Article Type : Research Article
Authors : Othmani A
Keywords : Global value chains; Monetary and fiscal policy interaction; OECD countries; Emerging Asian countries; Dynamic panel Model
This paper examines the effects of the monetary and
fiscal policies on global value chains (GVC's) in two panels of countries: the
Organization for Economic Cooperation and Development (OECD) and the emerging
Asian countries (EAC). We use a dynamic panel model during the period
(2007-2019). Our empirical results show that monetary and fiscal policies are
not significant in the OECD country model while they are significant and
negative in the model of emerging Asian countries. However, their interaction affects
positively the global value chains variable (GVCRT) in both groups of
countries. Finally, results reveal that global value chains are generally
explained by international trade variables namely the lagged year of global
value chains, the upstream and downstream participation of countries with their
trading partners also the export of final goods. Hence, these variables occupy
a decisive place in explaining the global value chains.
Global
value chains have become a dominant feature of global trade and Investment,
encompassing developing emerging countries and developed economies, Borin and
Mancini [1]. Approximately 70% of today's international Trade is based on
global value chains, GVC’s flows of services, raw materials parts, and
components crossing borders often multiple times. They are integrated at the
end of the chain into final products, which are then shipped to consumers
around the world, OCDE (2015). Exporting from one country to another is often
based on complex interactions between various local and foreign suppliers. Even
more than before, trade is determined by companies' strategic decisions to
outsource, invest and operate where the necessary skills and materials are available
on competitive terms in terms of cost and quality, OCDE (2015). The
unprecedented complexity of global networks potentially generates new
dimensions for the international transmission of monetary policy shocks that go
beyond standard textbook trade channels primarily driven by exchange rate
fluctuations, GFC (2013). The purpose of this study is to investigate the
effects of monetary and fiscal policies on global value chains related to trade
(GVCRT) in both groups of countries OECD
and emerging Asian countries EAC, according to Ca’ Zorzi, we try to
answer these questions in our study, can monetary policy spillovers account for
the high degree of cross-country co-movement in real and financial variables?
Can central banks in emerging Asian economies retain control over inflation and
real activity in the presence of monetary policy spillovers from systemic
advanced economies? This study examines the effects of monetary policy through
policy rates (INT (% GDP)) or long-term interest rates which are usually
averages of daily rates, measured in percentages. These interest rates are
implied by the prices at which government bonds are traded in financial
markets, not by the interest rates at which the loans were issued. These bonds
whose repayment of capital is guaranteed by governments. Long-term interest
rates are one of the determinants of business investment. Low long-term
interest rates encourage investment in new equipment and high-interest rates
discourage it. In a standard open economy environment, monetary tightening
influences the current account in two ways. On the one hand, imports are
affected by the contraction in domestic demand; on the other hand, the
subsequent appreciation of the national currency makes exports more expensive,
if we operate according to a local currency pricing paradigm, Ca’ Zorzi.
Foreign economies are only affected by the local monetary policy shock to the
extent that their Trade in goods and services depends on the local Trade
balance, Olamide [2]. Also, the effects of Fiscal policy using final
consumption expenditure (GGEXP (%GDP)) or general government expenditure,
indicating the size of government in countries. The wide variation in this
indicator highlights the diversity of countries' approaches to the provision of
public goods and services and social protection, not necessarily differences in
the resources spent. This indicator is measured in thousands of dollars per
capita and as a percentage of GDP. These two economic policies were carried out
by OECD countries and emerging Asian countries during the period (2007-2019),
our objective is to study their effects on the evolution of global value chains
linked to global trade (GVCRT (% Global trade) which is the dependent variable
and measures the value of goods and services exported by a sector or a country
that crosses more than one border, see Borin and Mancini. The interaction
variable (INTGGEXP (%GDP)) between monetary and fiscal policies is also
presented as an explanatory variable to check whether there is a coordination
of monetary and fiscal policies during this period. The other independent
variables which explain the endogenous variable are domestic production which
does not cross any border and measures the value of imported inputs in the
overall exports of a country (PDO(%Production overall du pays)), the variable
Participation pure upstream participation in GVCs (GVCPB (% Gross trade of a
country), the Pure downstream participation in GVCs variable (GVCPF (% Gross
trade of a country), Trade in intermediate goods (TIG (% Gross trade of a
Country) and Trade in Final Goods (TFG (% Gross Trade of a Country). The Stata
16 package program is used in the study. Data was compiled from the World Bank
(WITS). we find that the global value chains are generally explained by the
variables of international trade: the previous years of global value chains,
the upstream, and downstream countries, their trading partners also the export
of finished goods occupy a decisive place in GVCs, whereas in Asian countries
it is not significant. The export of intermediate goods is not significant in
both models group of countries, it is neglected by the presence of the export
of finished goods. Monetary and fiscal policies are not significant in the OECD
country model, while they are significant in the emerging Asian countries
model, but each affects the global value chains variable negatively, but their
interaction through the variable (INTGGEXP) positively affects the global value
chains variable (GVCRT). This paper contributes to the existing literature in
three ways. Firstly, this paper investigates the relationship between the
Global Value Chains and Monetary and Fiscal policy for the period 2007–2019.
The main objective is to verify on the one hand the diversion process of
international Trade through the global value chains from the OECD countries to
the Emerging Asian countries using the descriptive analysis. On the other hand,
test the effects of monetary and fiscal policy on global value chains in these
two panels of countries. Secondly, during the period of investigation including
the global financial crisis, most countries tended to improve financial
conditions and the significant monetary and fiscal policy measures put in place
by governments. However, several factors are expected to limit demand and
investment. Given these circumstances, we fill the vacuum in the topic by
analysing the effects of these policies’ measures on the evolution of global
value chains in the Organization for Economic Cooperation and Development
(OECD) and Emerging Asian countries (EAC). Thirdly, we apply both the static
and dynamic panel approach which allows us to test the relationship between the
variables related to Trade and the monetary and fiscal policy variables. This
study is becoming increasingly important after the current outbreak of
COVID-19. The pandemic is emerged in China and has swiftly spread globally. In
addition, its effects have quickly reverberated on the global value chain’s
process. Therefore, this survey can help policymakers to opt for the optimal
policies and to choose the better coordination of monetary fiscal policy in the
favour of growing Trade in a period characterized by the alternation between
normal and turbulent phases. The remainder of the paper is organized as
follows. Section 2 concludes with an
overview of the related literature. Section 3 explains the Research methods and
illustrates the choice of the dynamic panel Model. Section 4 presents empirical results
motivated by the model and discusses it. Section 5 concludes with a summary of
the main messages and policy implications that arise from the analysis.
The
relationship between global value chains and economic policies has been the
field of study for several economic authors, these studies are increasingly
adopted by scholars and practitioners as a means of understanding the global
organization of industries and their impact on development, De Marchi [3].
Global value chains will undergo a drastic transformation in the decade ahead.
The change will be driven by a push for greater supply chain resilience due to
COVID-19, which adds to existing pressures from the technology revolution,
growing economic nationalism, and the sustainability imperative. Based on
UNCTAD’s World Investment Report 2020, this column argues that the global trade
and investment landscape will be reshaped by the restructuring of global
chains, the build-up of new regional chains, and distributed manufacturing.
While these will present daunting challenges, they will also offer ample
opportunities for firms and states alike and will lead to a GVC development
paradigm shift Bolwijn. R, and Gereffi showed in his studies the attention paid
to Non-Equity Trade governed by powerful multinationals, the so-called “big
global corporations”, which are driving the development of industries around
the world [4-7]. Central banks generally only target inflation from the CPI, in
the presence of global value chains the economic literature launches the
studies between the CPI or the PPI is more appropriate for the objective of
monetary policy, De Paoli, De Gregorio [8,9]. These rapid and profound changes
in the structure of the world economy have influenced macroeconomic decisions
strongly on trade and industrial policy, which are becoming increasingly
complex also on monetary and fiscal policies strengthen domestic companies to
be more competitive and participatory in the global economy, Gereffi [10]. On the one hand, participation in global GVCs
led by lead firms can facilitate learning and enhance the development potential
of supplier firms and territories, especially in developing country contexts,
Gereffi. On the other hand, the ability to capture the value and scale through
participation in GVCs is not straightforward and requires government policy
intervention, given the diverse stakeholder interests and asymmetries between
large global corporations and their fragmented supplier base, [11-13]. Huang
and Liu [14] showed in their study that the interactions between multistage
production and economic openness and their implications for the design of
monetary policy have not been much explored. For example, as an economy becomes
more open, should the optimal upstream inflation weights increase or decrease
relative to the end-stage inflation weights? Should trade frictions such as a
tariff rate hike affect the design of monetary policy? The literature on monetary
policy introducing global value chains has been referred to by several economic
authors. In panel models, De Paoli demonstrates this in a model with a more
general parameterization but also a single production stage and focuses on the
externality of the terms of trade in the conduct of the optimal monetary
policy. Shi and Xu [15] developed a two-country New Keynesian model with trade
in vertical production, focusing on the cross-border ripple effect of the
productivity shock and the discussion of optimal money supply policy. To
explain international economic cycles, Huang and Liu constructed a two-stage
production model with staggering prices. Aoki, among the first works, studies
the optimal sector weights in the monetary policy rule when there are two horizontal
sectors. Lombardo and Ravenna examined the optimal monetary policy in two
(horizontal) sectors under a production stage with imported inputs or backward
participation for export. Matsumura also studied monetary policy in a small
open economy with several sectors but still with a single stage of production.
In his study, strum [16] showed that trade intensity and business cycle
synchronization depend on the substitutability between traded goods. Trade in
intermediate goods increases the synchronization of economic cycles between
countries, Kose and Yi. Global value chains involve greater trade in
complements, hence stronger co-movement between economic cycles of
production-sharing countries, De Soyres [17]. The COVID-19 crisis has reignited
discussions on the international fragmentation of production. In advanced
economies, most of the money has gone into preserving jobs and supporting
household incomes. The health sector was also strongly supported. Large
companies have benefited the most from business support compared to small and
medium-sized enterprises. In this group of countries, governments have mostly
introduced multi-year programs to promote and accelerate recovery. By
comparison, in emerging economies, most funds have gone into public works,
i.e., infrastructure projects, but also into maintaining employment through
other channels. Many measures tended to take place in 2020 and were only
minimally extended through 2021, IMF [18]. Although the efficiency gains from
GVC’s are well established, questions arise as to whether the gains from
deepening and expanding international specialization in GVCs outweigh the risks
and instability associated. The risks associated with GVC’s were initially
revealed in the very early phase of the pandemic. At the beginning of 2020,
when the public health situation in the People's Republic of China (hereinafter
"China) led to blockages. Most global manufacturers have operations in
China, and many companies have reported disruptions to production and trade in
this important GVC’s partner. Shortages of personal protective equipment (PPE)
as well as key respiratory medical devices, such as ventilators, have raised
greater concerns. Importantly, however, the global shortage of medical devices
stems from the unprecedented demand shock induced by the spread of the pandemic
around the world, not from the supply side. The global outbreak of COVID-19 in
2020 and its persistence in 2021 are generating severe economic damage and
deteriorating overall household health. Enderwick and Buckley [19] predict
stronger regionalization trends instead of deeper globalization. The authors
argue that the pandemic is raising concerns about the vulnerabilities of GVCs,
which may lead to de-globalization, hence however a renewed interest in
regional alliances, i.e., on neighbouring countries, could emerge well.
Regional alliances can overcome the disadvantages of small states while
benefiting from economies of scale. The macroeconomic effects caused by the
virus can be broadly categorized into demand-side and supply-side effects.
According to Meichenbaum [20], demand dynamics emerge because households
decrease consumption to reduce their likelihood of attracting the virus.
Evenett and Baldwin [21] argue that a renationalization strategy does not
promote resilience. Rather, they argue that alongside the COVID-19 crisis and
the threat of "vaccine nationalism", all the other pressing
challenges currently facing the global community be it the climate crisis, the
growing importance of digitization and e-commerce, and the trade war between
the United States and China, which also reflects a rivalry of modes of
capitalism must be addressed A number of recent columns in VoxEU warn against
the risks of a reversal in economic globalisation and an unprecedented
downsizing of the existing international production system [22-29]. In our
globalized world, production has become increasingly interconnected across
borders. The rise of global value chains (GVCs) implies that the production of
today’s manufactured goods hinges on inputs supplied from different corners of
the world, see Cantore. Such fragmentation of production increases global
efficiency as each country focuses on and specializes in the components it
produces best. Yet fragmentation also bears risks: when an unexpected event
disrupts production in a given location, its ripple effects quickly spread
across borders, potentially serious the global production of several goods. The
effects are even more severe if the disruption occurs in a location that contributes
strategic inputs in several value chains, Cantore. Global value chains have
become a dominant feature of global trade and Investment, encompassing
developing emerging countries and developed economies. Borin and Mancini
developed a natural measure of the importance of GVC’s Trade in total
international Trade. According to this measure from Borin and Mancini, the
overall share of GVC’s Trade in total world Trade increased very significantly
in the 1990 and early 2000, but it seems to have stagnated or even decreased
over the course of the decade of the last 10 years. Yet around half of the
global Trade appears to be GVC’s-related. The two components of GVC’s
participation, upstream (backward) and downstream (forward), can also be easily
calculated at the national and even sectoral levels. In doing so, it becomes
clear that the expansion of GVC’s activity has occurred unevenly around the
world. Participation in GVCs is much lower in other parts of the world, notably
in Latin America and Africa. The process of producing goods, from raw materials
to finished products, is increasingly fragmented and performed wherever the
necessary skills and materials are available at competitive cost and quality,
OECD (2015). Approximately 70% of today's international Trade is based on
global value chains, GVC’s flows of services, raw materials parts, and
components crossing borders – often multiple times. They are integrated at the
end of the chain into final products, which are then shipped to consumers
around the world. Exporting from one country to another is often based on
complex interactions between various local and foreign suppliers. Even more
than before, trade is determined by companies' strategic decisions to
outsource, invest and operate where the necessary skills and materials are
available on competitive terms in terms of cost and quality. The pandemic led
to massive capital outflows in emerging markets and massive currency
depreciations in this group of countries in early 2020. Confidence in the
stability of exchange rates and financial markets in many countries of the
Global South has collapsed. In particular, foreign direct investment (FDI) has
declined. Greenfield FDI fell in 2020 to 42% of the 1990s level, Altman and
Bastian [30]. Portfolio investment outflows showing signs of panic seemed to
trigger a new widespread financial crisis in many countries, IMF (2021) “a
crisis like no other” [31]. The world has been plunged into a recession. And
yet, many say the crisis could have been much worse, IMF (2021) [32]. The
macroeconomic consequences of the transformation of Global Value Chains, World
Bank Development Report [33]. Compared to traditional trade, in which producers
compete to serve foreign markets, GVCs are associated with a higher degree of
production complementarity across countries, as productivity and demand shocks
propagate backward and forward value chains. This results in faster and more
intense transmission of shocks between countries, as evidenced by natural
disasters such as the 2011 Tohoku Earthquake [34,35], and at a more aggregated
level, it also leads to higher co-movement of output and prices across
countries, i.e., larger firms, cycle synchronization and inflationary
spillovers, De Soyres and Gaillard. GVCs also weaken the effects of exchange
rate changes on the trade balance. For example, the positive effect of
depreciation on export competitiveness is counteracted by the increase in the
cost of foreign Value Added used in production. Disrupted supply chains of a
few essential goods and shortages of key medical products during the COVID-19
outbreak have highlighted the interdependence between countries through global
value chains (GVCs) and rekindled the debate on the costs and benefits of
globalization. More specifically, recent discussions emphasize the risks and
instability associated with the international fragmentation of production.
To
investigate the effects of monetary and fiscal policies on global value chains
related to trade (GVCRT) in both groups of countries OECD and emerging Asian
countries EAC, we have several countries from both groups (Appendices-Table1),
and we choose the variables explaining the global value chains in our model
(Table 1).
Descriptive statistics of OECD
country study variables
According to the results obtained, the number of observations used in the study is 429. We note that the maximum value of the global value chains variable (GVCRT) is 0.785, while the minimum value is (0.279). The standard deviation is (0.096), while the Average is (0.507) we can say that there is a strong dispersion around the Average. When the statistic skewness (0.310) is greater than zero, the distribution is said to be skewed and skewed to the right. Concerning the kurtosis coefficient. The value of this coefficient is equal to 3 if the distribution is normal, less than 3 if the distribution is more flattened, and greater than 3 if the distribution is less flattened. The kurtosis statistic here is equal to (2.837) is less than 3, and the distribution is less flattened. Concerning the probability of the Jarque-Berra statistic (0.025) we reject Ho, that is to say, that the distribution does not follow a normal law (Table 2,3) [36-46].
Descriptive statistics of the
variables of the study of emerging Asian countries
Depending on the results obtained; the number of observations used in the study is 208. We note that the maximum value of the variable of global value chains linked to trade (GVCRT) is 0.685, while the value minimum is (0.239). The standard deviation is (0.106), while the Average is (0.424) we can say that there is a strong dispersion around the Average. When the skewness statistic (0.652) is greater than zero, the distribution is said to be asymmetric and right-skewed. Regarding the kurtosis flattening coefficient. The value of this coefficient is equal to 3 if the distribution is normal, less than 3 if the distribution is more flattened and greater than 3 if the distribution is less flattened. The kurtosis statistic here is equal to (2.712) is less than 3, and the distribution is less flattened. Concerning the probability of the Jarque-Berra statistic (0.000) we reject Ho, that is to say, that the distribution does not follow a normal distribution. Looking at Figure 1 below, we can see that global value chains have evolved faster in emerging Asian countries than in OECD countries. This explains the diversion of international trade to Asia under the pressure of price competitiveness favoured by Asian countries, especially China (Figure 1).
The
correlation matrix of macroeconomic variables presented in which concerns the
study of OECD countries, informs us that five variables (PDO, GVCPF, TIG, INT,
and tries INTGGEXP) are negatively correlated to the Global Value Chain’s
variable. Values linked to trade (GVCRT) one of which is strongly correlated is
the variable (PDO) with a negative value (-0.8955), while the other three
variables (GVPCB, TFG, and GGEXP) are positively correlated with whose variable
(GVCPB) is highly correlated (0.8442). The correlation matrix of macroeconomic
variables presented, which concerns the study of emerging Asian countries,
shows us that six variables (PDO, GVCPF, TFG, INT, GGEXP, and INTGGEXP) are
negatively correlated to the variable of global value chains linked to trade
(GVCRT) one of which is strongly correlated which is the variable (PDO) with a
negative value (-0.8270), while the other two variables (GVPCB and TIG) are
positively correlated whose variable (GVCPB) is highly correlated (0.8485).
Stationarity of the variables
According
to the stationarity below and by testing the stationarity of the selected
macroeconomic variables, it was noticed that for the study of the OECD
countries, the variables are stationary at the level except the variable (PDO)
which n is not level stationery using the three stationarity tests but
stationary in first difference. For the macroeconomic variables whose sample is
the emerging Asian countries, we note that the majority of the variables are
stationary at a level including the variable (PDO) which is stationary
according to the test of (Levin, Lin, and Chu) at a level at 10%, while the
variables that are not stationary at the level are the variables of backward
participation in GVCs (GVCPB) and forward participation in GVCs (GVCPF) which
are stationary in first difference.
Static panel model
We
start our estimation by choosing a static model
Hausmann
test (Fixed effect model or random effect model)
The
test statistics are as follows:
Under
the null hypothesis of correct specification, this statistic is asymptotically
distributed according to a chi-square with K degrees of freedom, i.e., the
number of variable factors, Hausmann test. The estimate of the static panel
model in OECD countries shows that this model has a random effect (Prob>chi2
= 0.4908 > 0.05), we note that only three explanatory variables are
significant, of which two variables positively affect the variable of GVCs
which are backward participation in GVCs (GVCPB) with a coefficient of (1.591)
and the forward participation variable (GVCPF) with a coefficient of (1.515)
while the variable trade in final goods negatively affects the variable (GVCRT)
with a coefficient ( -0.074). The other variables including monetary and fiscal
policy variables are not significant. Estimation of the static panel model in
emerging Asian countries shows that this model has a fixed effect (Prob>chi2
= 0.0396 < 0.05). Only four explanatory variables are significant, two of
which positively affect the GVC variable, namely the upstream participation
variable (GVCPB) with a coefficient (0.612) and the downstream participation
variable with a coefficient of (0.453). The two other variables that negatively
affect the endogenous variable (GVCRT) are domestic production (PDO) with a
coefficient of (-0.194) and the trade in finished goods variable (TFG) with a
coefficient of (-0.238). The other variables including monetary and fiscal
policy variables are not significant.
Dynamic panel model
By
applying the dynamic panel model, the GMM estimator in the first differences of
Arellano and Bond (1991) takes the first difference of the equation to be
estimated to eliminate the individual specific effects for each period. We
obtain:
It
is then a question of instrumenting the endogenous variable delayed by its past
values of 2 periods and more. However, this method does not identify the effect
of time-invariant factors, see Nickell (1981). Moreover, Blundel and Bond
(1998) showed using Monte Carlo simulations that the system GMM estimator
performs better than the first difference estimator, the latter gives biased
results in finite samples when the instruments are weak. We formulate the
following equation:
With
X_ (i,t) : the set of explanatory variables which are the financial variables
GMM in first differences
The
GMM estimator in the first differences of Arellano and Bond (1991) consists of
taking the first difference of the equation for each period to be estimated to
eliminate the individual- effects. We obtain:
It
is then a question of instrumenting the endogenous variable delayed by its past
values of 2 periods and more. However, this method does not identify the effect
of time-invariant factors. In addition, Blundel and Bond (1998) showed using
Monte Carlo simulations that the GMM estimator in the system is more efficient
than that in the first differences, the latter gives biased results in finite
samples when the instruments are weak.
GMM in system
The system GMM estimator of Blundel and Bond (1998) combines first difference equations with level equations. The instruments in the first difference equation are expressed in level, and vice versa.
The
main dynamic panel tests are based on the following assumptions, which must be
accepted. Sargan's test has a null hypothesis (Ho): The Instruments as a group
are exogenous. Sargan's p-value should not be less than <5% and >10%.
However, according to Roodman (2006), it is recommended that the p-value of
Sargan be greater than 0.25.
Model:
With
X_ (i,t) : the set of explanatory variables which are the financial variables
Although the instruments are valid according to the
Sargan test in both estimates of the two samples. By comparing the two methods
of GMM one retains the method GMM in the system. According to the estimate of
the GMM dynamic panel model in the system in the two samples of countries:
OECD countries
By applying the system GMM method During the study
period (2007-2019), the global value chains variable (GVCRT) is explained only
by the country trade variables:
Emerging Asian countries
By
applying the GMM method in system during the study period (2007-2019), the
variable of global value chains (GVCRT) is explained by six macroeconomic
variables in the model studied: three of them are
Global
value chains offer a high degree of exposure and learning to the rapidly
changing, technology-based business models that characterize fragmented
production chains, even without the participating firms needing to engage in
the disposition of property, here comes the role of economic decision to
incentivize internationalization and the creation of added value to increase
the role of economic policy maker degree of participation in GVCs. By comparing
between patron the two trade blocks the OECD countries and the emerging Asian
countries, we find that the global value chains are generally explained by the
variables of international trade: the previous years of global value chains,
the upstream , and downstream countries, their trading partners also the export
of finished goods occupy a decisive place in GVC’s, whereas in Asian countries
it is not significant The export of intermediate goods is not significant in
both models group of countries, it is neglected by the presence of export of
finished goods. Monetary and fiscal policies are not significant in the OECD
country model, while they are significant in the emerging Asian countries
model, but each affects the global value chains variable negatively, but their
interaction through the variable (INTGGEXP) positively affects the global value
chains variable (GVCRT). The rebound in the global economy and the rise in
prices have pushed central banks on the path to the beginning of a reduction in
asset purchases, even if the challenges remain numerous: global logistics
tensions, commodity prices, and the trajectory of China. Market value action
levels and the end of “whatever it takes” are a challenge for 2022 which could
well mark the year of the normalization of the world economy. Analysis of
macroeconomic and market prospects for the coming months. The multiple
characteristics of GVCs that matter for production efficiency also determine
the exposure to international trade shocks and economic policy shocks and the
propagation of these shocks along the chain. A strong dependence of sales on
foreign demand and a strong dependence on foreign value added in production
govern the exposure to foreign supply and demand shocks. Governments still have
a role to play and practical economic policies can be outlined to foster the
diversification and resilience of GVCs while preserving the benefits of
specialization and ensuring efficient management an ensuring of essential
goods.
Static
panel model
H0:
random effect model
H1: fixed effect model
Dynamic GMM model
1st
Method: GMM in first differences
It is a dynamic method in first differences
·
(GVCRTi,t
– GVCRTi,t-1)
= ?(GVCRTi,t-1
– GVCRTi,t-2) + ?(Xi,t-X i,t-1) + (?i,t -?i,t-1)
·
?GVCRTi,t
=? ?GVCRTi,t-1+? ?i Xi,t
+ ??i,t
On
the other hand, we can use:
·
GVCRTi,t-2 as an instrument for
(GVCRTi,t-1 – GVCRT i,t-2)
·
GVCRTi,t-2 is a valid
instrument because it is not correlated with (?i,t - ?i,t-1)
·
GVCRTi,t-2 is a good estimator
because it is correlated with (GVCRTi,t-1 – GVCRTi,t-2)
·
The estimator in this case is said to be
identified
·
A better estimate is more efficient and
possible by using two additional lags of the dependent variable as an
instrument, (GVCRTi,t-3 – GVCRT i,t-4 …………….,
In
this case the estimator becomes over-identified, Anderson and Hsiao (1981) have
suggested:
A-First difference, the
model to be eliminated ?i
B-Using ?GVCRT i,t-2
= (GVCRTi,t-2 – GVCRTi,t-3) or
more simply
GVCRTi,t-2
As an instrument for ?GVCRTi,t-1 = (GVCRTi,t-1 – GVCRTi,t-2)
Sargan's
test specific
H0 : instruments are valid
H1 : instruments are invalid
OECD countries
·
In first difference (FD-GMM), p-value
=0.665 > 5%, we cannot reject H0 , therefore the instruments (L.
GGEXP, L2. GGEXP, L.INT, L2.INT, LD.PDO and L2D.PDO) are valid.
·
GMM in system (SYS-GMM), p-value = 0.537
> at 5%, we cannot reject H0, therefore the instruments (L.
GGEXP, L2. GGEXP, LD.PDO, L2D.PDO, L.TIG and L2.TIG) are valid.
Emerging
Asian countries
·
In first difference (FD-GMM), p-value
=0.878 > 5%, we cannot reject H0, therefore the instruments
(L.PDO, L2.PDO, L.TIG, L2.TIG, L. GGEXP, L2. GGEXP, L. GVCPB and L2. GVCPB) are
valid.
·
GMM in system (SYS-GMM), p-value =0.229
> at 5%, we cannot reject H0, therefore the instruments (L.PDO,
L2.PDO, L.INT, L2.INT, L. GGEXP, L2. GGEXP, LD.GVCPB and L2D.GVCPB) are valid.