/ Does the Introduction of a
Nationwide Minimum Wage affect
International price
Competitiveness?
Bachelor thesis
Name: Justus Cohen Tervaert Student number: 10048138 University: University of Amsterdam
Faculty: Economics and Business Field: International Economics Supervisor: Francisco Gomez Martinez
Date: April 23rd 2015
Abstract
Controversy arises whenever governments consider introducing minimum wage legislation. The effect of minimum wage on unemployment has been clarified by research. The effect on competitiveness is less clear. Wage structures are an important aspect of every competitiveness analysis. However, an emphasis on minimum wage legislation within research or policy analysis lacks. Empirical research is conducted to study the effect of a nationwide minimum wage on competitiveness. The presence of nationwide minimum wage legislation is included as a dummy variable, the natural logarithm of export numbers and of export differences over a year are used as a measure of cost competiveness. Over a sample of 24 European countries, using data from 1990 till 2014, results are inconclusive. Initial OLS on logExport are positive and significant. Additional regressions, on logExportDiff or using a First Differences approach, display contrasting results.
Table of content
1/ Introduction . . . . . . . 3
2/ Literature review . . . . . . . 5
Minimum wage and the Labour market . . . 5
Country Competitiveness . . . . . 5
Wages and Competitiveness . . . . . 6
Minimum wages and Competitiveness . . . 7
3/ Methodology . . . . . . . 9 Data . . . . . . . . 9 Method . . . . . . . 12 4/ Results . . . . . . . . 14 OLS: logExport . . . . . . 14 OLS: logExportDiff . . . . . . 15
OLS of First Differences of dependent variable only . . 16
OLS of First Differences of all variables . . . 17
OLS: logExport of 3 countries . . . . 18
OLS: logExportDiff of 3 countries . . . . 19
5/ Conclusion . . . . . . . . 20 6/ Further Research . . . . . . . 21 7/ Literature . . . . . . . . 22 8/ Appendices . . . . . . . . 25 Appendix 1 . . . . . . . 25 Appendix 2 . . . . . . . 26
1/ Introduction
Almost seven years after the financial crisis hit in 2008, Germany’s economy has survived the impasse strongly. Only as of 2004 the country was disapproved for not living up to its potential of being the European powerhouse it could be (The Economist, 2004). The German economy coped with falling competitiveness and a jammed labour market with high unemployment rates (Vishnevskaya, 2014).
As a response the so-called Hartz plans were implemented between 2003 and 2005. The reforms led to considerable improvements within short notice (Schmidt & Modrack, 2008). The Hartz-plans were labour-market reforms, that consisted of three aspects: improving employment services and policy measures, activating the unemployed, and fostering employment demand by deregulating the labour market (Jacobi & Kluve, 2006). In 2003 Germany was one of nine EU-countries without a minimum wage. Not including a nationwide minimum wage in the Hartz-plans was a notable choice.
General equilibrium theory reasons that as a minimum wage raises wage level above equilibrium, employers will hire fewer employees. Hence unemployment rises (Gärtner, 2009). For this reason the absence of a minimum wage is considered pivotal in Germany’s uprising of the last decade. The minimum wage would have prevented Germany from keeping its unemployment as low as it is. This logic is, however, largely contested by research. Card & Krueger (1993), Doucouliagos & Stanley(2008) and Baskaya & Rubinstein(2012) all have trouble drawing a conclusion which unilaterally supports the theoretical expectation.
The absence of the minimum wage in Germany might explain the exceptional economic performance for another reason. A minimum wage raises the wages above their equilibrium level. As minimum wages are decided upon on a national level, this creates asymmetries between countries. This way a minimum wage influences competitiveness; making a country less attractive for big companies to settle and creating a comparative advantage for countries without a minimum wage.
International competitiveness is a widely debated subject in economics (Aiginger, 2006). In the nineties its use in general was criticized by Krugman (1994). Since then, its relevance and meaning have been subject to on-going debate. There has been no clear consensus on how to measure it and which variables influence it. An indicator that is found in virtually all scientific or policy estimation is wage structure. High or vicious wages make a country less attractive compared to its international
competition. How this works has not been researched thoroughly. The specific effect of the presence of nationwide minimum wage legislation has hardly been researched at all.
As of the 3rd of July of 2014 the German Parliament has approved the country’s first federal minimum wage. Both in Germany up to the elections, and around the world when it was announced, there has been quite a lot of controversy about this decision (BBC, 2014). Again the threat for Germany’s competitiveness was articulated. This paper takes a look on this influence of having nationwide minimum wage legislation on international competitiveness. Since Germany has only agreed to install a minimum wage in July 2014, data about their economic performance with a minimum wage are scarce. Instead a wide sample is created consisting of 24 European countries and spanning over 24 years. This paper will address the research question: what effect does (nationwide) minimum wage legislation have on international price competitiveness?
To give an answer on the research question, there will be a combination between a literature analysis and empirical research. A literature review will shed light on the role of minimum wages in international competitiveness. The impact of having the minimum wage will be tested with OLS-regression. Exports is chosen as the dependent variable, as it is interpreted the same and easily accessible for all sample countries. Exports difference is also included as a separate dependent variable, to give insight in the effect of minimum wage on a change in exports over a year. The presence of a minimum wage is included as a dummy independent variable.
The remainder of this paper is as follows: section 2 provides a literature review, Section 3 contains the methodology where the data and method of the regression analysis will be discussed, section 4 contains the results from the empirical research, section 5 contains the conclusion, and sections 6 provides suggestions for further research.
2/ Literature review
Minimum wage and the labour market
Over the past thirty years the minimum wage has been a peculiar economic policy instrument. During a period generally determined by liberalisation, increasing numbers of (developed) countries have adopted a minimum wage, which intervenes with the price-setting process in the labour market. As the term reveals; a minimum wage sets a lower boundary for wages to be paid by employers. The most important incentive to introduce a minimum wage is to ‘combat poverty’ (Politics.co.uk, 2012), as was the case for the introduction in the United Kingdom in 1999. This combat is fought with two kinds of ammunition: reducing inequality and increasing aggregate demand (International Labour Organization, 2013).
Wage is set in equilibrium of labour supply and demand. Minimum wage raises wage
w above its equilibrium-level w*, thus causing (involuntary) unemployment. (Gärtner,
2009). However this theoretical effect is not supported by scientific evidence. In their book Myth and Measurement: The New Economics of the Minimum Wage (1995), Card and Krueger systematically explain and prove that contrary to economic theory and belief, the minimum wage does not increase unemployment. More recent study by Doucouliagos and Stanley (2008) and Baskaya and Rubinstein (2012) also fail to find evidence of unemployment as a consequence of the minimum wage, both for the short and the long run. If inequality were reduced without generating unemployment, a minimum wage would be very attractive policy tool, especially in the poverty increasing decades of the eighties and the nineties (Dolado, Felgueroso & Jimeno, 2000). However the effects of introducing a minimum wage could reach further than the labour market. An effect that is often mentioned is the effect on competitiveness. Country Competitiveness
The idea of country competitiveness has been distracted from the more familiar context of the firm. This idea of international competitiveness started gaining momentum at the end of 1970’s and materialized in the World Competitiveness Report, which the World Economic Forum started issuing in 1980 (Krugman, 1996). Krugman, one of the main critics of the use of competitiveness in a country context, claimed that by the 1990s the concept was ‘no longer even controversial among influential people’. In 1994 he criticized the widespread use of the concept calling it a ‘dangerous obsession’ (Krugman, 1994). However this obsession never vanished.
Debate, partly fired up by Krugman, has focused on the meaning of country competitiveness. An initial definition was given by Orlowski (1982) who dubbed it ‘the ability to sell’. This unilateral interpretation can be considered one side of the spectrum. A dominant extension of this view was Porter’s (1990) claim that the only relevant meaning is the level of national productivity. At the other end of the spectrum the emphasis lies not on the ability to sell, but on the ability to create welfare (Aiginger, 2006). Fagerberg(1988) gives a definition that lies in the middle of the spectrum:
The ability of a country to realise central economic policy goals, especially growth in income and employment, without running into balance of payment difficulties.
He combines both sides of the spectrum. Present-day policy institutions like the European Commission and the World Economic Forum find connection with this view (Aiginger, 2006). The World Economic Forum is the institute that issues the annual World Competitiveness Report.
Wages and competitiveness
Regardless of Krugman’s critique, competitiveness has kept relevance and attention both in policy and science circles. One of the focus points has been the identification of factors that contribute to or indicate competitiveness. A wide range of factors has been mentioned or researched.
In his paper on ‘International Competitiveness’, Fagerberg (1988) addresses this question of what factors influence the process. He identifies that in the 1980s relative unit labour costs was used as the principal measure of international competitiveness. By using Kaldor’s Paradox, the fact that the fastest growing countries in terms of exports and GDP in the post-war period have at the same time experienced much faster growth in relative Unit Labour Costs (ULCs), Fagerberg (1988) also marks the role of technology and investment as factors influencing competitiveness.
Ireland is an example of a country that has been researched quite extensively in the context of competitiveness. Honohan and Walsh (2002) underline that a gain in competitiveness was the catalyst of growth in output and employment that earned Ireland the nicknames ‘Celtic Tiger’ and ‘Irish Hare’. Cerra, Soikkeli and Saxena (2003) critically examine the competitiveness of the Irish manufacturing industry. They point out that not the entire industry has to be responsible for the growth, but possibly only a handful of strong sectors. They also highlight the role of external
factors. However, as Fagerberg started off, they put emphasis on the relevance of Irish wage costs. In another paper, Cassidy and O’Brien (2005) review the strong gains in market share Ireland experienced between 1990 and 2005. Again they specifically treat cost competitiveness, of which wage costs are a significant element.
Since 1997 Ireland has a National Competitiveness Council. It advises the Taoiseach (the Irish Prime Minister) on key competitiveness issues (NCC). In its annual competitiveness scorecard wages are considered as part of the ‘essential conditions’ (NCC’s Ireland’s Competitiveness Scorecard 2014). Other governments with government bodies for competitiveness include Spain, Greece and Saudi Arabia. The earlier mentioned World Competitiveness Report is the most prominent measure of competitiveness. Wages are also taken into account in the WRC.
So, as Cassidy and O’Brien (2005) state, wage costs can be considered ‘an important component’ of competitiveness. In all research and policy calculations this effect is considered to be negative. Dornbusch, Fischer and Samuelson (1977) explicate this effect as part of a Ricardian model of comparative advantage. The comparative advantage of an industry exists if its labour cost is inferior to the labour cost of the same industry in the foreign country. In other words; between two comparable industries in different countries, the industry with lower labour costs has an advantage. Carlin, Glyn and Van Rheenen (2001) researched a similar proposition of comparative advantage. They found a negative effect of relative unit labour costs on export market shares. Hence wage costs have an adverse effect on competitiveness. A nationwide minimum wage is an example of a policy measure that influences wage structures directly. The effect of nationwide minimum wage legislation on competitiveness is not clarified.
Minimum wages and competitiveness
Minimum wages could affect country competitiveness through two channels.
In the first place, it could mean a rise in domestic prices and hence, a deterioration of net exports. The labour market plays a role in macroeconomics; it is a component of the DAD-SAS model. This model is a dynamic interpretation of the interaction between the goods, the money, the foreign exchange and the labour market (Gärtner, 2012). Introducing a minimum wage raises the wage level and makes it less flexible. This increases domestic prices, which hurts competitiveness, and could reduce net exports. As Jeffrey Sachs (1980) describes it; in an open economy, which faces a
given world real interest rate, the path leads from real wages to international price competitiveness. The influence of wages on prices has been researched.
Aforementioned Card and Krueger (1994) recorded that higher minimum wages result in higher fast-food prices. With a bigger sample, covering both the United States and Canada, Aaronson (2001) also notes restaurant prices rising with rising wages. In a summary of several studies about the influence of rising wages on product prices, Lemos (2006) states that a 10% US minimum wage increase raises food prices by 4% and overall prices by 0,4%. Lee, Schluter and O’Roark (2000) temper the effect in a study on food prices, articulating that ‘price increases are small’.
Second, the minimum wage might make a country less attractive for countries to settle. The introduction of a minimum wage raises the wage level and makes wages less flexible. Withstanding international competition pushes firms to modify their production process to reduce costs, if not to relocate in countries with more flexible labour markets (Méjean & Patureau, 2010). This way a minimum wage reduces competitiveness, as firms will not settle in or relocate to a competing country with a minimum wage. Méjean & Patureau (2010) underline the importance of taking into account the impact of labour market policies on firms’ location decisions, but point out the relevance of substitutability between low-skilled and high-skilled labour. Javorcik and Spatareanu (2005) show that a more flexible labour market attracts Foreign Direct Investment (FDI).
The question this thesis poses and researches is if a minimum wage affects competitiveness, and so affects the economies of countries that introduce one. To research this question we use a narrow interpretation of country competitiveness in accordance with Siggel (2006):
The most consistent interpretation of the concept of competitiveness is the microeconomic notion of cost competitiveness.
This is also in line with the definition of the OECD, the Organization for Economic Co-operation and Development:
Competitiveness is a measure of a country's advantage or disadvantage in selling its products in international markets.
In the remainder of this paper this relation between minimum wage and country (cost) competitiveness is researched empirically.
3/ Methodology
Data
For this empirical research a dataset is created consisting of 24 countries and spanning over 24 years. The countries chosen are all the European member states of the OECD. In the first place, these countries are chosen on the basis of their membership of the OECD. This means the OECD has recorded data of all the countries investigated. For the same reason the OECD is one of the primary data-sources. By choosing only the European member states, geographic and cultural differences are initially reduced to those of the European continent.
In the same line of reasoning the time-sample is chosen for convenience. The period of 1990-2014 is well recorded. Besides, diversity in minimum wage legislation exists between the sample countries: in this period eight countries with a nationwide minimum wage throughout the time frame, eight without a nationwide minimum wage and six countries that introduced one during the sample, are present. As visible in Appendix 1.
The variables of interest are logExport (the natural logarithm of exports of a country) and logExportDiff (the logarithm of the difference between export in a year and the export number of the year before) as the dependent variables, and Minimumwage (a dummy variable of the presence of nationwide minimum wage legislation) as the independent variable.
Exports and the export differences are chosen as measures of competitiveness. The export number measures the aggregate of products and services produced in one country and sold to another. It is taken in million US Dollars. If competitiveness improves, countries’ products (or services) become more attractive relative to those of other countries, hence exports from that country will increase (Gärtner, 2009). In some literature Exchange Rates or Real Effective Exchange Rates are also used as a measure of competitiveness (Boltho, 1996), however as many countries in the sample use the same currency (i.e. the Euro) for part of the time sample, this is not preferred with our sample. As differences in Exports can vary widely, the natural logarithm is used to bring them more in proportion.
Extra regressions are conducted with logExportDiff as dependent variable. This is the natural logarithm of the increase or decrease in Exports over a year:
The independent variable is Minimumwage. The presence of a nationwide minimum wage is taken as a dummy variable: 1 means there is nationwide minimum wage legislation, 0 means there is not. The presence of nationwide minimum wage means that the minimum wage is set by the national government to apply to all industries in a country. Years of introduction for the national minimum wage legislation are found in Appendix 1.
The Exportnumbers that are used as a measure for economic competitiveness correlate and depend on a lot of factors, which need to be taken into account to get reliable estimates of the variable of interest (Minimumwage). To get a reliable result, control variables are used for GDP and will be included in the model. This will decrease omitted variable bias. These control variables are Neighbours (the number of neighbouring countries reachable by road or railway), GDP per capita (GDP per head of population), TradePercentage (the aggregate of imports and exports measured as a percentage of GDP), ExchangerateDiffPerc (percentage change of exchange rate with the previous year), HighLabourDummy (GDP per hour worked, divided in a dummy with value 1 for labourproductivity values of over 35), TertiaryDummy (a dummy for the percentage of the population with tertiary education, percentages of over 25% correspond with a value of 1), inflation (in %) and WorldEconomicPerformance (the change in the aggregate GDP’s of all countries in % relative to the previous year). The number of neighbouring countries is used as a control variable as the lion’s share of a countries’ trade usually flows to and from neighbouring countries. Germany is principal trading partner to 17 European countries. Nearly all of them are neighbouring or nearby countries (CBS, 2012). There are several reasons for this. Transport costs depend on geography and infrastructure (Limão & Venables, 2001). Cultural similarities might also contribute to significant trade between neighbouring countries. For this reason an increased number of neighbouring countries is seen as a stimulatory power for export numbers and, hence, used as a control variable.
GDP per capita is used as control variable. The Gross Domestic Product (GDP) of a country is the sum of the final uses of goods and services (all uses except intermediate consumption) measured in purchasers' prices, less the value of imports of goods and services, or the sum of primary incomes distributed by resident producer units (OECD). It is the main indicator of a countries’ economic performance. As GDP is not compensated for population size it is also partially a measure for country size. For this reason GDP per capita (i.e. per head of the population) is used as control
variable. With GDP per capita, the effect on exports is expected to be adverse. If welfare increases, products become more expensive and imports will increase at the cost of exports. Data on GDP per capita come from the OECD.
ExchangerateDiffPerc is also one of the control variables. It is a variable for exchange rate. If the exchange rate rises, domestic currency becomes more expensive relative to foreign currencies. That means domestic products become more expensive, which means they become less attractive and exports will decrease. The ExchangerateDiffPerc variable is expected to have a negative effect on exports. As the countries in the sample have different currencies in use, not the regular exchange rate is used as a variable. The change in exchange rate over a year is calculated, measured in percentages:
𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒𝐷𝑖𝑓𝑓𝑃𝑒𝑟𝑐!=
(𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒!− 𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒!!!)
𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒! (𝐴. 1)
The exchange rates inserted in the formula were measured in US$ and taken from the OECD. Off course, for all euro countries this value is equal from 2002 onward.
GDP measures the market value of final goods and services produced. Complementary Trade in % of GDP is used as measure of openness of a country. A higher percentage means a country is more open, which is a stimulatory indicator of exports. Data on Trade in % of GDP come from the World Bank.
Labourproductivity is measured as GDP per hour worked. In the first case it is a measure for productivity of the labour force. Second, it also indicates the level of technology in a country, as productivity increases if better production instruments are at hand. If productivity increases relative to competing countries in the world market, prices decrease relatively, which stimulates Exports. Fagerberg (1988) highlights the role of labour productivity and technology in relation to competitiveness. Carlin, Glyn and Van Reenen (2001) also conclude that technology factors have an effect on export performance. Labourproductivity is included as a dummy variable with values of over 35 corresponding to a value of 1, values below 35 correspond to a value of 0. Data on labour productivity come from the OECD.
TertiaryDummy is a dummy variable for the percentage of the population that has enjoyed tertiary education. Tertiary education is all education after the high school level (i.e. professional education). Literature suggests that the competitiveness of developed countries or countries with a high-skilled labour force is less influenced by labour market structures than less developed nations or nations with a less
high-skilled labour force. Dummy values of 1 are based on tertiary education percentages of 25% or higher. Data come from the OECD.
Inflation is the percentage rate at which prices increase. If inflation in a country is high, prices are rising relatively fast, which makes products less attractive. So exports decrease. Hence inflation is expected to have a negative effect on exports. Data on inflation come from the OECD database.
The variable WorldEconomicPerformance stands for the change in aggregate GDP of the world in %. For this measure the weighted average is taken of the increase or decrease every country in the world has made in a year relative to the year before. By averaging this aggregate, the global economic performance is measured. If world GDP is increasing it means the global economy is performing well. If the economy is doing well, both production and trade are increasing. Hence exports are as well. Data on the variable WorldEconomicPerformance come from the International Monetary Fund.
A complete overview of all the variable descriptions and its summary statistics can be found in appendix 2.
Method
To research the effect of Minimumwage on competitiveness, several Ordinary Least Squares (OLS) regressions are conducted.
The assumptions for OLS can be considered met. Of these assumptions exogeneity is the most questionable. As a nationwide minimum wage exists both in developed and in developing economies, has been introduced both in periods of high and low economic performance, with high and low employment-levels or inflation, we conclude that the introduction of a minimum wage is a political, not an economical, decision. Hence the exogeneity-assumption is met. The independent variable is assumed not to correlate with the error term.
The regression model used is:
𝑙𝑜𝑔𝐸𝑥𝑝𝑜𝑟𝑡! = 𝛽!+ 𝛽!𝑀𝑖𝑛𝑖𝑚𝑢𝑚𝑤𝑎𝑔𝑒!+ 𝛽!𝑁𝑒𝑖𝑔ℎ𝑏𝑜𝑟𝑠 + 𝛽!𝐺𝐷𝑃𝑝𝑒𝑟𝐶𝑎𝑝𝑖𝑡𝑎! + 𝛽!𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒𝐷𝑖𝑓𝑓𝑃𝑒𝑟𝑐!+ 𝛽!𝑇𝑟𝑎𝑑𝑒𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒!
+ 𝛽!𝐻𝑖𝑔ℎ𝐿𝑎𝑏𝑜𝑢𝑟𝐷𝑢𝑚𝑚𝑦!+ 𝛽!𝑇𝑒𝑟𝑡𝑖𝑎𝑟𝑦𝐷𝑢𝑚𝑚𝑦!+ 𝛽!𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! + 𝛽!𝑅𝑒𝑎𝑙𝐺𝐷𝑃!+ 𝜀! (1.1)
The two dependent variables are considered complementary. As logExport is a static number, which because of its size, might make results opaque in meaning. logExportDiff displays a increase or decrease of export numbers. This is a more specific interpretation of competitiveness:
𝑙𝑜𝑔𝐸𝑥𝑝𝑜𝑟𝑡𝐷𝑖𝑓𝑓!
= 𝛽!+ 𝛽!𝑀𝑖𝑛𝑖𝑚𝑢𝑚𝑤𝑎𝑔𝑒!+ 𝛽!𝑁𝑒𝑖𝑔ℎ𝑏𝑜𝑟𝑠 + 𝛽!𝐺𝐷𝑃𝑝𝑒𝑟𝐶𝑎𝑝𝑖𝑡𝑎!
+ 𝛽!𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑟𝑎𝑡𝑒𝐷𝑖𝑓𝑓𝑃𝑒𝑟𝑐!+ 𝛽!𝑇𝑟𝑎𝑑𝑒𝑃𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒!
+ 𝛽!𝐻𝑖𝑔ℎ𝐿𝑎𝑏𝑜𝑢𝑟𝐷𝑢𝑚𝑚𝑦!+ 𝛽!𝑇𝑒𝑟𝑡𝑖𝑎𝑟𝑦𝐷𝑢𝑚𝑚𝑦!+ 𝛽!𝐼𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛! + 𝛽!𝑅𝑒𝑎𝑙𝐺𝐷𝑃!+ 𝜀! (1.1)
The same models are used with First Differences (FD) as well. With FD, again the differences between the values of one year are reduced by the previous year. A difference with aforementioned model is the order of calculation and logarithm. With logExportDiff the logarithm is taken of the difference of 𝐸𝑥𝑝𝑜𝑟𝑡!− 𝐸𝑥𝑝𝑜𝑟𝑡!!!, whereas with FD first the difference is calculated and then the logarithm is taken. Two versions of FD are conducted; one of only the dependent variable, one of all variables.
An extra advantage of conducting FD is the clustering of groups. With FD the sample is divided in groups based on country. These clusters facilitate a better view into country-specific aspects as now virtually 20 regressions are conducted with 24 observations, instead of 1 with (an estimated) 480.
Three countries in the dataset (the United Kingdom, Ireland and Slovenia) have introduced a nationwide minimum wage during the time sample1. For these countries it is possible to compare the effect of a nationwide minimum wage with the same country. To enhance the strength of the results even more, two extra regular OLS-regressions are performed with a sample consisting of only those three countries.
1 In reality Czech Republic, Estonia and the Slovak Republic also introduced a nationwide minimum
4/Results
The impact of a nationwide minimum wage on the competitiveness of countries is investigated with the regression models outlined in the previous section. The different regressions and their results are discussed sequentially. As mentioned, the most important result to examine is the p-value of the variable.
OLS: logExport
The OLS estimates show that the presence of a Minimumwage has a significant impact on logExport at the 1% significance level. Table 1 shows that the p-value for Minimumwage is 0.000.
Because Exports is tested as a natural logarithm, the presence of a Minimumwage is associated with a 100% times the coefficient of Minimumwage change in Exports. Minimumwage increases Exports with 42.07% according to the OLS regression analysis. This effect is not in line with the expectation of an adverse effect.
The R-squared of regression (1.1) is 0,5615, which means that the chosen variables are a moderate estimate of the contributions to Exports, because a R-squared of 1 will completely describe the effect on Exports.
The other variables confirm expectations, which underlines the strength of the model. Positive coefficients are given for the variables; Neighbours, HighLabourDummy, TertiaryDummy and WorldEconomicPerformance, meaning Exports will increase with an increase in these variables. The coefficients of GDPperCapita, ExchangerateDiffPerc and Inflation are negative, meaning that if prosperity, prices, or currency increase, exports will decrease. This is in line with the theoretical expectation explained above. TradePercentage, our measure of the openness of the economy, displays a small, but significant negative relation. This is against our expectations. So, except for TradePercentage, the coefficients of the control variables behave as expected.
Notable is that the variables GDPperCapita and WorldEconomicPerformance are highly insignificant.
Dependent variable: logExport Estimation method OLS
Number of obs 451
OLS: logExportDiff
As explained under ‘method’, extra regressions are conducted to nuance and strengthen our interpretation of the relation between nationwide minimum wage legislation and export numbers. Export differences are calculated.
The number of observations is reduced a little with this regression.
The R-squared reduces, to a value of 0.4187. The Minimumwage variable is still significant at the 5%-level. Again, its effect on the dependent variable is positive, in contrast with expectations. The effect does shrink by 3% to an effect of 39,61% on Export difference.
The other variables behave identical to the first regression. TradePercentage is negative once more. GDPperCapita on the 5%-level and WorldEconomicPerformance on 1%-level are now significant in contrast to before. The TertiaryDummy now is not, and nor is the constant.
Variables Coef. Std. Err. t P>t (95%)
Minimumwage .4207261 .114203 3.68 0.000
Neighbours .2714158 .0234189 11.59 0.000
GDPperCapita -4.31e-06 7.46e-06 -0.58 0.564
ExchangeRateDiffPerc -.014294 .0015498 -9.22 0.000 TradePercentage -.0078157 .0011109 -7.04 0.000 HighLabourDummy 1.268259 .1556295 8.15 0.000 TertiaryDummy .3377172 .0045099 1.68 0.003 Inflation -.0977204 .0150231 -6.50 0.000 WorldEconomicPerf. .0079609 .028884 0.28 0.783 _cons 3.275185 .2300752 14.24 0.000
Dependent variable: logExportDiff Estimation method OLS
Number of obs 328
Full sample R-squared 0.4187
Variables Coef. Std. Err. t P>t (95%)
Minimumwage .3960816 .1731272 2.29 0.023
Neighbours .2410925 .0405553 5.94 0.000
OLS of First Differences of dependent variable only
In the first First Differences OLS-regression only the first differences (i.e. increase or decrease relative to previous year) of the dependent variable are used. As explained the difference with the logExportDiff is the order of calculating the difference value and taking the logarithm.
An extra function of FD is the clustering, or grouping, of the observations. The effect of this grouping is best illustrated with a fictive example. If only one country would have nationwide minimum wage legislation, and that country would also have very high (and increasing) export numbers, an apparent causality would be found between the two. Clustering is a method to reduce this effect by treating every country as a separate regression in itself, besides considering the entire sample.
This new method gives three different R-squared values. The ‘within’-value is the one we use, as is it best comparable to the regular R-squared. It does show a low value of 0.2498, meaning that the model is not able to explain the entire effect.
The effect of Minimumwage is insignificant, but negative. It is the first regression to show the expected, adverse effect of Minimumwage on logExport. However the effect is highly insignificant.
The control variables also behave different than in previous regressions. HighLabourDummy and TertiaryDummy both show negative effects, which is contrast with expectations. Again both variables are highly insignificant. Inflation shows a positive and significant relation, also contradicting expectations.
The Neighbours variable is omitted as the values of the variable remain constant throughout the time-frame. This makes the variable obsolete with FD.
Dependent Variable:
logExport Number of obs 450
ExchangeRateDiffPerc -.0166223 .0022123 -7.51 0.000 TradePercentage -.0044943 .0017012 -2.64 0.009 HighLabourDummy 1.489255 .2211311 6.73 0.000 TertiaryDummy .2625748 .1673434 1.57 0.118 Inflation -.0470904 .0171636 -2.74 0.006 WorldEconomicPerf. .3322425 .0640134 5.19 0.000 _cons -.2947241 .4105331 -0.72 0.473
Group Variable country Number of groups 24
R-squared within 0.2498 Obs per group min 13
between 0.0001 avg 18.8
overall 0.0001 max 22
Variables Coef. Std. Err. t P>t (95%)
Minimumwage -.0141674 .0303818 -0.47 0.645 Neighbours 0 (omitted) GDPperCapita -.0000103 3.07e-06 -3.37 0.003 ExchangeRateDiffPerc -.0060435 .0013421 -4.50 0.000 TradePercentage .0015269 .0005654 2.70 0.013 HighLabourDummy -.0029752 .0146435 -0.20 0.841 TertiaryDummy -.0129753 .0124606 -1.04 0.309 Inflation .0083999 .002428 3.46 0.002 WorldEconomicPerf. .0505736 .0037825 13.37 0.000 _cons 1.030024 .2503196 4.11 0.000 sigma_u 54.564.125 sigma_e .17165385 Rho .9990113
OLS of First Differences of all variables
Another First Differences regression is conducted in which the first differences of all variables are taken. It is the most nuanced regression method applied in this paper. Again the sample is clustered on country. The R-squared decreases further, to a value of only 0.2002.
The Minimumwage variable however shows a negative relation on a significance level of 1%. Both HighLabourDummy and TertiaryDummy are insignificant. And Inflation, again shows a remarkable positive relation with logExport. Although, only on a 10% significance level.
Dependent Variable:
logExport Number of obs 421
Group Variable country Number of groups 24
R-squared within 0.2002 Obs per group min 12
overall 0.1962 max 21
Variables Coef. Std. Err. t P>t (95%)
Minimumwage -.1251907 .0390075 -3.21 0.004 Neighbours 0 (omitted) GDPperCapita .0000352 .0000136 2.59 0.017 ExchangeRateDiffPerc -.0042331 .0005357 -7.90 0.000 TradePercentage .0045044 .0015309 2.94 0.007 HighLabourDummy .0255229 .0261778 0.97 0.340 TertiaryDummy -.0236041 .0159619 -1.48 0.153 Inflation .0097617 .0041481 2.35 0.028 WorldEconomicPerf. .0048134 .0023324 2.06 0.051 _cons .0509733 .0094333 5.40 0.000 sigma_u .03015731 sigma_e .1814376 Rho .02688408
OLS: logExport of 3 countries
In addition to the regressions with the entire sample, two other regular OLS-regressions are included using only the data on the three countries that introduced nationwide minimum wage legislation during the sample. The reduced samples contain only 54 and respectively 40 observations. This means the statistic value is low and might be marginal. The economic implication of these regressions is just the more interesting. For this reason, the regressions are included nonetheless.
The first regression uses logExport as dependent variable. The R-squared of this regression is substantially higher than previous regressions.
Minimumwage displays a negative effect at the 10% significance level. Neighbours, TradePercentage and ExchangeRateDiffPerc show effects contradicting expectations. The other control variables behave as expected.
Dependent variable: logExport Estimation method OLS
Number of obs 54
UK, Ireland & Slovenia R-squared 0.9532
OLS: logExportDiff of 3 countries
The final regression is an OLS-regression of Minimumwage on logExportDiff with a sample consisting of the aforementioned three countries. The regression does show a R-squared of 0.7413 and a negative relation of Minimumwage on a significance level of 10%.
However most of the variables are highly insignificant. Only exceptions are TradePercentage and WorldEconomicPerf.
Minimumwage -.405879 .2156447 -1.88 0.066 Neighbours -.2680161 .1036394 -2.59 0.013 GDPperCapita .0000948 .000018 5.27 0.000 ExchangeRateDiffPerc .0056906 .0049627 1.15 0.258 TradePercentage -.0151164 .001049 -14.41 0.000 HighLabourDummy .4525147 .1317507 3.43 0.001 TertiaryDummy .05638 .1458269 0.39 0.701 Inflation -.05918 .0220774 -2.68 0.010 WorldEconomicPerf. .0107619 .0257825 0.42 0.678 _cons 4.052.738 .5741752 7.06 0.000
Dependent variable: logExportDiff Estimation method OLS
Number of obs 40
UK, Ireland & Slovenia R-squared 0.7413
Variables Coef. Std. Err. t P>t (95%)
Minimumwage -.9783076 .5030601 -1.94 0.061 Neighbours -.2257152 .1036394 -0.86 0.399 GDPperCapita .0000325 .00005 0.65 0.521 ExchangeRateDiffPerc -.0206794 .0166578 -1.24 0.224 TradePercentage -.0153526 .0031077 -4.94 0.000 HighLabourDummy .6317984 .3405006 1.86 0.073 TertiaryDummy .3451494 .2868591 1.20 0.238 Inflation .1579279 .1220929 1.29 0.206 WorldEconomicPerf. .4379053 .1219172 3.59 0.001 _cons 1.119.421 1.455.754 0.77 0.448
5/Conclusion
The intention to introduce a nationwide minimum wage is always surrounded by discussion and debate. As is often the case, this discussion is fed by indistinctness. The most prominent debate is usually on the effects of the minimum wage on the labour market. Research on these effects however has increased knowledge and reduced emotion-driven argumentation. Another effect that is mentioned as a relevant consequence of the introduction of the minimum wage is the effect on competitiveness. On this topic research has provided less certainty.
This paper aimed at this topic. To research this effect six regressions were conducted. Exports and Export differences were chosen as indicators of competitiveness. If a country becomes more competitive, demand for goods from that country increase and so do exports. A sample consisting of the 24 European OECD-member states spanning over the period from 1990 to 2014 was constructed.
Results of the regressions were inconclusive. Initial OLS-regression displayed a positive and significant effect; the introduction or presence of a nationwide minimum wage caused exports to increase by 41%. However consecutive regressions on Export differences and with the use of First Differences, showed smaller and negative relations between minim wages and export numbers. This was in line with the expectations set put in the theoretical explorations.
Two extra OLS-regressions were conducted, considering only the countries that introduced nationwide minimum wage legislation during the time sample. Although the statistical value of these regressions might be low due to the small number of observations, its economic implications might be just the more interesting. These regressions reduce country-specific effects, i.e. the influence of cultural, demographic and political, but also economic, factors that differ between countries. Both regressions showed a negative effect on a 10% significance level.
Although these latter regressions suggest a negative relation between minimum wages and export performance, initial results were positive. The contradicting results in this paper make it hard to draw conclusions. If anything, the results make a direct relationship between nationwide minimum wage legislation and export performance less obvious.
6/ Further research
The ambiguous results stimulate further research in the relation between wage structures and international competitiveness. This paper aimed at reducing country-specific effects. However the means to do so were not optimal. More research might be put into comparative research between countries.
The data used in this paper are all coming from European countries. These match relatively well. The potential effect of a minimum wage in developing countries could not immediately be deduced from our results. Different or more diverse datasets could shed more insight.
In the literature review two possible effects are set put that could explain the negative effect of Minimumwage on exports. To explain the effects found in this paper, research could aim at these two mechanisms. How would minimum wage influence competitiveness? This could give insight into when the effect might be positive, and when it might be negative.
7/ Literature
Aaronson, D. (2001). Price Pass-Through and the Minimum Wage. The Review of
Economics and Statistics, vol. 83, no. 1, pp. 158-169
Aiginger, K. (2006). Competitiveness: From a Dangerous Obsession to a Welfare Creating Ability with Positive Externalities. Journal of Industry, Competition
and Trade, 6, pp. 161-177
Baskaya, Y.S. & Rubinstein, Y. (2012). Using Federal Minimum Wages to Identify the Impact of Minimum Wages on Employment and Earnings across the U.S. States. Presenting Paper at SOLE, 2012
BBC (2014). Germany approves first-ever national minimum wage. Retrieved from http://www.bbc.com/news/business-28140594
Boltho, A. (1996). The Assessment: International Competitiveness. Oxford Review of
Economic Policy, vol. 12, no. 3
Card, D. & Krueger, A.B. (1993). Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania. The American
Economic Review, 84/4
Card, D. & Krueger, A.B. (1995). Myth and Measurement: The New Economics of the Minimum Wage. Princeton University Press.
Carlin, W., Glyn, A. & Van Rheenen, J. (2001). Export Market Performance of OECD countries: An Empirical Examination of the Role of Cost
Competitiveness. The Economic Journal, vol. 111, no. 468, pp. 128-162 Cassidy, M. & O’Brien, D. (2005). Export Performance and Competitiveness of the
Irish Economy. Quarterly Bulletin, 3
Centraal Bureau voor Statistiek (2012). Germany is the main trading partner of 17 EU Countries. Retrieved from
http://www.cbs.nl/en-GB/menu/themas/dossiers/eu/publicaties/archief/2012/2012-3661-wm.htm
Cerra, V., Soikkeli, J. & Saxena, S.C. (2003). How Competitive is Irish
Manufactering? The Economic and Social Review, vol. 34, no. 2, pp. 173-193 Dolado, J., Kramarz, K., Machin, S., Manning, A., Margolis, D., Teulings, C., Saint
Paul, G. & Keen, M. (1996). The Economic Impact of Minimum Wages in Europe. Economic Policy, Vol. 11, No. 23 (Oct., 1996), pp. 317-372
Dolado, J.J., Felgueroso, F. & Jimeno, J.F. (2000). The Role of the Minimum Wage in the Welfare State: An Appraisal. Scheiz. Zeitschrift für Volkswirtschaft
Dornbusch, R., Fischer, S. & Samuelson, P.A. (1977). Comparative Advantage, Trade, and Payments in a Ricardian Model with a Continuum of Goods. The
American Economic Review, vol. 67, no. 5, pp. 823-839
Doucouliagos, H. & Stanley, T.D. (2008). Publication Selection Bias in Minimum Wage research? A Meta-Regression Analysis. Faculty of Business and Law
School Working Paper, 2008/14
Gärtner, M. (2009). Macroeconomics. Harlow: Pearson Education Limited. Honohan, P. & Walsh, B. (2002). Catching up with the leaders: The Irish Hare.
Brookings papers on Economic Activity, vol. 2002, no. 1, pp. 1-57
Fagerberg, J. (1988). International Competitiveness. The Economic Journal, vol. International Labour Organization (2013). Can minimum wages help rebalance
the economy? Retrieved from http://www.ilo.org/newyork/voices-at- work/WCMS_219658/lang--en/index.htm
Jacobi, L. & Kluve, J. (2006). Before and after the Hartz reforms: the performance of active labour market policy in Germany. IZA Discussion Papers, No. 2100 Javorcik, B.S. & Spatareanu, M. (2005). Do Foreign Investors Care about Labour
Market Regulations? Review of World Economics, vol. 141 (3), pp. 375-403 Krugman, P. (1994). Competitiveness: A Dangerous Obsession. Foreign Affairs,
vol. 73, no. 2, pp. 28-44
Krugman, P.R. (1996). Making Sense of the Competitiveness Debate. Oxford Review
of Economic Policy, vol. 12, no. 13, pp. 17-25
Lee, C., Schluter, G. & O’Roark, B. (2000). Minimum wages and food prices: an analysis of price pass-through effects. International Food and Agribusiness
Management Review, 3, pp. 111-128
Lemos, S. (2006). A Survey of the Effects of the Minimum Wage on Prices.
University of Leicester, Deparment of Economics Working paper, no. 06/9
Limão, N. & Venables, A.J. (2001). Infrastructure, Geographical Disadvantage, Transport Costs, and Trade. The World Bank Economic Review, vol. 15, no. 3, pp. 451-‐479
Méjean, I. & Patureau, L. (2010). Firms’ location decisions and minimum wages.
Regional Sciences and Urban Economics, 40, 45-49
NCC (2014). Ireland’s Competitiveness Scorecard 2014. Retrieved from
http://www.competitiveness.ie/media/ICS%202014%20ONLINE%20FINAL.pdfNear
vol. 37, no. 2, pp. 197-213
Nederlands-Duitse handelskamer (2014). Handel met Nederland. Retrieved from
http://www.dnhk.org/nl/niederlande/fakten/handel-met-nederland/
OECD (1998). Chapter 2: Making the most of the minimum: statutory minimum wages, employment and poverty. OECD Employment Outlook 1998 Politics.co.uk (2012). National Minimum Wage. Retrieved from
http://www.politics.co.uk/reference/national-minimum-wage
Sachs, J. (1980). Wages, Flexible Exchange Rates, and Macroeconomic Policy. The
Quarterly Journal of Economics, vol. 94, no. 4, pp. 731-747
Schmid, G. & Modrack, S. (2008). Employment dynamics in Germany: Lessons to be learned from the Hartz reforms. WZB Discussion Paper, No. SP I 2008 102
98, no. 391, pp. 355-374
The Economist (2004). How to pep up Germany’s economy? Retrieved from http://www.economist.com/node/2647306
Vishenvskaya, N. (2014). The Labour Market Reform in Germany: a Special Case or an Example to Follow? Ekonomicheskiy zhurnal VShE, vol. 18, no 2, pp. 249-284
World Economic Forum (2015). The Global Competitiveness Report 2014-2015. Retrieved from http://www.weforum.org/reports/global-
8/ Appendices
Appendix 1: introduction years of minimum wage legislation
Country Year of Introduction
Austria None Belgium 1975 Czech Republic 1991 Denmark None Estonia 1991 Finland None France 1950 Greece 1953 Germany 2015 Iceland None Ireland 2000 Italy None Luxembourg 1944 Netherlands 1968 Norway None Poland 1990 Portugal 1974 Slovak Republic 1991 Slovenia 1995 Spain 1963 Sweden None Switzerland None United Kingdom 1999
Appendix 2: variable information
Variable Measure Mean Std. Dev. Obs.
Dependent variables
logExport Billion US$ 4.521256 1.396.597 316 logExportDiff Billion US$ 2.246986 150.132 236
Independent variable
Minimumwage Dummy variable .5766667 .4944995 600
Control Variables
Neighbours - 5.473542 8.577.817 600
GDPperCapita GDP per head of
population in US$ 27980.69 11016.98 566
ExchangerateDiffPerc % change of average exchange rate per year
in US$
189.715 8.950.114 589
TradePercentage Import + Export as %
of GDP 9.584.503 5.169.721
565
HighLabourDummy GDP per hour worked
in US$ - 1 if over 35 39.4539 2.022.376 548
TertiaryDummy % of labour force with tertiary education – 1
if over 25
25.28463
25.28463
468
Inflation Annual growth rate in
% 8.292257 3.523.836 587
WorldEconomicPerf. % change of aggregate
GDP of the world 3.596 1.268.016 600
