaggwa
Author content
All content in this area was uploaded by Martin Kaggwa on Nov 19, 2015
Content may be subject to copyright.
GDN_Role of System Dynamics in Policy Alignment1 Page 1 of 24
The role of system dynamics in aligning policy goal and actions: Reflections on South
Africa’s automotive industry growth (1995-2004)
*
Martin Kaggwa
Abstract
South Africa has been has been giving investment and export-based incentives to the
automotive industry as a means of supporting the industry attain global competitiveness.
Whereas industry investment, production and exports have increased in the period 1995 to
2004, employment and sourcing of domestic components have lagged behind general industry
growth. Industry trade balance has deteriorated over the same period; moreover, investment
in Research and Development (R&D) activities, as an indicator for future competitiveness
remains insignificant.
The paper argues that the offer of investment and export-based incentives to the automotive
industry is a necessary but not a sufficient condition for making the industry competitive in
the long-term. The industry support model does not take into account internal
interrelationships, and interdependencies among industry sectors critical to understanding
industry behaviour. Processes underlying industry performance are also omitted in the model.
Using a system dynamics qualitative model, the paper shows how a system dynamics
approach provides means of incorporating effects of systemic factors and processes, active
within the industry, in understanding how government incentives are influencing industry
performance. The system dynamics approach provides some understanding onto potential
sources of the uneven industry performance. In bringing to the fore key processes underlying
industry performance, the approach opens up opportunity for policy intervention to target
processes that are critical in determining industry outcomes. System dynamics provides useful
means of aligning the long-term industry objective of competitiveness with current policy
intervention.
*
Martin Kaggwa is a PhD Candidate with the Institute of Technological Innovation, University of Pretoria –
South Africa. Email: mkaggwa@aidc.co.za
GDN_Role of System Dynamics in Policy Alignment1 Page 2 of 24
1. Introduction
Automotive production in South Africa started in the 1920s. Government used tariff
regulation and local content requirements to guide industry growth (Black, 2001). By early
1990s, it was evident that the hitherto adopted inward looking policy stance was not
sustainable in the long run. The industry had to comply with General Agreement on Tariffs
and Trade (GATT) and World Trade Organisation (WTO) trade regulations (Damoense &
Simon, 2004). Domestic market growth constraint meant that exports had to play a big role in
industry growth. Government realised that the industry needed encouragement with a number
of “sticks and carrots” to change and improve its competitiveness (Motor Business
International, 2000). Of major importance to government were ways through which to
maintain and grow the industry in a less protected trade environment. Table 1 summarises
development of automotive policy in South Africa.
In 1995, government adopted an outward looking policy framework for the automotive
industry – the Motor Industry Development Programme (MIDP). The MIDP had specific
objectives to achieve – international and domestic competitiveness, vehicle affordability,
employment, local supplier development and improvement of industry trade balance. The
programme also provided a framework for government support of the automotive industry
through offer of incentives. Formulation of the MIDP was a consultative process and all
stakeholders agreed that the MIDP provided required impetus to grow the industry in a
liberalised trade environment.
Up to the early 2000, stakeholders concurred that the MIDP was the most successful industry
support in the country. There were suggestions to replicate the same model of support in other
industries. Industry was adamant to programme critics that claimed that the programme had
an implicit cost to domestic consumers (Flatters, 2002).
Doubts on the programme’s long-term success began to emerge after mid-term review of
2002. Job creation and local component sourcing increase did not commensurate with overall
industry growth. Industry trade balance was not improving despite high increase in
automotive exports from the country. The questioning of the programme’s compliance with
the WTO regulation by Australia in 2005 further exacerbated uncertainty on the programme
as long-term strategy to support industry competitiveness. The uneven success of the
GDN_Role of System Dynamics in Policy Alignment1 Page 3 of 24
programme has weakened the initial consensus on the programme yet, without the MIDP or
comparable alternative support, the automotive industry in South Africa can collapse.
Despite being a consultative process, the formulation of the MIDP put less emphasis on how
incentives were to lead to industry competitiveness in the long term. As a result, processes,
systemic factors and feedbacks effects within the industry have received little attention in the
management and implementation of the programme yet these factors are critical in
understanding industry performance. The future of the MIDP requires a re-examination of the
initial thinking of the programme and search for potential causes of the unexpected industry
outcomes.
2. Data sources and methodology
The intention of the study was to come up with a formal MIDP model that was cognisant of
interrelationships between and feedback effects among industry variables and reflects on how
systemic and process factors enrich the MIDP policy framework. The study used both
quantitative and qualitative data collection techniques to support industry performance
analysis and MIDP model building process.
Data Sources
Quantitative historical data was collected from two sources:
1. The Department of Trade and Industry South Africa (thedti): Thedti carries out annual
surveys to capture industry performance data as part of its monitoring mandate.
Although part of this data is confidential, data relating to general trends in industry
performance is published in the department’s annual publication “Current
developments in automotive industry” and is available in the public domain. Thedti
data is triangulated with other internal but confidential data sources, thereby
increasing its reliability. One issue of concern on thedti data is its representativeness
of the industry. Response to the industry survey questionnaire is voluntary; therefore,
the sample size is self-selected. Depending on the number of companies that respond
to the survey, thedti data may not be representative of the industry. A review of the
2006 annual survey, however, revealed that the sample size, for the vehicle
manufacturing, from which data was captured, was representative of the industry.
Questionnaires were sent to all of the 8 local vehicle manufacturers in the country; all
GDN_Role of System Dynamics in Policy Alignment1 Page 4 of 24
of them responded. For the component sector that comprises of some 278 first-tier
suppliers, 40 questionnaires were sent and half of the companies responded.
2. The National Association of Automotive Manufacturers of South Africa (NAAMSA):
NAAMSA is the national association of all domestically based light, medium and
heavy commercial vehicle manufacturers. NAAMSA is also the representative
organisation for franchise holders marketing vehicles in South Africa. NAAMSA
membership stood at 25 companies at the beginning of 2006. The association collects
performance data from all its members. The data is published in the organisation’s
annual reports and is periodically disseminated to the public through press briefings.
The NAAMSA data is more comprehensive and disaggregated but can potentially be
biased due to vested interests. NAAMSA data was compared with thedti data and in
cases of significant deviation between the two data sets, thedti data was preferred.
Thedti and NAAMSA data was supplemented by qualitative data collected from archive
documents. Board of Tariffs and Trade reports and government gazettes on the rationale for
the introduction of the MIDP were specifically reviewed. Furthermore, for a period of over a
year and half, the researcher interacted with the study situation under the Motor Industry
Development Council (MIDC) - a forum that brings together industry stakeholders to
deliberate on the MIDP. Attendance of the MIDC was critical in understanding the working of
and stakeholders’ thinking on the MIDP. Qualitative Causal Loop Diagrams (CLDs) depicting
industry interrelationship were developed using qualitative data from the MIDC mainly. The
CLDs were improved upon as more data was received until some level of consistence and
stability were established. Whenever required, clarity was sought from industry
representatives on positions taken and deliberations at the MIDC.
System dynamics
System dynamics (SD) is a methodology for building qualitative and quantitative models of
complex situations so that they can be better understood and managed (Caulfield and Maj,
2001). The method is based on the premises that the internal structure of any system
determines the system's behavior. At the core of the approach is the capturing feedbacks and
non-linear relationships within a system. One of the advantages of the approach is that it
allows progression from qualitative models to simulation models. SD simulation models can
accept input of both quantitative and qualitative data; they incorporate feedbacks and other
GDN_Role of System Dynamics in Policy Alignment1 Page 5 of 24
dynamics in the analysis, at the same time allowing carrying out sensitivity analysis on the
model (Sterman, 2000).
From a SD perspective, the automotive industry in South Africa can be seen a system, that is,
a group of independent but interrelated elements comprising a unified whole.
3. MIDP Mental Model
A mental model is an enduring and accessible, but limited, internal conceptual representation
of an external system whose structure maintains the perceived structure of that system (Doyle
and Ford, 1998); a filter through which one interprets experiences, evaluates plans, and
chooses among possible courses of action (Sterman, 1991). A mental model contains ideas,
opinion, assumptions, generalisations, etc. with respect to a policy problem and related issues
(Vennix, 1990). It describes facts and concepts that constitute ones understanding of a
particular phenomenon (Morecroft, 1998). Mental models are the starting point of formal
system dynamics modeling.
The MIDP policy framework is guided by a uni-directional, static and non-interactive mental
model. The model presupposes that providing the automotive industry with Import Rebate
Credit Certificates (IRCCs), duty rebates and allowances to import inputs free of duty can
positively influence industry competitiveness. Companies use IRCCs and duty rebates to
offset import duty payable on vehicles and automotive components in excess of the duty free
allowance. Figure 1 provides an interpretation of the thinking behind the MIDP mental
model.
Vehicle and component manufacturers earn IRCCS by exporting automotive products under
the Import-Export Complementation (IEC) dispensation and earn import duty rebates by
investing in productive assets under the Productive Asset Allowance (PAA) dispensation. The
MIDP also specifies conditions for import of industry inputs free of duty.
The import-export complementation arrangement allows duty reduction on cars and light
commercial vehicles imported based on the value of local content exported. For every
Completed Built Unit (CBU) exported, a percentage determined by the Value of Exports
Performance (VEP) of CBUs can be imported free of duty. The Value for Export Performance
started at 94% in 2003, reducing by 4% per annum.
GDN_Role of System Dynamics in Policy Alignment1 Page 6 of 24
The Productive Asset Allowance (PAA) provides vehicle manufacturers in the Southern
Africa Custom Union (SACU) rebates equal to 20% of the value of investment in new
productive assets. Benefit from the PAA is spread over a five-year period.
The duty free allowance (DFA) is a proportion of the wholesale value of the vehicle that may
be imported free of duty.
4. Automotive industry performance under the MIDP
Production
In the period, 1995 to 2004, for which the MIDP has been operational, industry performance
has recorded success on some but not all industry deliverables. The number of vehicles
produced remained below pre-MIDP volumes for a while. On average, production volumes
decreased by 1.4 percent per year between 1995 and 2000. Production volumes for 2000 were
8.4% percent lower than at the inception of the MIDP. A gradual increase in production
volumes has been recorded since the year 2000 (Figure 2).
Investment
On the other hand, total industry investment increased from R847 million in 1995 to R2.2
billion in 2004 (Figure 3). Industry investment peaked in 2002 after which there has been a
gradual but not significant decrease. One reason behind the increase in industry investment
has been the introduction of new models for which major investment was required. The 8
vehicle manufacturers in the country have each introduced at least one new model within the
period.
Exports
The value of automotive exports from South Africa increased from R4.2 billion in 1995 to
R39.2 billion in 2004, an increase of more than 8 fold (Figure 4). Increase in exports is seen
as evidence of successful integration of the previously protected South African automotive
industry in the global business.
Trade Balance
Despite increase in industry investment and exports, industry trade deficit has worsened from
R12.2 billion to R18.8 billion in the period 1994 to 2004 (Figure 6). The increase in
GDN_Role of System Dynamics in Policy Alignment1 Page 7 of 24
automotive exports has been accompanied by high imports of automotive products into the
domestic market. There are concerns that if the trend in automotive imports is allowed to
continue unabated, it will negatively affect domestic vehicle production and the local sourcing
of components in the long term. Even after accounting for the country’s exchange rates trend
effects on industry performance, the impact of the MIDP on the industry trade seems
negative. The deteriorating industry trade balance a significant unintended consequence of the
MIDP.
Employment
In stating MIDP objectives, government, industry and labour agreed on a compromise to tone
down on the employment objective. Considering that attainment of globally competitive
production efficiency in the industry often has a negative impact on employment, it was stated
that the MIDP was to stabilise rather than create employment.
According to the government-commissioned study on automotive industry employment of
2005, combined vehicle manufacturing and component sector employment increased by 9%
between 1995 and 2004. The increase in employment, emanated from the component
manufacturing sector only. Employment in the vehicle manufacturing sector has been on
decline since the commencement of the MIDP. The head count of 34,152 recorded for the
vehicle manufacturing sector in 2004 was 17% lower than the pre-MIDP employment of
41,364 head count (Figure 6).
Supplier Development and Sourcing of local component
The MIDP was supposed to facilitate integration of domestic component manufacturers into
the global automotive value chain. The pre-MIDP period was characterised by low production
levels that could not adequately support a vibrant component sector. The import-export
complementation scheme was intended, in part; to assist component suppliers generate
volumes that would make them more efficient. The component sector often has a bigger
potential to create jobs and to stimulate domestic technological capabilities through spillovers
effects (Humphrey and Memedovic, 2003).
The extent to which supplier development has taken place under the MIDP is an elusive
aspect. Neither government nor industry keeps explicit data on the issue. The use of proxies
GDN_Role of System Dynamics in Policy Alignment1 Page 8 of 24
is the only viable way to assess local supplier development. This paper makes use of local
content use and domestic component sourcing as proxies for supplier development.
The share of locally sourced components used in domestic vehicle assembly was on decline
from 1992 to 1994. It remained low but stable between 1994 and 1995 (Bell and Madula,
2003). There was substantial reduction in the share of locally sourced components as a
proportion of total component usage from 40.1% in 1996 to 33.8% in 2000 (Table 2).
In essence, local component manufacturers were benefiting less from each domestically built
vehicle. If the proportions of local components per each manufactured vehicle were to
continue on the same declining trend of 1996, it would mean that the MIDP will become less
and less effective in supporting local component manufacturers despite industry growth.
Post 2000 data is not available but given the trade deficit that has characterised the industry,
the expectation is that participation of local component sector in the industry value added
activities has not improved much. A rapidly emerging new group of mainly foreign-owned
firms (Black, 2001; Miozzo, 2000) has been responsible for the bulk of expansion in
automotive component export realised under the MIDP.
5. Industry performance and long-term competitiveness
In assessing whether industry is on the right path to graduation to global competitiveness, one
has look beyond short-term industry performance parameters. In this respect, the nature of
investment provides a useful starting point.
The nature of investment undertaken has a bearing on the process towards achieving
competitiveness by an industry. According to Waddock & Graves (1994), R&D investment
as opposed to capital investment is associated with improved industry competitiveness.
Investment in plant, machinery and tooling is important in the realisation of short to medium
term profitability of firms, but in the long run it is the R&D investment and the subsequent
potential to innovate that is likely to determine industry competitiveness (Fan, 2006; Taymaz
et al, 2004; Koschatzky et al, 2001; Lee, 2000). R&D investment intensity can also be
indicative of the willingness of firms to commit themselves to new products and improved
processes within a particular location (Graves, and Waddock, 1994). By deciding to undertake
GDN_Role of System Dynamics in Policy Alignment1 Page 9 of 24
R&D and innovation activities, enterprises signal the importance they attach to a location in
terms of future competitive strategy.
Investment in R&D is one of the main determinants of innovative capacity. For a domestic
automotive industry to continue supplying automotive products competitively, it has to keep
pace with the ever-improving technological specifications of global automotive vehicle
manufacturers. Innovation is an important element in achieving both production processes and
resultant products that meet global standards (Koschatzky et al, 2001). To offer an industry
investment incentives and increase in investment does not guarantee support of the industry’s
competitiveness objective. The offer of non-targeted investment incentives can even
potentially lead to enterprises switching to less costly technological investment that yields
quicker returns on investment in the short run at the cost of long-term competitiveness (Zhu et
al, 2006). The nature of industry investment that has taken place under the MIDP dispensation
can, therefore, provide insights on the extent to which the programme is supporting the
process towards competitiveness.
For the period 1998 to 2004, investment in plant, machinery and tooling constituted more than
80 percent of the total annual vehicle manufacturers’ investment. Investment in support
infrastructure that included R&D was less than 10 percent of total expenditure (Table 3). Land
and buildings accounted for the rest of the investment. Moreover, the bulk of the investment
categorised under investment support infrastructure related to technical fees paid to foreign
experts that provided specialised services at the launch of new models in the country.
In spite of the general increase in industry investment, investment in R&D activities has
remained minimal. The low level of R&D in the automotive industry is in line with the
findings of the South African Innovation Survey of 2001, which showed that 51% of firms in
the country were not engaged in R&D in terms of persons working on R&D activities (Table
4). On average firms in South Africa allocated less than 2 percent of their annual turnover to
R&D innovation activities (Oerlemans et al, 2003). At a national level, the percentage of
gross domestic expenditure on R&D was below 1% of the country’s gross domestic product;
lower than most developed countries (Department of Science and Technology, 2005).
GDN_Role of System Dynamics in Policy Alignment1 Page 10 of 24
Considering the type of investments that have benefited from MIDP incentive thus far, as well
as the national effort towards R&D, the potential of the programme to support the industry’s
progress towards sustainable global competitiveness appears to be weak.
In terms of actual industry competitiveness, industry performance indicators show mixed
results. According to the European Competitiveness Report (2004), competitiveness can be
defined as the ability of an industry to defend, and/or gain market share in open markets
relying on price and/or quality of its goods. Common indicators for assessing industry
competitiveness include the growth rate or increase in domestic market share of locally
produced vehicles and export growth rates (Narayanan, 1998).
The weak support of the competitiveness process by the MIDP through R&D is compounded
by the seemingly diminished industry competitiveness. Domestic market share of locally
produced vehicles decreased from 93.2% in 1995 to 71.6% by 2004 (Table 5). According to
Auto Insight (2006), the sale of locally produced vehicles increased by 19.6%, while the sale
of imported cars increased by 155% from 2004 to 2005.
On the other hand, industry realised reasonably high growth rates in vehicle exports between
1997 and 2001 (Table 5), indicating that more automotive products from South Africa were
being put on the global market. In the context of the automotive industry in South Africa,
however, export growth rates can be a weak proxy for international competitiveness. The
offer of export-based import rebate credit certificates, as an incentive under the MIDP,
cushions domestic vehicle manufacturers from competitive pressure. Although increase in
exports is a desirable effect of the MIDP, one has to take into account government support
received on exports before a qualified statement on industry competitiveness based on an
increase in exports can be made.
6. MIDP and Policy resistance
Policy resistance is central issue of concern in systems dynamics methodology. According to
Meadows (1982) policy resistance occurs when policy intervention leads to delayed, dilute, or
defeat the intended purpose. Sterman (2000) refers to policy resistance as a tendency for
intervention to be defeated by the response of the system to the intervention itself. Policy
resistance often leads to the opposite of the intended results (Forrester, 1969). System
dynamics singles out policy resistance as the main reason behind ineffective policy
GDN_Role of System Dynamics in Policy Alignment1 Page 11 of 24
intervention. Forrester (1991) contend that as high as 98% of policies in a system have little
effect on the behaviour because of the ability of the system to compensate for changes in most
policies.
Whereas investment, production and exports have increased in the period 1995 to 2004,
employment and sourcing of domestic components have lagged behind the general industry
growth. Industry trade balance has deteriorated over the period; moreover, investment in
Research and Development (R&D) activities, as an indicator for future competitiveness
remains insignificant. It is also notable that production response to investment rate has been
very low. The performance of the automotive industry under the MIDP reveals both
unintended and diluted outcomes of policy intervention characteristic of a policy resistance
phenomenon.
The uneven industry performance coupled with lack of a clear sense whether industry is
becoming globally competitive poses a big challenge on how to take forward government-
industry support programmes.
7. MIDP as Systems Problem
The MIDP seems like a simple concept, but its ramification on industry dynamics are vast.
The working of the MIDP shows interrelationships between sectors and industry variables
without explicit cause and effects, characteristic of a complex system.
The MIDP mental model reveals two conspicuous shortcomings from a systems dynamics
perspective:
It does not capture feedback effects between the model variables. The model assumes,
for example, a positive relationship between the value of IRCCs earned and industry
production levels, without taking into account that production levels may in turn affect
the value of IRCCs through the export variable. Increase in productions has a positive
effect on export levels through low average cost realisation, and subsequently on the
value of export-based IRCCs.
Second, the model assumes that MIDP objectives as captured by the expected
outcomes matrix have no effect on each other. Possible trade-offs and
complementation between programme outcomes is not acknowledged.
GDN_Role of System Dynamics in Policy Alignment1 Page 12 of 24
The thinking behind the MIDP exposes gaps in capturing systemic relationships, processes
and feedback effects active within the industry. Ignoring such feedbacks and variable
interrelations in the model leads to inaccurate and incomplete perceptions underlying a
particular policy and leads to ineffective policy intervention (Sterman, 2000). In order to get a
better insight into effect of government support to the motor industry, these factors have to be
incorporated in the MIDP mental model.
Under the MIDP, government uses two major policy tools to influence industry performance
– the stock of IRCCs and the level of duty-free import allowable. After specifying rules
governing the policy tools, what transpires within the industry and subsequent performance
are largely dependent on triggered industry dynamics, which in turn depend on industry
structure. A qualitative system dynamics model provides a useful tool of articulating internal
dynamics active within the industry.
The following section presents a recast of the MIDP’s Productive Asset Allowance and the
import-export complementation arrangement into a qualitative system dynamics model. High-
level casual loop diagrams for the two incentives are included that capture internal factors at
play, some of which are omitted through presenting the MIDP as a uni-directional, static and
non-interactive model.
The Productive Asset Allowance (PAA) Causal Loop Diagram
The PAA incentive presupposes that potential industry investors will increase investment
based on the level of investment incentives obtainable. Since the value of PAA is specified as
a constant proportion of the level of investment, a feedback loop is created between
investment and investment incentives obtainable.
The level of investment incentives receivable under the PAA increases with aggregate
industry investment. Holding other factors constant, an increase in investment will increase
investment incentives obtainable. The increased value of investment incentives received will
motivate further investment, which will in turn increase the value of investment incentives
receivable. A reinforcing investment process is created. In reality, perpetual increase in
investment is not feasible. At some point, the level of desired-investment constraint comes
into effect, to stop the continuous increase in industry investment. A casual loop diagram
capturing the systemic and feedback effect between investment and investment incentives is
GDN_Role of System Dynamics in Policy Alignment1 Page 13 of 24
presented in figure 7. Three ‘desired-investment’ determining factors, relevant to the
automotive industry - potential market growth, vehicle-model change interval and
depreciation are included.
It is further noted that for each level of production, there is an optimal level of investment.
Based on the famous Cobb Douglas production function that specifies output as a function of
capital and labour, investors in the automotive industry can crudely estimate required
investment to produce particular outputs levels. Planned output will then determine the
desired level of investment.
Although the investment-investment incentive loop is supposed to be reinforcing, this cannot
practically happen, due limitations arising from planned production levels. One has to look at
factors that determine planned production and subsequent desired-investment in South
Africa’s automotive industry in order to understand the effectiveness of the investment
incentives in stimulating industry investment.
Import-export complementation Casual Loop Diagram
The functioning of the Import-export complementation is a major source of industry dynamics
and feedback effects. Under the arrangement, companies exporting automotive products
receive IRCCs based on the local content value exported. The arrangement is based on the
understanding that increase in exports will drive domestic industry growth. For domestic
companies to be able to export, they must be able to produce world-class auto products at
globally competitive prices. Indirectly the incentive motivates the domestic industry to attain
economies of scale and efficient production means. The later element is related to the
investment in productive assets and R&D efforts within the industry.
In terms of systemic relationships, increase in exports will increase the value of IRCCs
receivable by the industry. Since companies can only benefit from the IRCCs received
through offsetting duties payable on imports, the value of available IRCCs increase industry
propensity to import. The level of duty free allowance further augments the propensity to
import since IRCCs are used to pay import duty net of the duty free allowance.
By default, therefore, the import-export complementation arrangement has a delayed feedback
effect on domestic production through the import propensity effect. To the extent that local
GDN_Role of System Dynamics in Policy Alignment1 Page 14 of 24
market growth does not keep pace with imported automotive products, the domestic market
share for locally produced products will decrease. Unless the increase in export is significant
enough to offset reduction in domestic market share of locally produced automotive products,
domestic production will fall in the long run. Decrease in domestic production will decrease
the value of IRCCs relievable via the export variable, mitigation against the perpetual
decrease in local production. The import-export complementation loop is therefore a counter
balancing.
In acknowledgment of the process towards long-term competitiveness, the delayed but
positive effect of domestic R&D is included in the analysis. Figure 8 presents the functioning
of import-export complementation in a high-level casual loop diagram.
Of much interest recently, is the impact of the import-export complementation arrangement
on industry trade balance. Exports improve industry trade balance, yet on the other hand,
exports increase propensity to imports via export-based IRCCs. The resultant effect is another
counter-balancing, but incomplete trade balance loop X in figure 8.
Although Figure 7 and 8 are separately presented, they are part of a single industry system.
Due to complexity of cause and effects involved between different variables, the human mind
may not fully comprehend and anticipate outcomes of the combined interrelationships
presented. Indications of likely outcomes are only possible probabilistically, through
quantification of the model and simulation of scenarios. Articulation of the MIDP model
from a system dynamics perspective provides a necessary foundation for quantification and
simulation of the effects of incentives to the automotive industry.
8. Insights and Conclusion
A qualitative system dynamics model of the MIDP incentives brings new insights, regarding
industry performance under the programme that have thus far received little attention:
The approach shifts focus from outcomes, to factors underlying realised industry
outcomes. The role of internal processes and interrelationships within industry
including R&D come to fore. The approach opens up opportunity to target and
incentivise processes that are seen as critical in determining industry outcomes.
System dynamics approach extends the thinking around the offer and effectiveness of
investment incentives to include factors that affect planned production levels and the
GDN_Role of System Dynamics in Policy Alignment1 Page 15 of 24
desired level of industry investment. Investment incentives may have limited effect on
industry investment unless if supported by factors that cause an upward shift in the
level of desired investment. In this respect, economic performance and vehicle
manufacturers’ company strategies become important.
The approach sheds light on the uneven performance of the MIDP, sometimes
bordering to policy resistance. Signs of policy resistance under the MIDP policy
regimes further strengthening the case to system dynamics approach in the MIDP
policy articulation, implementation and management.
Understanding systemic and structural issues underlying dynamics of a system like the MIDP
incentive model is useful in highlighting relationships, processes, feedbacks and industry
dynamics over time. It puts policy makers in a better position to plan and evaluate policy
action based on possible effects to the industry as a whole. Specifically for South Africa’s
automotive industry, the approach provides useful means of aligning the long-term objective
of competitiveness with policy interventions to ensure that benefit from industry growth
outlives the period of government offer incentives.
GDN_Role of System Dynamics in Policy Alignment1 Page 16 of 24
Table 1: Development of Automotive Policy in South Africa
Policy Measure Period
1. High tariffs 1920 to 1995
2. Local content requirements by mass 1961 to 1987
3. Local content requirements by Value 1989 to 1995
4. Import-export Complementation (MIDP) 1995 to date
5. Investment Incentives (MIDP) 2000 to date
Source: Damoense (2004)
Table 2: Imported and locally sourced components as a proportion of total component
usage, and as a proportion of wholesale vehicle turnover: 1996 – 2000 (%)
Year
Imported
components/Total
components
Local
components/Total
components
Imported
components/WVT
a
Local
components/WVT
Total Local
Content/WVT
1996 59.9 40.1 41.9 28.1 58.1
1997 61.2 38.8 42.8 27.2 57.2
1998 58.3 41.7 40.8 29.3 59.2
1999 60.0 40.0 42.0 28.0 58.0
2000 66.2 33.8 46.3 23.7 53.7
Notes: WVT stands for wholesale vehicle sale turn over;
Source: Derived from data from Trade and Investment South Africa (TISA) presented in Bell and Madula (2003)
Table 3: Automotive industry investment expenditure – South Africa (1998-2004)
Year Investment (Rm)
1
Investment in support
infrastructure (incl. R&D) as a %
of total OEM investment
Investment in plant, machinery
and tooling -as a % of total OEM
investment
1995 847 9.2 86.6
1996 1,171 11.1 85
1997 1,265 8.8 81
1998 1,342 10.4 85.2
1999 1,511 7.6 87
2000 1,562 9 83.9
2001 2,078 11.8 86.6
2002 2,726 9.6 84.8
2003 2,325 8.3 85.5
2004
3,577 10.1 86.9
Source: Current Developments in Automotive Industry 2003/2004 – thedti South Africa and NAAMSA Annual
Report 2001/2006
GDN_Role of System Dynamics in Policy Alignment1 Page 17 of 24
Table 4: R&D intensity classes 2000
R&D Intensity
*
Percentage of firms Cumulative percentage
0% 51.2 51.2
0.01 to 1.05% 14.9 66.1
1.50 to 3.00% 16.9 83
3.00 to 4.50% 8.9 91.9
4.50 to 6.00% 1.4 93.3
6.00% or more 6.7 100
*
R&D intensity refers to the percentage of workers in the total workforce of an organisation performing
R&D activities
Source: South African Innovation Survey 2001
Table 5: Domestic market share of locally produced vehicles and vehicle
export growth rate of the South Africa automotive industry
Year
Domestic market share of
locally produced vehicles
(%)
Vehicle export –
Annual growth rate
1995 93.2 -
1996 88.7 -26.7
1997 85.6 69.4
1998 81.1 32.3
1999 81.5 130.6
2000 81.3 13.9
2001 77.9 59.2
2002 75.8 15.7
2003 77 1.1
2004 71.6 -12.8
Source: Calculations based on NAAMSA Data – Annual Report 2005
GDN_Role of System Dynamics in Policy Alignment1 Page 18 of 24
Figure 1: MIDP Incentives Mental Model
388,442
385,252
361,316
310,333
325,222
356,250
406,149
404,441
421,335
455,052
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000
450,000
500,000
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Units
Figure 2: South Africa’s vehicle production (Units)
Source: Naamsa Annual Report 2005
1. Export based
IRCCs
2. Investment
based duty
rebates
3. Duty free
allowance on
industr
y
in
p
uts
Locally Based:
Vehicle
manufacturers
Component
manufacturers
a. International
competitiveness
b. Positive trade
balance
c. Sustain employment
d. Vehicle quality and
affordability
e. Increased domestic
sourcing of
components
Incentive
Recipients
Expected outcomes
GDN_Role of System Dynamics in Policy Alignment1 Page 19 of 24
660
697
858
400
492
847
1,171
1,265
1,342
1,511
1,562
2,078
2,726
2,325
2,220
0
500
1,000
1,500
2,000
2,500
3,000
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Investment (R million)
Figure 3: South Africa’s automotive industry investment
Source: Naamsa Annual Report 2005
0.8
1.1
1.5
2.3
2.8
4.2
5.1
6.6
10.1
14.8
20.0
30.0
40.1
40.7
39.2
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Export value (Rbn)
Figure 4: Value of South Africa’s automotive exports
Source: Naamsa Annual Report 2005
GDN_Role of System Dynamics in Policy Alignment1 Page 20 of 24
5.9
6.9
5.5
5.2
5.1
6.8
9.2
12.2
14.1
10.6
9.8
8
9.7
8
10.1
9.1
18.8
0
2
4
6
8
10
12
14
16
18
20
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Year
Trade Deficiet (R billion)
Figure 5: Automotive Trade Deficit South Africa
Source: Naamsa Annual Report 2005
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
Year
Headcount
Vehicle Assembly
60,800 64,144 64,337 64,401 64,659 67,568 69,730 72,031 73,040 76,911
Component Sector
41,364 40,346 39,341 35,774 34,593 35,543 35,128 33,789 33,637 34,152
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Figure 6: Employment in the South African Automotive Industry, 1995 to 2004
Source: FRIDGE Study on Employment on SA Automotive Employment 2005
GDN_Role of System Dynamics in Policy Alignment1 Page 21 of 24
Desired investment
Potential market
growth
Depreciation
Model change
interval
Investment
Investment
incentives
+
+
+
+
+
+
Investment multiplier Loop
Figure 7: MIDP investment-investment incentive casual loop diagram
Exports
Import rebate credit
certificates
Imports
Domestic market
+
+
-
Import-export complementation
X
Production
Car prices
+
-
-
Trade balanc
e
+
-
Duty free
allowance
+
Domestic R&D
+
Figure 8: Import-export complementation casual loop diagram
Notes on the causal loop diagrams:
The arrows denote a cause-effect relationship between the connected variables
“+” at the head of the arrow denotes that the connected variables change in the same
direction, i.e., when one increases, the other variable will also increase and vice versa.
“-” at the head of the arrow denotes that the connected variables change in opposite
directions. When one variable increase, the other will decrease and vice versa.
denotes an overall reinforcing effect
denotes an overall tendency towards an equilibrium state
GDN_Role of System Dynamics in Policy Alignment1 Page 22 of 24
References
Barnes, J. 2000. Domestic Market Pressure Facing the South African Automotive
Component Industry. Research Report No.33: Industrial Restructuring Project, School
of Development Studies, Natal University.
Bell, T. 1997. ‘Trade Policy’, in Michie, J & Padayachee, V, (eds). The political economy of
South Africa, London: Harcourt.
Bell, T, & Madula, N. 2003. South African motor industry policy in a cloud of uncertainty.
National Institute of Economic Policy paper commissioned by National Union of
Metal Workers of South Africa, Johannesburg.
Black, A. 2001. Globalization and restructuring in the South African automotive industry,
Journal of International Development, 13, pp 779-796.
Caulfield, C.W, & Maj, S.P. 2002. A Case of System Dynamics. Global Journal of
Engineering Education, 16(1): 23-33.
Damoense, M.Y., & Simon, A. 2004. An analysis the impact of the first phase of South
Africa’s Motor Industry Development Programme (MIDP), 1995-2000, Development
Southern Africa, 21(2), pp 251-269.
Department of Science and Technology South Africa. 2005. National survey of research
and experimental development (R&D) 2003/4 fiscal year.
Http://
www.naci.org.za/pdfs/sa_nat rdsurvey_ 2003_4.pdf. Accessed on 5 May 2006
Doyle, J.K., and Ford, N.F. 1998. Mental model concepts for system dynamics research,
Systems Dynamics Review, 14(1), pp 3–29.
European Commission. 2004. European Competitiveness Report. http://europa.eu.int/
comm/enterprise/automotive/pagesbackground/competitiveness/comprep_2004_en_
automotive. PDF. Accessed on the 10 May 2006.
Fan, P. 2006. Catching up through developing innovation capability: evidence form China
Telecom-equipment industry, Technovation, 26, pp 359-368.
Flatters, F. 2002. From import substitution to export promotion: Driving the South African
motor industry. Paper for the on-going USAID-funded project with the Service Group,
SADC Secretariat on Trade and Investment Policies.
Forrester, J.W. 1969. Urban Dynamics, Portland: Productivity Press.
Forrester, J. 1991. Usefulness of system dynamics approach in case of policy resistance. In
(Ed): Systems basis of policy making in the 1990s, Greene, K.B.D. 1991
Humphrey, J., & Memedovic, O. 2003. The global automotive industry value chain: What
prospects for upgrading by developing countries? United Nations Industrial
Development Organisation (UNIDO). Geneva.
GDN_Role of System Dynamics in Policy Alignment1 Page 23 of 24
Koschatzky, K., Bross, U., & Stanovnik, P. 2001. Development and innovation potential
of Slovene manufacturing industry: analysis of an industrial innovation survey.
Technovation, 21, pp 311-324
Lall, S. 1993. Understanding technological development. Development and change, 24(4): pp
719-753.
Lee, J. 2000. Challenges of Korean technology-based ventures and governmental policies in
emergent-technology sector. Technovation, 20: pp 489-495.
Meadows, D.H. 1982. Whole earth models and systems, CoEvolution Quarterly, Summer pp.
98-108.
Miozzo, M. 2000. Transnational corporations, industrial policy and ‘War of incentives’: The
case of Argentine automotive industry. Development and Change, 31: pp 651–680
Motor Business International, Second Quarter 2000, South Africa’s automotive sector: On a
long-terms growth path? pp 110-129.
Morecroft, J.D.W., 1998. System Dynamics and micro worlds for policymakers. European
Journal of Operational Research, 1(59), pp 9-27.
Narayanan, K. 1998. Technology acquisition, de-regulation and competitiveness: a study of
India automotive industry. Research Policy, 27: pp 215-228.
Oerlemans, L.A., Pretorius, M.W., Buys, A.J., & Rooks, G. 2003. Industrial innovation in
South Africa: 1998 -2000. University of Pretoria: Pretoria
Rhys, G. 2000. Report on the Economic Prospects for the Automotive Industry in the UK and
Europe and its Impact on Ford Dagenham. Centre for Automotive Research, Cardiff
University Business School.
Sterman, J.D. 1991. A Sceptic guide to computer models, in Barney, G.O., et al (Eds).
Managing a nation: the microcomputer software catalog, Boulder, CO: Westview
Press, 209-229.
Sterman, J.D. 2000. Business Dynamics: Systems Thinking and Modelling for a Complex
World, Irwin McGraw-Hill.
Taymaz, E., & Ozcelik, E. 2004. Does innovation matter for international competitiveness
in developing countries? The case of Turkish industry, Research Policy, 33, pp 409-
424.
Thedti (Department of Trade and Industry South Africa). 2004. Current Developments in
SA Automotive Industry. Pretoria: Government Printer.
GDN_Role of System Dynamics in Policy Alignment1 Page 24 of 24
Vennix, J.A. 1990. Mental models and computer models: Design and evaluation of a
computer-based learning environment for policy-making. Den Haag: CIP-Gegevens
Koninklijke Bibliotheek.
Waddock, S., & Graves, S. 1994. Industrial performance and investment in R&D and capital
goods. Journal of High Technology Management Research, 5(1): 1-17.
Zhu, P, W., & Lundin, N. 2006. The impact of government’s fundings and tax incentives on
industrial R&D investment – Empirical evidences from industrial Shanghai. Economic
Review China, 17: pp 51-69.
