Remittances and Money Supply on the Economic Growth in the Philippines

I. INTRODUCTION
Background of the Study
The Philippines has become a major exporter of labor in which we possess one of the highest rates in Southeast Asian counties (Burgess & Haksar, 2005). The reported number of Overseas Filipino Workers (OFWs) according to the National Statistics Office was around 2.2 million last September 2011, and it was around 5.5% of the country’s total labor force. Given the number of OFWs, the inflow of remittances from other countries is staggering. The remittances recorded for 2011 were estimated to be 156.3 billion pesos which contributed to the Gross Domestic Product (GDP) of 2, 723.257 billion pesos for the same year. Remittances are estimated to be around 5.80% of the country’s GDP.

It verifies that labor migration significantly contributed to Philippines’ development through remittances (Orbeta & Abrigo, 2009). Remittances has helped stabilized the consumption and the disposable income of the household in the country. Indeed, remittances show that it is an important factor to support the Philippines especially in the balance of payments (Burgess & Haksar, 2005). Even before, as former President Ferdinand Marcos alleged that OFWs give a boost to the Philippine economy in the form of remittances. Also, the persistent influx of remittances to the Philippines could keep the economy safe from crisis. Such as the Asian Financial Crisis during the late 1990s wherein remittances surged up the country’s foreign reserve (Ang, 2009). Consequently, it spurred policymakers to empirically study on the probable substantial effect of remittances given that flow of remittances in the Philippines is increasing. Also, remittances for being a “potential source of funding for economic development” urged policymakers in studying remittances (Chami, Barajas, Cosimano, Fullenkamp, Gapen & Montiel, 2008).

Similarly, monetary policy can augment or affect the economic growth as part of its objectives (Hameed & Amen, 2011 May). According to Liang and Huang (2011, April), the objectives of implementing monetary policy is “price stability and maximum sustainable economic growth.” These objectives of monetary policy can be achieved by controlling money supply in the economy. Mathai (2012 March 28) mentioned that money supply, in attaining price stability and economic growth, “is a meaningful policy tool.” For instance, Pakistan attains price stability and economic growth through “targeting monetary aggregates (broad money and reserve money) in accordance with real GDP growth” (Hameed & Amen, 2011 May). Another, Feldstein and Stock (1994) showed that money supply, particularly M2, can possibly be used in controlling inflation and sustaining GDP growth rate. Likewise, Philippines tries to achieve these objectives through monetary policy of Bangko Sentral ng Pilipinas (BSP) (Philippines Monetary Policy, n.d.).

Besides, it is indeed that money supply has a positive relationship with Gross Domestic Product (GDP). For the reason that “an increase in real quantity of money causes the nominal interest rate to decline and real output to rise” (Hsing, 2005). In his study, increase of money supply (M2) has contributed to the increase of Venezuela’s real GDP. While in a study about relationship of money supply and inflation also have found a positive relationship between money supply and Gross Domestic Product (Qayyum, 2006). However, adjustments on money supply do not take effect on Gross Domestic Product right away. Mathai (2012 March 28) mentioned that current adjustments on money supply take effect on the production of goods market in the short run.

From there, the role of money supply control in achieving price stability and economic growth is observed to be significant. For the reason that the relationship of money supply to prices and economic growth is strong (Bando, 1998 September). As a result, it spurred researches on finding out the consequences of monetary policy (money supply) on macroeconomic variables.

Objectives
Accordingly, this study aims to study the substantial effect of remittances and money supply specifically to GDP over time in the Philippines. The following are the objectives of this paper: * To show the effect of remittances to economic growth

* To maximize the probable developmental benefits of having a substantial flow of remittances through policy recommendations * To show the effect of monetary policy (money supply) on economic growth.

Significance of the Study
The rationale of studying the effect of remittances to GDP in the Philippines is to empirically prove the relationship since the importance of remittances is emerging among Filipinos and the economy. In a developing country like the Philippines, we want to investigate the implications of remittances being a positive factor to GDP. Also, Chami et. al ( 2008) stated in their study that with the growing flow of remittances is “the challenge of unlocking the maximum societal benefit from these transfers.” While, the rationale of studying the effect of monetary policy (money supply) on economic growth is to infer the future economic consequences in implementing monetary policy (money supply) in the case of a developing country like the Philippines.

As we proceed, the background of Remittances and Money Supply in the Philippines will be presented. Also, the overview and roles of remittances, and monetary policy in economic growth are included to show the basis of this study. Afterwards, we present the theories that will compellingly prove the relationship of variables included for the model. Then, the data and methodology will be presented to show the techniques to be used in answering the objectives of this paper.

Scope and Limitations
Since the objective of this study is to show the effects of remittances and money supply (M3) to GDP, we disregard other determinants of GDP. We all know that the government also makes an adjustment with regards to fiscal policy- government spending and taxes- that can affect the economic growth. Yet, the model in this study only includes the monetary policy measured by money supply (M3) and remittances. This is so because this study aims to determine if monetary policy (money supply) and remittances affect economic growth. In doing so, we will use dataset of remittances and money supply (M3) from 1977 to 2011 giving us 35 observations. II. REVIEW OF RELATED LITERATURE

Labor Migration and Remittances
In today’s world where ‘globalization’ keeps on proliferating, the search for resources such as labor and capital becomes very common and prevailing. Countries are able to efficiently reduce their costs of production by means of importing inexpensive means from all over the world (McDonald, 2012). Likewise, those countries who afford to offer low-cost inputs will be able to capture a bigger market in accordance to the trending of globalization. More often than not, those emerging economies such as the Philippines have been continually exporting part of its labor sector to the rest of the world. With the support of the government, migration should be promoted in order to access foreign labor market (O’Neil, 2004). Given so, fertility rate is important in propagating labor force for the development of the country (Abella, 1994). Those countries that are abundant in labor tend to have the comparative advantage of specializing in labor-intensive goods. From the context of labor exports, it requires that these laborers have to migrate to designated workplace abroad; hence it introduces us to the term ‘labor migration’.

The reasons for labor-abundant countries to outsource part of its labor force to the world lies within the purpose of reducing the pressure of domestic unemployment while earning foreign exchange at the same time namely remittances (Wickramasekera, 2002). In the field of economics, the term remittances have evolved into different definitions based on its uses and contributions in economics. These are the different definitions of remittances in economics (Alfieri & Havinga, 2006): * It is an important source of income for households; it is the least responsive with respect to economic downturns. It may be personal, such as personal transfers, personal remittances and total remittances. * Remittances also reduce poverty cases in home countries, in this case the differences in the values of exchange rates positively affects the home countries.

From there, it can be observed that remittances can be macroeconomic and microeconomic level. For instance, remittances as important source of income can affect the household level of consumption which is at microeconomic level. While remittances as mean to reduce poverty cases in home countries is a macroeconomic level effect of remittances. As mentioned, this study focuses more on the effect of remittances at macroeconomic level particularly on Gross Domestic Product (GDP).

The Inflow of Remittances and its Role in the Economy
The role of remittance and its potential in boosting economic growth was already recognized in the beginning of the 1980s (Strauhaar and Vadean, 2006). As mentioned, definition of remittances varies across the organization and even in the central banks. According to International Organization of Migration (IOM), remittances refers to personal monetary transfers, which might include investment, funds and donations, a migrant makes to the home country (International Organization of Migration, n.d.). Remittances can be interpreted into 3 different categories of the balance of payments according to the International Monetary Fund (IMF): * Compensations of employees are the gross earnings of workers residing abroad for less than 12 months, including the value of in-kind benefits * Workers’ remittances are the value of monetary transfers sent home from worker residing abroad for more than one year * Migrants’ transfer represent the net wealth of migrant who move from one country of employment to another

Remittances may come in various forms depending on its usage. Indeed, the receiving countries deemed remittances in varying economic view. Although remittances have different impact on different countries, International Organization of Migrant (2003) generally agreed that remittances help reduce poverty at the house levels since the funds which migrant send home can be an important source of income and foreign exchange, which will enable a country to afford scarce imports or pay off external debts. Other paper also discussed the effect of remittance in relation to income distribution, poverty alleviation and individual welfare. The paper written by Strauhaar and Vadean (2006) explained the use of gini index in relation to income distribution. Some papers show that remittances are positively related to inequality in income distribution by saying that richer families are more able to pay the costs associated with international migration. It also named other papers which showed that increase in remittance had reduced unequal income distribution.

From this it concluded that remittance might be income convergence or divergence depending on empirical methods applied to measure income inequality. From the paper of Ratha (2003) and Taylor (1999), they discussed that remittance have impact to economic growth in the country. For the first reason, remittances are an important source of income for many low and middle income households in developing countries. And secondly, remittance provides the foreign currency for importing scarce inputs that are not available domestically. And lastly, remittances are deemed as an additional savings for economic development. But the magnitude of the development through the remittance will highly depends on the way the remittances have used in the country receiving remittances. Aside from the consumptions, remittances can also help stimulate economic growth when they are made in good use (Ali & Mim, 2012). For developing countries, it is believed that workers’ remittances will eventually become one of the largest sources of financial capital for the country when it is being used as savings by households or investments (Barajas, Chami, Fullenkamp, Gapen, & Montiel, 2009)

Labor Migrants and Remittances in the Philippines
In the Philippines, it is observed that the importance of remittances among Filipinos is growing. This is brought by the increase in the number of labor migrants also known as the Overseas Filipino Workers (OFWs) which is continuously increasing each year. The Philippine Overseas Employment Administration (POEA) reports that Filipinos have found employment in some 200 destinations in all over the world continents (Tan, 2012). As mentioned in the discussion paper of Tan (2012), total number of OFWs across the border is estimated at 3.8 million, about 4% of the population and 10% of the labor force in 2010. In 5-years’ time from 2005-2010, the amount of foreign earnings remitted by these OFWs constitutes around 11% of our GNP. The outflow or newly hired OFWs averaged at 320,000 which comprised about 7% of the new entrants to the labor force in year 2010 (Tan, 2012). Though the increase in the number of newly hired workers may indicate that there is an increasing demand for OFWs, it is indeed skeptical to have this kind of belief since the demand for laborers lies upon the economic performance of those countries.

As we can see, the amount of migrants increases significantly every ten years. However, the total number of migrants should be classified as either permanent or temporary. Permanent emigrant pertains to those people who sort to stay abroad permanently (Tan, 2012). This type of migrant will in turn deal a small scale negative effect to the domestic economy since this will lead to the brain-drain effect (Rivera, 2011). Despite the continuous increase in labor migration, annual economic growth did not happened to rise in line with the increase in the amount of remittances to the domestic economy. One possible explanation would be the economy might not be doing well in other sectors even with the help of remittances. In fact, the amount of remittances may also vary depending on their sectors and professions of these migrants.

As we can see from the table above, household service workers still accounts for the highest number of OFWs every year. These workers have been making significant contributions to the Philippine economy in the form of remittances for a long time already (Sicat, 2012). With these OFWs being deployed in various destinations, remittances are seen as one of the major economic agents for stabilizing growth (Cortez, 2007). By the end of 2009, the Asian Development Bank (ADB) found that the Philippines’ economic growth is less affected by the Global Financial Crisis mainly due to its few or limited exposure to global financial markets and is supported by remittances (Asis, 2011).

Monetary Policy and its Role
The government or a country’s central bank has the jurisdiction in the implementing monetary policy. Under this policy, the government or the central bank has a control over the money supply and eventually interest rates so to achieve stability of prices and economic growth (Hameed and Amen, 2011 May). Money supply can be of control by implementing contractionary and expansionary monetary policy. Contractionary monetary policy and Expansionary monetary policy is about decreasing and increasing the money supply in the economy, respectively (Monetary Policy, n.d.). By decreasing money supply, it results to an increase in interest rates which cause the investments and GDP to decrease. Investments are to decrease because profits will be lesser for them with high interest rates. Conversely, investments and GDP will increase due to money supply growth and low interest rates. Both of these can be employed by using the different monetary policy tools such as Open Market Operations, Reserve Requirement and Discount Rate (Liang and Huang, 2011 April).

Open Market Operations can be used to increase or decrease money supply by selling or buying bonds (Investopedia, n.d.). By buying bonds, the central bank actually releases money in the circulation which results to an increase in money supply. Whereas selling of bonds decrease money supply because central bank takes the money from people. Reserve Requirement can also affect the money supply because this is the “amount of funds that banks must hold in reserve” (Investopedia, n.d.). If the reserve requirement increases, the money supply decreases. This is so because the banks are limiting the outflow of money to the circulation. Then, the money supply increases once the reserve requirement decreases because the outflow of money to the circulation is not limited. Lastly, the rediscount rate is the “rate of interest charged to member banks” (InvestorWords.com). The money supply decreases if the rediscount rate increases, vice versa.

Money Supply
The money supply that has been mentioning above can be the cash in your pocket or at time deposit. As defined by Investopedia (n.d), money supply is a collection of “currency and other liquid instruments” in the economy which is controlled by the government using different tools of monetary policy. Moreover, money supply in an economy is actually divided according to its type. Each country may have its own classification of money supply. But, money supply can be categorized into four namely M0, M1, M2 and M3 in general (Investopedia, n.d.). Given so, the cash in your pocket and time deposit can be included in a different category. The category M0 and M1 also known as the Narrow Money includes the type of money that can be exchanged into cash easily. Usually, the money included in the Narrow Money category is used as a medium of exchange which includes currency in circulation, demand deposits, traveler’s check and other checkable deposits.

For M2 or Broad Money, it is the “broader measure that also reflects money’s function as a store of value” (Schwartz, n.d.). This includes money under the M1 category, savings deposits, time deposits, money market accounts and other close substitute for transaction. Lastly, the category M3 or Total Domestic Liquidity is “still broader measure that covers items that many regard as close substitutes for money” (Schwartz, n.d.). It is regard as broad because it contains all the money from M2 category and government securities. Provided that, numerous studies used one of these types of money supply in assessing the relationship of money supply with macroeconomic variables such as economic growth and inflation. For instance, Hameed and Amen (2011 May) studied the impact of monetary policy on GDP IN Pakistan for the past 30 years.

They used M2 category in assessing the effects of changes in monetary policy to GDP. They found out that implementation of monetary policy in Pakistan has contributed in achieving Pakistan’s economic growth. In the case of the United States, Ling and Huang (2011 April) also studied the effect of monetary policy to economic growth using M2 measuring money supply and GDP measuring economic performance. They found out that change in M2 leads to change in GDP. Since these countries’ objectives of implementing monetary policy are the same with the Philippines, this study attempts to show the effect of changes in monetary policy to GDP. We want to show if sustainable economic growth is truly achieved and contributed by the adjustments made by the government in the supply of money. Thus, we will show this by using M3 as the measurement for monetary policy and GDP for economic growth. The category M3 will be used because it is the “broader definition of money supply” among others (Central Bank of Egypt, 2011).

III. THEORETICAL FRAMEWORK
Classical or Developmentalist theory
In the developmentalist era of the 1950s and 1960s, it was widely assumed that, through a policy of large-scale capital transfer and industrialization, poor countries would be able to jump on the bandwagon of rapid economic development and modernization. In the same period, large-scale labor migration from developing to developed countries began to gain momentum. Many developing countries became involved in the migration process amidst these expectations of the dawning of a new era. Governments of developing countries started to actively encourage emigration since they considered it as one of the principal instruments to promote national development.

Developmentalist migration optimists tend to think that migration leads to a North-South transfer of investment capital and accelerates the exposure of traditional communities to liberal, rational and democratic ideas, modern knowledge and education. From this perspective, return migrants are perceived as important agents of change, innovators and investors. The general expectation was that the flow of remittances as well as the experience, skills and knowledge that migrants would acquire abroad before returning would greatly help developing countries in their economic take-off (de Haas, 2007). Return migrants were expected to invest large sums of money in enterprises in the country of origin.

Informal Contracts
Aside from the Developmentalist theory, a theory of informal contracts supports our intuition that remittances do guard the economy against “temporary shocks” by providing monetary and in-kind transfers to the household in times of crisis [ (Englama, 2009) ]. According to the literature, there are actually two types of transactions between agents – the formal contracts and the informal contracts [ (Dubois, Jullien, & Magnac, 2007) ]. According to the study of Dubois, Jullien, & Magnac (2007), formal contract is used as an enforcement tool for informal lending-borrowing agreements in a risk-sharing context. In line with remittances, individuals receiving large transfer of money due to an income shock are more likely to concede or exchange favourable terms on the formal contracts [ (Dubois, Jullien, & Magnac, 2007) ]. Henceforth, this condition raises the future prospects of the other party by providing more incentives. On the other hand, informal contracts refer as the transfer of money with no legal contract enforceability (Thomas and Worrall, 1988). This happens when there is a voluntary transfer of money from lending-borrowing activity in the presence of limited commitment. Hence, the theory of informal contracts of insurance can be used in the context of remitting activities. By all means, when there is an income shock in the family of the migrant, informal contracts usually take place so that the family will be better off from bad times.

Monetary Theory
Adjustments made in the money supply have effects on achieving the economic goals such as sustainable economic growth. This statement is proven by the Monetary Theory that defined “monetary policy manipulates the money supply and rate of interest in such a way to achieve the goals of the manifestation of the ruling party (Shoaibk, 2010). Moreover, Monetary theory also asserts that “monetary policy provides a logical relationship between its variables stipulated to affects the outcomes” (Gul, Mughal & Rahim, 2012 May). Monetary theory sets as the ground for claiming that changes in money supply indeed affect economic growth. The government uses money supply as a tool to achieve or sustain economic growth. Given that the sustainable economic growth is part of the objectives in monetary policy, it compels the relationship of money supply to economic growth. IV. Methodology

Data Variables
For this study, we will be using annual data on GDP (Gross Domestic Product), remittances, and M3 of the Philippines from 1977 to 2011 obtained from the World Bank data base. The unit of GDP and remittances is current US dollars and the unit of M3 is percentage of GDP. The table below shows the summary of variables included. EndogenousVariable: lnGDP| The specification of Gross Domestic Product based on current US dollars transformed to logarithmic form.| Exogenous Variable| Description| A-priori Expectation|

lnREMITTANCES| The specification of Remittances based on current US dollars transformed to logarithmic form.| As remittances increase, the Gross Domestic Product will also increase because remittances also have the potential to promote economic growth (Mohamed & Sidiropoulos, 2010).| lnM3| The specification of M3 (Domestic Liquidity) which is in percentage of Gross Domestic Product transformed to logarithmic form.| As M3 (Domestic liquidity or Money supply) increases, the Gross Domestic Product will also increase (Krugman & Obstfeld , n.d.).| Initial Tests

Since the data we are using is time series data, we must first determine whether or not our data is stationary or non-stationary. One of the important assumptions of the Classical Linear Regression Model (CLRM) is that time series data used in empirical work must be stationary. The violation of this assumption or the presence of non-stationary data may lead to spurious regressions which is very undesirable because it gives misleading results.

We want to determine if the inflow of remittances and Money supply (M3) affect the GDP over time in the Philippines. So, the general model we will be using for this study is:

A summary of the data used in the regressions is shown below.

Initial Regression
Using Ordinary Least Squares, we take the regression of lnGDP, lnREMITTANCES, and lnM3. The results presented below show that lnREMITTANCES and lnM3 are both significant exogenous variables since their p-values are less than 0.05 and the R-squared for the model is 0.9570 which indicates that the percentage of variation in the endogenous variable is greatly explained by the variation in the exogenous variables.

Unit Root Test
Data is said to be stationary if its mean and variance are constant over time and the value of the covariance between two time periods depends only on the distance between these time periods and not the actual time at which the covariance is computed (Gujarati & Porter, 2009). To check the stationarity of our data, we used the unit root test or the Augmented Dickey-Fuller Test in Gretl. The Augmented Dickey-Fuller Test allows us to check if unit roots are still present because the presence of a unit root will allow us to convert a non-stationary time series data set to a stationary one by the use of the difference transformation. The null hypothesis of the Augmented Dickey Fuller test is that there is still a unit root, while the alternative hypothesis states that the data series is already stationary.

Using the software Gretl, the results of the ADF test for the initial level are presented below:

The majority of the asymptotic p-values of the three variables are greater than 0.05 so we accept the null hypothesis therefore, they are both non-stationary at the initial level. Since the variables are not stationary, then unit roots must still exist. Running the stationarity test again on each variable’s first difference, we get the following results:

Majority of the ADF Test Statistics for the1st difference of each variable are lower than their critical values at the 5% level, which means that the null hypothesis of the existence of a unit root is rejected, making the variables stationary at 1st differencing. Therefore, lnGDP, lnREMITTANCES, and lnM3 are all stationary at their 1stdifference. For more confirmation on the stationarity of the three variables at their 1st difference, let us look at their corresponding correlograms:

The data are stationary if the red lines are within the two blue lines. Looking at each correlogram, no red line exceeds the blue line therefore, the correlograms presented above further confirms the finding that lnGDP, lnREMITTANCES, and lnM3 are all stationary at their 1st difference.

Spurious Regression
In regressing non-stationary time series data, there is a tendency that the results will become non-sensical spurious regressions. Some indicators of spurious results are having a high R-squared, significant p-values, and intuitive signs. To test if the model suffers from spurious results, we may use the Durbin-Watson statistics. If the value of R-squared is greater than the Durbin-Watson d-statistics, then there is a “great” likelihood of having spurious results.

Based from the result above, the Durbin Watson d-statistics is 0.4773035 which is less than the R-squared of 0.9570 presented in the initial regression. Hence, we can say that there is high likelihood for our result to become spurious.

Cointegration Test
Cointegration means that there exists a long run or equilibrium relationship between variables (Gujarati, 2003). Two or more variables may be cointegrated even if they are not stationary, which guards us against spurious regression results. If the data are cointegrated, then this means that there is a relationship between the variables and a meaningful inference could be made from the results. Moreover, a cointegrating relationship signals that there are related studies supporting our model. If, however, the variables are not cointegrated, then the resulting interpretation and results may be considered spurious. For cointegration, the cointegrating relationship should be at most (M-1) where M refers to the number of variables used in our model.

Using the Johansen Test for Cointegration, we arrive at the result below:

Since there is a star in rank 1 under the trace statistic, it tells us that there is at least one cointegrating relationship in this model. These results tell us that indeed, the variables have a relationship with each other, and thus, the results would not be spurious.

Optimal Lag Structure
Next step is to determine the optimal lag length for this model by using the top-down approach.

Based from the results above, since there are more stars in lag 3, it means that the optimal lag length to be used in this model should be three lag periods.

Causality Test
To test for causality, we used the Granger Causality Analysis in assessing which economic agent affects which. This test allows us to see whether or not there is a directional relationship between the three variables of interest. Our null hypothesis tells us that the variable under consideration does not granger cause the other. We lagged lnGDP for three periods since it is the optimal number of lags.

Based from the results above, we can see that the p-value of lnGDP to lnREMITTANCES and lnM3 at all periods are greater than 0.05. Therefore, we accept the null hypothesis by saying that lnGDP does not granger cause lnREMITTANCES and lnM3 at the current period and future periods. Furthermore, since the p-value of current lnGDP to future lnGDP is higher than 0.05, we can say that current lnGDP does not granger cause the future lnGDP. Moving on to the next variable, we can see that lnREMITTANCES at the current period granger causes lnGDP at the current period and two periods from now and it does not granger cause lnM3 at the current period. Moving on to the last variable, we can see that lnM3 at the current period granger causes lnGDP at current period and does not granger cause lnREMITTANCES at the current period. In addition, since the p-value of lnM3 at the current period to the future periods of lnGDP is greater than 0.05, we can say that lnM3 does not granger cause lnGDP in the future.

STATIC MODELS
Since we’ve already tested for the requirements of a time-series data set, we proceed to the Ordinary Least Squares regression (OLS), or the static model. This regression is yet accurate for it has yet to be tested for Classical Linear Regression Model (CLRM) violations. However, from here we could check the significance of each regressor on determining the regressand. The results of the OLS regression and the resulting checks for violations are shown below.

OLS regression

The result of the regression shows that both lnREMITTANCES and lnM3 are significant variables giving us the model:

Test for Multicollinearity

If the VIF value is less than 10, then the model does not suffer from multicollinearity. Given that the result from the VIF test presented above has a value of 7.50, then multicollinearity does not exist in our model.

Test for Autocorrelation

In the Breusch-Godfrey LM test for autocorrelation, the null hypothesis is that there is no autocorrelation while the alternative hypothesis suggests that there is autocorrelation. The p-value shown in the results above is less than 0.05. Given this, we reject the null hypothesis and accept the alternative hypothesis, therefore the model suffers from autocorrelation.

Correction Measure for Autocorrelation
The tests for violations show that autocorrelation is present in the model. Autocorrelation is endemic in time-series data and autocorrelation in the data is further verified by having a result of 0.0000 under “Prob> chi2.” To correct for autocorrelation, we used the first difference correction measure.

The results show that both lnREMITTANCES and lnM3 are still significant variables. To check again for autocorrelation, we run the Breusch-Godfrey LM test for autocorrelation once again.

Based from the Breusch-Godfrey LM test for autocorrelation, the p-value is now greater then 0.05, which means that we accept H0 by saying that the corrected model presented above is now free from autocorrelation.

Results of the Model
lnGDPt=-0.7316106+0.0814406lnREMITTANCESt-0.2768997lnM3t
The results for the final model show that lnREMITTANCES is a significant variable due to a p-value of 0.012. Based on the final model, remittances have a positive effect on the gross domestic product of the Philippines as shown above wherein the GDP increases as remittances increase. From the results above, we can say that 1% increase in remittances will increase GDP by .081%. With money supply (M3), it has a negative effect on the gross domestic product of the Philippines yet it is a significant variable due to a p-value of 0.014. Since the results from the static model show that money supply (M3) is counter-intuitive, we will execute its dynamic model. Dynamic Models

The Static model uses ordinary least square regression which only takes into account the immediate effect of the changes in the exogenous variables on endogenous variable. However, we should consider that the policies being implemented do not have an immediate effect which is why we need to use dynamic econometrics to further analyze the effect of remittances on gross domestic product. Dynamic models allow us to analyze the effect of economic actions over time or to see the lag effect of X on Y.

There are three types of dynamic models namely Distributed Lag model (DL (p)), Autoregressive model (AR (q)), and Autoregressive Distributed Lag model (ARDL (q, p)). The Distributed Lag model only considers the effect of a policy variable at time t up to time t-p. The Autoregressive model only considers the effect of target variables at time t up to t-q. The combination of the first two types of Dynamic model is the Autoregressive Distributed Lag model. This includes the values of both policy and target variables over time (Gujarati & Porter, 2003).

In this study, we will be using the Distributed Lag model given that the regression model only includes the effect of policy variable at time t up to t-p. In using this, we will undergo two approaches specifically the Koyck Infinite Distributed Lag Model and the 2 Stage Least Squares Approach.

Koyck Infinite Distributed Lag Model
After checking that our variables are stationary, we will now proceed with estimating the effect of remittances on GDP using Koyck or Infinite Distributed Lag Model.

In measuring the effect of money supply to GDP,
β1=0.002447572×0.01035258×100%=0.002534%
This means that as money supply increase by 1% in the previous year, GDP will increase by 0.002534% this year. Likewise, another explanation could be: “for every 1% increase in money supply this year, GDP next year will increase by 0.002534%. Based from the result above, we see that only GDP lag at period 1 is significant. Hence, we decided to measure only the impact of remittances and money supply to GDP with 1 lag period.

Under the Koyck Model, the Long Run (LR) Multiplier is given as:
β=j=0∞βj=β01-λ
From table 1, we know that β0is 0.001872372 or 0.1872372%and the coefficient of decay (λ) is 0.01035258or 1.035258%, we can solve for the long run multiplier by substituting the values into the equation, we get:

β=β01-λ=β011-λ=0.001872372×11-0.01035258×100%=0.1892%
The long run multiplier tells us that a sustained growth in remittances will increase our GDP by 0.1892% in the long run.

For money supply, the long run multiplier can be obtained through:
β=β01-λ=β011-λ=0.002447572×11-0.01035258×100%=0.2473%

This tells us that a sustained growth in money supply will increase our GDP by 0.2473% in the long run.
Testing for Autocorrelation

Using the Breusch-Godfrey test for autocorrelation, the null hypothesis tells us that there is no autocorrelation in the model. Since the p-value is greater than 0.05, this means that we accept the null hypothesis and therefore, the model is free from autocorrelation.

Mean Lag jis given by:
j=λ1-λ=0.010352581-0.01035258=0.0105Years
Median Lag jis given by:
j=-log10(2)logλ=-log10(2)log0.005667=0.1517 Year
It takes less than a year to attain the first half of the total change in the target variable (GDP).

Two Stage Least Squares (2SLS)

The first equation is given by:
lnREMITTANCES=28.37573-2.478304lnINTEREST+u1t

Based from the results above, we can see that the p-value of lninterest is less than 0.05; therefore it is significant in this model. Given the results of the regression, a 1% increase in interest will decrease remittances by 2.4783%.

To get the second equation, we first have to predict y1hat.

lnGDP=15.20652+0.2419573Y1+1.223357lnM3+u2t

The second equation is given by:
lnGDP=13.75668+0.5058984lnREMITTANCES+u2t

Based on the results above, a 1% increase in remittances will increase GDP by 0.5059%.

Results of the Two Stage Least Squares (2SLS)

lnremittances=-34.13432+2.223614lnGDP+0.3095966lninterest+u1t
lnGDP=13.31824+0.5857155lnREMITTANCES+0.3699496lnM3+u2t

The final results of the Two Stage Least Squares show that a 1% increase in remittances will increase GDP by 0.5857155% and a 1% increase in M3 will increase GDP by 0.3699496.

Using the Hausman test, we check for simultaneity bias – if the coefficient associated with e1hat is significant, then there exists simultaneity bias.

Since the p-value of e1hat is less than 0.05, it shows that our model has simultaneity bias.

V. Conclusion
It is said that our greatest export is overseas Filipino workers or OFWs who go abroad to earn for their loved ones here in the country. With the millions of Filipinos working overseas comes millions of dollars worth of remittances and these remittances have surely been a major source of funds for many Filipino families. The staggering amount of remittances that enter our country from different countries has certainly caught the attention of several economists and has been a focus of several studies. The question of how the inflow of remittances affects the GDP of the Philippines is what this study aims to answer. We also included the role of money supply, particularly M3, in this study to have a policy variable at hand. In addition to determining the effects of remittances, we also want to see how M3 affects the GDP.

The data we used in this study is time series data from 1997 to 2011 and the data is not stationary in its initial level however after using the Augmented-Dickey Fuller test, we were able to transform it to a stationary one using the differencing technique: all three variables were stationary at their first difference. After that, we checked whether or not our variables were cointegrated with each other using the Johansen Cointegration Test. The results showed that the variables are in fact are cointegrated and that there is at least one cointegrating relationship in this model. We also checked to see the relationship of the variables using the Granger Causality Test and it showed that remittances and M3 do have an effect on GDP.

After doing all the tests necessary for time series data and a static model, we moved on to the processes involved in a dynamic model. Our results from the Koyck Infinite Distributed Lag Model show that for every 1% increase in remittances, GDP will increase by 0.1872% and for every 1% increase in M3, GDP will decrease by 0.2445%. This presents a positive and significant relationship between GDP and remittances and M3 which satisfies our a-priori expectations and is consistent with economic theory and literature. We also used the Two Stage Least Squares approach to determine how remittances and M3 affect GDP. The results show that a 1% increase in remittances will increase GDP by 0.5857155% and a 1% increase in M3 will increase GDP by 0.3699496 which again is intuitive. Given that both the results of Koyck and 2SLS satisfy our a-priori expectations, we conclude that remittances and M3 have a positive relationship with GDP.

BIBLIOGRAPHY

Abella, M. I. (ed.) 1994 “Turning points in labor migration”, Asian and Pacific Journal 3 (1): special issue.

Alfieri, A. &Havinga, I. (2006). Issue paper: Definition of Remittances. United Nations department of economic and social affairs statistics division. Retrieved fromhttp://unstats.un.org/unsd/tradeserv/TSG3-Feb06/tsg0602-14.pdf

Ali, M., & Mim, S. (2012, August 27). Through Which Channels Can Remittances Spur Economic Growth in MENA Countries? Retrieved from University of Sousse: http://www.google.com.ph/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&sqi=2&ved=0CBwQFjAA&url=http%3A%2F%2Fwww.economics-ejournal.org%2Feconomics%2Fjournalarticles%2F2012-33%2Fversion_1%2Fcount&ei=HD6NUIGYH8H1mAWr2ICYCQ&usg=AFQjCNFuKHHdGbIaL7RUvX3CPGQ

Ang, A. (2009). Workers’ remittances and economic growth in the Philippines. Retrieved from http://www.degit.ifw-kiel.de/papers/degit_12/C012_029.pdf

Asis, M. M. (2011). The Global Financial Crisis and International Labor Migration in the Philippines. Retrieved from http://smc.org.ph/misa/uploads/country_reports/1285920115.pdf

Bando, S. (1998 September). Money Supply Control in ASEAN countries.
Retrieved from http://www.jri.co.jp/english/periodical/rim/1998/RIMe199804moneysupply/

Barajas, A., Chami, R., Fullenkamp, C., Gapen, M., & Montiel, P. (2009, July). Do Workers’ Remittances Promote Economic Growth? Retrieved from International Monetary Fund: http://www.imf.org/external/pubs/ft/wp/2009/wp09153.pdf

Barnes, R. (2010). Investopedia. Retrieved from http://www.investopedia.com/articles/06/gdpinflation.asp#axzz2BMDL7OXJ

Burgess, R. & Haksar, V. (2005).Migration and Foreign Remittances in the Philippines. IMF retrieved from: http://www.imf.org/external/pubs/ft/wp/2005/wp05111.pdf

Chami, R., Barajas, A, Cosimano, T., Fullenkamp, C., Gapen, M., & Montiel, P. (2008). Macroeconomic Consequences of Remittances. Retrieved from http://www.nd.edu/~tcosiman/pdf/Publications/IMFOP/MacroeconomicConsequences_of_Remittances_IMFOP_259.pdf

Cortez, M.A.A. (2007). The human development implications of temporary labor migration: The case of the Philippines. Retrieved from the Proceedings of the International Migration, Multi-local Livelihoods and Human Security: Perspectives from Europe, Asia and Africa.

de Haas, Hein(2007). Remittances, Migration and Social Development: A Conceptual Review of the Literature. Retrieved from: http://www.unrisd.org/80256B3C005BCCF9/(httpAuxPages)/8B7D005E37FFC77EC12573A600439846/$file/deHaaspaper.pdf

Dubois, P., Jullien, B., & Magnac, T. (2007, June). Formal and Informal Risk Sharing in LDCs:Theory and Empirical Evidence. Retrieved from The Institute for the Study of Labor: http://ftp.iza.org/dp2842.pdf

Englama, A. (2009, March). The Economics of Remittances: Theories. Retrieved from Central Bank of Nigeria: http://www.waifem-cbp.org/v2/dloads/THE%20ECONOMICS%20OF%20REMITANCE.pdf

Ericta, C. N. (2012, June 26). Total Number of OFWs is Estimated. Retrieved from Census: http://www.census.gov.ph/data/pressrelease/2012/of11tx.html

Feldstein, M. and Stock, J. H. (1994) The Use of a Monetary Aggregate to Target Nominal GDP. National Bureauof Economic Researchp. 7-69

Gujarati, D. & Porter, D. (2009). Basic Econometrics. Singapore: McGraw Hill International.

Gul, H., Mughal, K.& Rahim, S. (2012 May). Linkage between Monetary Instruments and Economic Growth. Retrieved from http://cprenet.com/uploads/archive/UJMSS_12-1150.pdf

Hameed, I. & Amen, U. (2011, May). Impact of Monetary Policy on Gross Domestic Product (GDP). Retrieved from http://mpra.ub.uni-muenchen.de/35562/1/Impact_of_Monetary_Policy_on_Gross_Domestic_Product_GDP_by_Irfan_Hameed.pdf

Hsing (2005). Impact of Monetary Policy, Fiscal Policy, & Currency Depreciation on Output: The Case of Venezuela, Briefing Notes in Economics.

International Organization of Migration.(n.d.).IOM and Remittances: definition, scare and importance of remittances. Geneva. Retrieved from
http://www.iom.int/jahia/webdav/site/myjahiasite/shared/shared/mainsite/published_docs/brochures_and_info_sheets/IOM_Remittance_eng_pdf.pdf

Investopedia (n.d.). Money Supply. Retrieved from http://www.investopedia.com/terms/m/moneysupply.asp#axzz2EwYWmykX

Investopedia (n.d.). Open Market Operations. Retrieved from http://www.investopedia.com/terms/o/openmarketoperations.asp#axzz2EwYWmykX

Investopedia (n.d.). Reserve Requirements. Retrieved from http://www.investopedia.com/terms/r/requiredreserves.asp#axzz2EwYWmykX

InvestorWords.com. (n.d.). Rediscount Rate. Retrieved from http://www.investorwords.com/14727/rediscount_rate.html

Jeansonne, A. (n.d.). Loglinear Models. Retrieved from http://www.education.umd.edu/EDMS/fac/Hancock/Course_Materials/EDMS771/readings/LogLinearModels%20reading.pdf

Khathlan, K. (2012, August 6). The Link Between Remittances and Economic Growth in Pakistan: A Boon to Economic Stability. Retrieved from http://www.google.com.ph/url?sa=t&rct=j&q&esrc=s&source=web&cd=33&cad=rja&ved=0CDQQFjACOB4&url=http%3A%2F%2Fwww.sciencedomain.org%2Fdownload.php%3Ff%3D1344430150-Khathlan_2-3-12BJEMT1426.pdf%26aid%3D581&ei=OH-XUM64DaOkmQWi-4DoCQ&usg=AFQjCNH0nBbxQnrIFOILnnBOXoJ5efE1kA&sig2=cLSwiigTLaitIg3pgADogg

Liang, F. and Huang, W. (2011 April). The relationship between money supply and the GDP of United States.Retrieved from http://lib-sca.hkbu.edu.hk/trsimage/hp/08050597.pdf

Mathai, K. (2012 March 28). Monetary Policy: Stabilizing Prices and Output. Retrieved from http://www.imf.org/external/pubs/ft/fandd/basics/monpol.htm

Mohamed, S.E. &Sidiropoulos, M.G. (2010). Does Workers’ Remittances Affect Growth: Evidence from Seven MENA Labor Exporting Countries. Retrieved from http://www.eurojournals.com/IRJFE_46_14.pdf

Monetary Policy. (n.d.). Retrieved from http://edwardmcphail.com/intromacro/lecture14/lecture14.html

McDonald, B. (2012, March 28). International Trade: Commerce among Nations. Retrieved from International Monetary Fund: