The effect of minimum wage on unemployment in European countries from 1999-2015

(1)

The effect of minimum wage on unemployment in

European countries from 1999-2015

Amsterdam Business School

Bachelor Thesis

Name

Sheyma Mamakheel

Student number

10372806

Program

Economics & Business

Specialization

Economics & Finance

Supervisor

N.Ciurila

(2)

[Typ hier] 2

Statement of Originality

This document is written by Sheyma Mamakheel, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

[Typ hier] 3

1 Introduction

In October 2016, the minimum wage in the Netherlands has increased by 2.1 percent compared to the minimum wage in 2015. This is the highest increase of the minimum wage in the past 7 years. The fact that the CBS increased the minimum wage is due to the improved economy in the labormarket. The minimum wage has been introduced in many countries all over the world for many years. One of the main reasons for the introductionof the minimum wage was to improve the living conditions of the low paid workers. Moreover, another important reason was to lower the market power of the employers (Dolado et

al,1996, p.334). Although the minimum wage has been introduced many years ago, there are still discussions of its effects. Therefore, there are several studies about the effects of the minimum wage on unemployment, employment and wages. Many studies are based on the effects of the low-skilled workers or teenagers since the minimum wage generally affects them the most because they are most probable to be paid at or near the minimum wage. Some studies are convinced of the standard economic theory that an increase of the minimum wage will lower the employment of low-wage workers (Card & Krueger, 1995, p.238). Brown et al (1983, p-3-31) came to the same conclusion. The research of Brown et al was based on the time-series data of the United States and the negative effects on the employment rate were already expected. However, there are also scientists who are convinced of positive effects of the minimum wage on employment. For example, Neumark (2015) concluded that it is suitable to counterbalance the cost of potential job losses from a higher minimum wage against the increased wage benefits for the other workers.

Not only scientists but also policymakers are debating about the influence of the minimum wage on unemployment, employment and wages.

Since scientists constantly conduct new research on the effects of the minimum wage on unemployment, which means new findings are constantly published, this subject remains scientifically relevant. Because of the many studies of the effect of the minimum wage on youth unemployment and low-skilled workers, this thesis will analyze the effect of the minimum wage on the total unemployment. Moreover, while Brown et al (1982) found a significant effect of the minimum wage on the youth employment, they did not find a significant effect on the total unemployment.

This thesis will research whether this effect on total unemployment does exist. To do the research a panel-data analysis of 14 European countries from 1999-2015 will be used, since not every country has a minimum wage. Moreover, this thesis focuses on European countries, because there are fewer studies that are based on European countries compared

(5)

[Typ hier] 5

to studies that are based on the United States, for example.

In the first section of this thesis, the available literature on the subject is reviewed and the standard econometric models are explained. The second section lays out the

econometric analysis with information on the model, hypotheses, data and results. The third section consists of a discussion of the results, the limitations, and suggestions for further research are made. Finally, the conclusion is provided.

2 Literature Review

A vast amount of literature has been written on the subject of minimum wages and its relationship to employment and unemployment. As mentioned before, no concrete answer has been found. In this section the theoretical insights and the economic models will be introduced.

2.1 Theoretical insights

2.1.1 The standard model of competitive labor markets

As mentionedbefore, the standard model of competitive labor markets suggests that a higher minimum wage will increase the unemployment among low-skilled workers (Neumark, 2015). Thus, the simplest scenario assumes that product and labor markets are competitive, which means that firms take prices as given. Additionally, there is only one type of labor and the output is produced with this labor and capital (Neumark & Wascher, 2008, p.39). There are two reasons that a minimum wage that is set higher than the competitive equilibrium wage will reduce the employment, since the low-skilled workers are now more expensive because of the ‘binding’ minimum wage. Therefore, the first reason is that employers will substitute away from the low-skilled labor towards equipment or other capital. This is called the substitution effect. Moreover, Machin & Manning (1996, p.668) state that firms chose to take their labor to countries with lower wages. Also as mentioned earlier, Neumark(2015) suggests that there is a labor-labor” substitution, which generally is suitable if scientists do not only focus on the cost potential job losses but also on the increased wage benefits of the other employees. A “labor-labor” substitution means that employers will hire more high-skilled workers and fewer low-high-skilled workers if there is an increase of the minimum wage while, from a policy perspective, fewer jobs for the low-skilled workers are the most important, because the minimum wage is intended to help them.

(6)

[Typ hier] 6

product and labor demand (Neumark, 2015). The reason behind these effects are also explainedin Pindyck & Rubinfield’s paper (2009). When the minimum wage is selected at a level above the equilibrium market wage, then this will lead to a difference in supply and demand as stated in figure 1 below. This in turn lead to unemployment. Because an increase in the wages will increase the production costs, which in turn will cause higher prices and this will then lower demand. Therefore, there will be less production and less workers will be needed. Thus, there will be less demand for labor.

Figure 1: The Standard model of competitive labor markets

* Introducing a minimum wage (Wmin) leads to a lower Labour Demand (Qd)

In some labor markets, however,the standard economic model is not supported by empirical work. In these labor markets employers have some degree of monopsony power. This is called the “monopsony model”, where the minimum wages not only allow to have a positive effect on the unemployment but also an ambiguous or negative impact (Dickens et al, 1999, p.1)

(7)

[Typ hier] 7

2.1.2 The monopsony model

In labor markets with a “monopsony” model, worker mobility is limited. There is a single buyer of labor and there are several sellers of labor. Therefore, this single buyer of labor is free in setting wages to a certain limit. The effect of a higher minimum wage will then become ambiguous (Neumark, 2015). In this model labor will be hired until marginal cost and demand are equal. Since the individual employers have more power in this market, the wages will also be lower (Dolado et al, 1996, p.329). However, there are also scientists who are critical of this model because they are convinced that this model is not realistic. Neumark concluded that evidence in literature is more important than theoretical assumptions (Neumark, 2015).

Figure 2 : Monopsony model

(8)

[Typ hier] 8

2.2 Empirical studies

In this section, the empirical findings from countries all over the world will be discussed, since there are many studies with different results.

Over the years, there have been many studies about the influence of minimum wage on unemployment. Several of these studies conclude that there is a higher unemployment rate among low-skilled workers and teens as a result of a higher minimum wage.

For example, as Neumark(2007) mentioned in his paper based on several studies, there is evidence that the number of jobs are reduced nationally by about 100,000-200,000 as a result of the current minimum wages in the US, compared with the period before the

Recession. Neumark and Wascher (2007) also concluded that 85% of newer minimum wage studies found the same results of the job loss effects on low-skilled workers.

Card 1992; Katz and Krueger 1992; Card, Katz and Krueger 1994; Card and Krueger 1994, and Machin and Manning 1994 have all found that a minimum wage can either lead to positive effects on employment or not have an effect at all.

Deere et al (1995) and Neumark and Wascher(1995), however, found results that contradict this view.

The study of Card and Krueger (1994) is one of the most important papers in the field of the minimum wage. Card and Krueger researched the effect of a minimum wage on the labor market. Their findings stated that the minimum wage in New Jersey had increased, while the minimum wage in Pennsylvania had remained the same. It is, however, important to note that they only investigated the fast-food industry, by comparing the labor growth of the two countries. Instead of a negative effect on employment due to increased minimum wage, the employment was increased. Their findings are explained by the monopsony theory.

Neumark and Wascher (2000) investigated the same states as Card and Krueger did, but they used payroll data instead of surveys. They found a significant decrease in

employment with the increased minimum wage in New Jersey. In their paper, a panel data analysis was used instead of a time-series analysis. One of the main advantages of a panel data analysis is that a panel data analysis takes different minimum wages between states into account, but this can also lead to errors in estimation in time-series analysis.

Additionally, Neumark and Wascher (2006) mentioned that it might take some time for the minimum wage to have an effect on the employment. Therefore, the results between studies might differ.

(9)

[Typ hier] 9

Brown et al (1982) have written one of the oldest papers in the field of the minimum wage. They investigated the effects of the minimum wage on employment and

unemployment and found that a 10 percent increase of the minimum wage led to a 1-3 percent decrease of the employment of teenagers. The analysis was based on the minimum wage of the employment of teenagers of 24 separate studies from 1970-1980. Moreover, Brown et al (1982) also analyzed the other age groups, but found smaller negative effects compared to the employment of teenagers. They stated that the smaller effects of the young adults might have to do with the fact that young adults were no longer competing with teenagers, because both groups were more expensive but the low-skilled workers could be fired during that time. Thus, since young adults are generally more experienced, there is a possibility that teenagers could be fired. Pereira (2003) also found this substitution effect in her research. Neumark and Wascher (2004) did a cross-sectional analysis and also

concluded the negative effect on youth employment. They came to the conclusion that higher regulated markets have had smaller effects on unemployment than lower regulated markets. However, Lang and Kang (1998) found that the increase of the employment of

teenagers is greater than the decrease of adults and therefore conclude that there might be positive effects on the minimum wage on employment.

Clemens and Wither (2014) investigated the effects of the increased minimum wage on the employment of low-skilled workers. The estimated effect on employment was

negative. Besides the effect on employment, they also found negative effects on income and income growth. The 30 percent increase of the minimum wage from 2006-2012, for example, led to a 0.7 percentage point decrease of the national employment-to population ratio. Michel and Cahuc (1996) elaborate on a model of the minimum wage, unemployment and growth. They suggest that minimum wage legislation may have positive effects on the growth rate by inducing more human capital accumulation. This means that a lower demand for unskilled labor induced by a minimum wage may create an incentive for workers to accumulate human capital. Thus, an increase in the minimum wage leads to an increase of employment from the point of view of this paper.

A few empirical analyses on the European countries do exist. Dolado et al (1996), for example, investigated the effects of minimum wages in 4 European countries. They found mixed results in every country. Dolado et al (1996) came to the conclusion that the effect on total employment and the effect on youth employment are not similar to each other. For example, in Spain the minimum wage effect was negative on youth employment, but positive on adult employment as also concluded from the paper of Brown et al (1982) and Pereira (2003). They conclude that the minimum wage does not cause the unemployment to increase in Europe.

(10)

[Typ hier] 10

Moreover, this conclusion is also made by Dube et al( 2010). Especially based on restaurants, they conclude that wages and employment should not be affected by a change in the minimum wage. Dube et al(2010) conclude that there are three important questions relating to studies about the minimum wage. Namely, whether the different empirical findings result from a lack of adequate controls for unobserved heterogeneity which is in case in most panel regressions , the lack of sufficient lag time or the overstatement precision of estimates especially in the local case studies.

3. Empirical analysis

In this section, the econometric analysis of the research will be introduced. First, the chosen regression model will be discussed. Second, there will be an overview of the variables included in the model with additional information and, after this, the econometric model will be explained.

3.1 Methodology and Data

The econometric analysis that will be conducted in this thesis is based on the effect of the minimum wage on the unemployment rate.

For this analysis, a strongly balanced panel regression with a random effects model will be used, in order to fully make use of the available information. Theanalysis of this thesis is based on the paper of Neumark and Wascher (2004), but whereas Neumark and Wascher used a pooled OLS-regression, in this thesis a panel regression with a random effects model will be used.

In general, there are no differences between these two regressions but a random effects model is preferred to a pooled OLS-regression, because a random effects model is more efficient (Woolridge, 2009, p.489). The significant outcome of a Breusch-Pagan Lagrange Multiplier test shows that a random effects model is preferred over an OLS-regression. Secondly, the insignificant outcome of a Hausman-test proves that a random effects model is preferred over a fixed effects model. The differences between the last two models are lay in the relation between the unobserved effect αiand each explanatory variable xitj. Namely, in

the random effects model, the unobserved effect αiis uncorrelated with each explanatory

variable xitj, which is not the case with a fixed effects model (Woolridge, 2009, p.493).

Although this research is a micro panel, the serial correlation is tested and has a

significant outcome, which means that there is autocorrelation and this must be controlled in the panel-regression. Serial correlation will cause the standard errors of the coefficients to be smaller than they are in reality. Finally, robust standard errors will be used due to the

(11)

[Typ hier] 11

are reported in table 4, which can be found the results section.

3.1.1 Countries

Although a panel regression takes into account the differences betweencountries – for example, the differences in minimum wage policies, – only European countries are analyzed. Moreover, the countries in the sample were chosen based on the availability of data. This analysis consists of the effect of the minimum wage of 14 European countries between 1999-2015. The reason that this period of time is researched in this analysis is because of the fact that the United Kingdom introduced the minimum wage The 14 European countries are shown below in table 1.

Table 1

Belgium Greece Netherlands Slovenia

Czech Republic Hungary Poland United Kingdom France Ireland Portugal

Germany Luxembourg Spain

The variables consist of the independent variable, the main dependent variable and control variables, and are discussed in the sections below.

3.1.2 The dependent variable and independent variable

The data for this analysis is collected from the website of the OECD(2016) and

Eurostat(2016), and every variable is calculated yearly. The dependent variable is the unemployment rate in percentage of the active population, the same variable as the variable that Neumark and Wascher (2004) used in their paper. The independent variable is the ratio of the nominal minimum wage relative to the annual wage of the workers, since this is the standard indicator used in the literature on the minimum wage, which is also used in the paper of Neumark and Wascher (2004). The use of the definition of the ratio as the minimum wage, as mentioned in the last sentence, is intended to capture the effects of the minimum wage on the wage distribution (Neumark & Wascher, 2004, p. 2260.Moreover, Dolado et al (1996) also divided the nominal minimum wage by the annual wage of workers for the same reason.

(12)

[Typ hier] 12

3.1.3 The control variables

The control variables are included in the model to diminish omitted variable bias. The control variables that are used are the interest rate (long-term interest rate for each country), the inflation rate (in percentage of the growth on the same period of the previous year), and the GDPgrowth rate (GDP volume in percentage of the growth on the same period of the previous year).

Choudry Marelli and Signorelli (2012) added these macroeconomic variables in his research, which might have a significant effect on the unemployment rate.

The other control variables are the Strictness Employment Protection (SEP) and the Trade Union Density (TUD). Strictness Employment Protection consists of procedures and costs involved in dismissing or hiring employees, individuals or groups of workers (OECD Database, 2016). This variable is also included in the paper of Neumark and Wascher (2004), since the degree of hiring or firing employees might affect the employment. Neumark and Wascher (2004) also included the trade of density in their research. Trade Union of Density is the ratio of the wage and salary earners that are members of trade union relative to the total number of wage and salary earners. A country with a higher Trade Union Density will have a higher bargaining power, which could lead to a lower unemployment since it is harder to fire employees in these countries.

3.1.4 Macroeconomic view of the control variables

Okun’s law, which is explained in the Macroeconomics book of Mankiw (2012, p.277), states that there is a negative correlation between the gross domestic product growth and the unemployment rate. The reason behind this assumption is that the unemployed workers do not help to produce goods and services and that an increase in the unemployment rate thus automatically means a decrease in real GDP. Rabins and Stevens (2002) also researched the impact of the unemployment rate and the real GDP, and came to the same conclusion about Okun’s law.

The curve is explained in the book of (Mankiw, 2002, p.406). The Phillips-curve is based on a short-time period evidence, which states that there is a negative correlation between inflation and unemployment as well. The theory behind this is that if there is more demand than supply in the market of goods, which in general means that the economy is improved, this will lead to an inflation, which will in turn lead to more production. This increase in production means that there is a higher demand for labor and less

unemployment.

The last macroeconomic variable is the interest rate. The theory behind this is that if there is a lower interest rate, households will lend more money to consume rather than save

(13)

[Typ hier] 13

money in the bank. Moreover, at a lower interest rate, investors will lend more. From both sides there might be more productions because of the increased investment and increased consumption, which means that the demand for labor would also increase. Theoretically, there is a positive correlation between interest rate and the unemployment rate expected.

3.1.5 Econometric model

There are 6 models used for the analysis in this thesis, which is reported below. The first model consists of the direct effect of the minimum wage ratio on the unemployment rate. But the unemployment rate does not only depend on the minimum wage ratio. Therefore, there are other variables added to the model, which are supposed to have an effect on the

unemployment rate. However, the unemployment rate also depends on other variables then the ones that are used in these models. Not all of these variables are included in this thesis, since there are a lot of variables that might affect the unemployment.

The variables that are correlated with the dependent variable (unemployment rate) are added one by one to analyze the significance of the variables. First, the interest rate and the GDPgrowth are added to the models to observe what the effects of macroeconomic variables could be on the unemployment rate. Secondly, the Trade Union Density is also added in the next model, which might have an effect on the unemployment. Lastly, the other macroeconomic variable (inflation rate) and the Strictness Employment Protection are added to the last model to investigate if there is a significant effect on the total unemployment.

Table 2

*This table shows the correlation of the variables with the dependent variable (unemployment rate).

Finally, the independent variable and the control variables are tested on perfect

multicollinearity. The independent variable and every control variable are tested on their correlation with each other. There is no perfect multicollinearity as is stated below in table 3.

(14)

[Typ hier] 14

Table 3

* This table shows the correlation between the explanatory variables

The models:

(1) _{UEit =}β0 + β1MWit αi + uit

(2) UEit = β0 + β1MWit + β2INTERESTit + αi + uit

(3) UEit = β0 + β1MWit + β2INTERESTit + β3GDPGROWHTHit + αi + uit

(4) UEit = β0 + β1MWit + β2INTERESTit + β3GDPGROWHTHit + β4TUDit + αi + uit

(5) UEit = β0 + β1MWit + β2INTERESTit + β3GDPGROWHTHit + β4TUDit +

β5NFLATIONit + αi + uit

(6) UEit = β0 + β1MWit + β2INTERESTit + β3GDPGROWHTHit + β4TUDit +

β5NFLATIONit + β6SEPit + αi + uit

β0 is the constant and uit is the error term in the model. The details of the variables are as

followed:

UE Unemployment rate

MW Minimum wage ratio

GDPGROWTH GDPgrowth

INFLATION Inflation rate

INTEREST Interest rate

TUD Trade Union Density

(15)

[Typ hier] 15

The first model consists of the direct effect of the minimum wage ratio on the unemployment rate , without control variables to compare it after the added control variables.

4 Hypotheses

4.1 Effect of the minimum wage on unemployment

From the view of the standard economic theory model – as stated in the literature review section – the minimum wage will have a significant positive effect on the unemployment in a country.

However, the minimum wage might not have a significant effect on the total unemployment, because the studies that have found significant effects were specifically based on youth employment or on low-skilled workers in the United States

4.2 Effect of the macroeconomic variables on unemployment

4.2.1 GDPgrowth rate

When Okun’s law is taken into account, the GDPgrowth is expected to be negative in the empirical part of this thesis.

4.2.2 Inflation rate

From the view of the Phillips-curve, the inflation rate is also expected to be negative in the research of this thesisl

4.2.3 Interest rate

The classic macroeconomic theory states that an increase in the interest rate will lead to a lower investment, since lending becomes more expensive. Therefore, fewer workers are needed and this will affect the unemployment rate in a negative way. Thus, the interest rate is expected to be positive in this analysis.

4.3 Effect of the other control variables on unemployment

4.3.1 Trade Union Density

Since it will be harder for employers to fire people, the degree of hiring or firing people is expected to be negative in this analysis.

(16)

[Typ hier] 16

4.3.2 Strictness Employment Protection

Since it will be more difficult to fire employees due to an increase in the bargaining power, this variable is also expected to be negative in the analysis.

(17)

[Typ hier] 17

5 Results

Table 4 shows the results of the panel random effects model.

5.1 Panel-regression: Random effects model

Table 4: Panel Data Random Effects Model

Dependent variable: unemployment rate

Independent variables (1) (2) (3) (4) (5) (6) CONSTANT 0.904*** 0.676*** 0.823*** 1.126*** 1.060*** 1.267*** (0.09136) (0.150) (0.183) (0.186) (0.176) (0.332) MW -0.076** -0.050* -0.077** -0.096** -0.063* -0.065** (0.033) (0.030) (0.033) (0.039) (0.0346) (0.032) INTEREST 0.048 0.033 0.038 0.058** 0.055** (0.030) (0.029) (0.030) (0.025) (0.025) GDPGROWTH -0.031** -0.028* -0.020 -0.021 (0.014) (0.015) (0.014) (0.014) TUD -0.013 -0.007 -0.010 (0.008) (0.007) (0.008) INFLATION -0.096*** -0.092*** (0.028) (0.026) SEP -0.057 (0.067) Observations 238 229 229 229 229 229 Countries 14 14 14 14 14 14 R-squared 0.0091 0.1352 0.1742 0.2008 0.3317 0.3181 Breusch Pagan/ Cook Weisberg Test 0.000 Woolridge test for autocorrelation 0.000

(18)

[Typ hier] 18

Robust standard errors are in parentheses below the estimated coefficients. * = significant at 0.10 level, ** = significant at 0.05 level and *** = significant at 0.01 level.

N is decreased from 238 to 229 because of the lack of the minimum wages in Germany until 2015, since the minimum wage was introduced in that year.

6. Discussion

In this section, the results from the panel data regression effects model will be discussed.

6.1 The dependent variable

The results from the regressions of this thesis show that the minimum wage ratio has a significant negative effect on the total unemployment rate, which was also found by the empirical study of Card & Krueger (1995) by comparing the two states in the United States. In model (1), (3), (4) and (6) the significant effects are higher than in model (2) and (5). It seems that when a macroeconomic variable is added to the model, the minimum wage has a lower significant effect on the unemployment. These results may be affected by several conditions, since this is a small research with limitations and the conclusion of this research cannot be generalized over the world. However, it is possible that the unemployment rate and the minimum wage ratio depends on these added variables and therefore lower the effect of the minimum wage ratio on the unemployment rate.

In all of the models, the minimum wage has a negative effect on the unemployment, which means that an increase in minimum wage will lead to a decrease in unemployment. This result was not expected and could only be explained by the monopsony model, not by the standard model of competitive labor markets.

6.2 The control variables

The interest rate has a positive effect on unemployment, which is in line with the

macroeconomic view. This effect was expected. The interest rate only has a significant effect on the unemployment rate in model (5) and (6). The difference with the other models (2), (3), (4) is that the inflation rate was included in the last two models. From the macroeconomic view a higher interest rate will lead to produce and consume less, because firms and households will lend less. Therefore, less production is needed and thus the unemployment will increase.

The GDPgrowth rate has a negative effect on unemployment, which is correct from the view of Okun’s law. The reason for this effect is explained under the empirical analysis section. The GDPgrowth rate is significant in model (3), while the significant effect on

(19)

[Typ hier] 19

unemployment rate is decreasing in model (4) when the Trade Union Density is added. There is no paper found which could explain this result of the two variables. However, after adding more control variables, the GDPgrowth is not significant anymore. Therefore, no hard conclusions can be drawn from these results.

The inflation rate shows a significant effect on unemployment in both models when added. The effect of the inflation rate on the unemployment rate is negative which is in line with the Phillips-curve, which is also explained in the empirical analysis section.

The Trade Union Density does not show a significant effect on unemployment and, therefore, no conclusions can be drawn from these results. However, the effects are negative, which was expected. Therefore, no conclusions can be drawn from this result. Maybe more control variables should be added to the model or trade union density might have an effect on the employment but not automatically on the unemployment.

The Strictness Employment Protection also does not show a significant effect on unemployment. The effects were also expected to be negative, and so this analysis has proved my hypothesis.

The last model includes all of the variables that are introduced in this paper. The minimum wage is still significant. This means that the overall conclusion is that the minimum wage has a significant effect on the unemployment.

This effect is negative, which is not in line with the hypothesis of this paper. Many studies found a significant positive effect on young unemployment. Moreover, the standard economic competitive labor markets model also supports the positive effect on the unemployment. However, in the monopsony model the positive effect on the unemployment is possible. Card and Krueger (1994) also used this model to discuss their results.

The interest rate and the inflation rate also have a significant effect on unemployment. The effect of the interest rate is positive and the effect of the inflation rate is negative, which is in line with the hypothesis of this paper. The theoretical background of these expectations are explained under the hypothesis section.

(20)

[Typ hier] 20

7 Conclusion

In this section, the theoretical insights and the empirical study that has been conducted in this thesis will be summarized and concluded.

There are several studies about the effect of the minimum wage on the employment and unemployment. Since scientists still do not agree on one point of view (there is no clear statistical evidence for either a positive or negative effect), this subject remains scientifically relevant. Moreover, since there are fewer studies about the effect of the minimum wage on the total unemployment rate rather than on the unemployment rate in low-skilled worker groups or teenagers, this thesis focused on the effect of the minimum wage on the total unemployment rate.

This thesis has undertaken an analysis of unemployment in 14 European countries from 1999-2015. It uses a strongly balanced panel regression with a random effects model. The independent variable in this research is the minimum wage ratio, the dependent variable is the unemployment rate and the control variables are added to diminish omitted variable bias. The results show that the increase in the minimum wage has a significant effect on the unemployment rate. But this effect is negative, which was not expected, since the standard economic competitive labor markets model states that the effect is positive. This result is in line with the monopsony model, although there many studies express doubts about this model compared to the labor markets in reality (Dolado et al, 1996). However, Card and Krueger (1994) also mentioned the monopsony model to discuss the results of their

research. They concluded that an increase in the minimum wage might lead to an increase in the employment (sample New Jersey).

The macroeconomic variables are in line with the theory that is used to explain the hypothesis of this paper. As well as the interest rate and the inflation rate that both have statistically significant effects on unemployment, while the GDPgrowth rate is only significant in model (3) and not significant when more variables are added to the model. The

macroeconomic theory (Mankiw, 2012) states that a decrease in the interest rate will lead to a higher lending from both the perspectives of firms and households and this, in turn, will lead to more production and consumption. To produce the goods, more workers are needed. From the view of the Phillips-curve, there is inflation due to the improved economy.

Therefore, more people will be hired to produce goods and the unemployment will decrease (Mankiw, 2002, p.406)

The Trade Union Density and the Strictness Employment Protection showed insignificant effects on the unemployment, therefore no conclusions can be drawn from these results.

(21)

[Typ hier] 21

Because the complete model of this thesis is not based on a paper, the results also cannot be linked to other empirical studies.

8 Limitations and further research

This section will mention the shortcomings of the research made in this thesis and it will also provide recommendations for further research

This research only consists of 14 European countries and investigates the period between 1999-2015. It is possible to add more countries and increase the number of years that are researched in further research, which might lead to different results. For a panel-data analysis, especially, the number of years is important to draw a clear conclusion.

The data that was available to do the research was limited. This could possibly have led to a different outcome. In further research more control variables could be added to avoid omitted variable bias. Moreover, a few control variables showed no significant effects at all (Trade Union Density and Strictness Protection Employment). Therefore no conclusions were made based on these variables. Furthermore, the quality of the independent variable could also be improved, since the minimum wage ratio of full time workers is used. It could be useful, for example, to include the minimum wage ratio of part time workers.

Germany has introduced the minimum wage in 2015. However, this country is included in the regression. This might lead to an inordinate effect on the overall coefficient estimate, because the implied minimum wage change of Germany would relatively be very large compared to to the other countries’ minimum wage changes. Moreover, for Germany, zero values have been used until 2015 to observe the effects of introducing the minimum wage, but this might lead to a misspecification of the model in this research, since there is a floor well above zero below which wages would not fall (Neumark and Wascher, 2004, p. 229).

This thesis focused on the total unemployment and the effect of the minimum wage was negative. Further research could also investigate the unemployment of teenagers compared to the total unemployment, since the minimum wage might affect the

unemployment of teenagers more. Namely, Brown et al (1982) found studies that proved the significant effect of the minimum wage on the unemployment of teenagers in the United States.

As many studies did not use lagged variables, this thesis also did not make use of lagged variables. Neumark and Wascher(2006 , p.105), however, suggest that using lagged variables could make sense in finding negative effects of the minimum wage on the

(22)

[Typ hier] 22

findings will differ when there will be sufficient lag time in the case studies. This could also be the reason of the unexpected significant negative effect of the minimum wage on the

unemployment in this paper. It could thus prove valuable for further research to add lagged variables to their model.

(23)

[Typ hier] 23

References

Brown, C., C. Gilroy, & Kohen, A. (1983). “ The Effect of the minimum wage on Employment and Unemployment.’ Journal of Human Resources 18 (1) , pp. 3-31.

Cahuc, P., Michel, P. (1996). “ Minimum wage unemployment and growth.” The economic

Review 40 ; pp.1463-1482.

Card, D. (1992).” Using regional variation in wages to measure the effects of the federal minimum wage”. Industrial & labor relations review 48(1), pp. 7-37

Card, D., & Krueger A. B. (1995). “ Time-series Minimum Wage Studies: A meta-analysis.”

The American Economic Review 85(2)pp. 238-243.

Card, D., & Krueger A. B. (1995).” Minimum wages and employment: A case study of the fast food industry in New Jersey and Pennysylvania. The American Economic Review, 84(4) , pp. 772-793.

CBS ( 2016). Retrieved from : https://www.cbs.nl/nl-nl/nieuws/2016/40/grootste-cao-loonstijging-sinds-2009

Choudry, M., Marelli, E., & Signorelli, M. (2012). “Youth and total unemployment rate: the impact of policies and institutions.” Rivista Internatzionale di Scienze Sociale.

Clemens, J., & Wither, M. 2014. “ The minimum wage and the great recession: evindence of effects on the employment and income trajectories of low-skilled workers.” NBER Paper 20724.

Deere, D., Murphy, K.M., & Welch, F. “Employment and the 1990-1991 Minimum Wage-Hike.” The American Review, 85(2) , pp. 232 – 237.

Dickens,R., Machin, S., & Manning, A. (1999). “The effects of minimum wages on

employment: Theory and evidence from Britain.” Journal of Labor Economics, 17(1), pp. 1-22.

Dolado, J., Kramarz, F., Machin, S., Manning, A., Margolis, D., & Teulings, C. (1996). “The economic impact of minimum wages in Europe.” Economic policy, 11(23), pp. 319-372.

(24)

[Typ hier] 24

Dube, A., Lester, T.W., & Reich, M. ( 2010) “ Minimum wage effects across state borders: estimates using contiguous counties”. The Review of economics and statistics. 92(4), pp. 945-964.

Katz, L. F., & Krueger, A.B (1992). “ The effect of the minimum wage on the fast-food industry”. Industrial and Labor Relations Review. 46 (1) , pp. 6-21.

Lang, K., & Kahn, S. (1998). “The effect of minimum-wage laws on the distribution of employment: theory and evidence.” Journal of Public Economics, 69(1),pp. 67-82.

Mankiw, N.G. (2012)” Macroeconomics.” Eight edition.

Neumark, D. (2015) “The effects of minimum wages on employment” Economic letter

Neumark, D., & Wascher, W. (2004). “Minimum wages, labor market institutions, and youth employment: a cross-national analysis.” Industrial & labor relations review, 57(2), pp 223-248.

Neumark, D., & Wascher, W.(2006) “Minimum wages and employment: A review of evidence from the new minimum wage research.” National Bureau of Economic Research.

Neumark, David, Wascher, William L. (2008). “Minimum wages.” Massachusetts Institute of Technology, London.

Neumark, D., & Wascher, W. (2000). “Minimum wages and employment: A case study of the fast-food industry in New Jersey and Pennsylvania : Comment.” The American Economic

Review, 90(5), pp. 1362-1396.

Neumark, D., & Wascher, W. 2007. “ Minimum Wages and Employment.” Foundations and

Trends in Microeconomics 3(1-2), pp. 1-182.

OECD(2016) Obtained via: http://stats.oecd.org/

Pereira, S.C. (2003). The impact of minimum wages on youth employment in Portugal.

European Economic Review, 47(2), pp. 229-244.

Pindyck, R., & Rubinfield, D. (2009). Microeconomics. Pearson: London

(25)

[Typ hier] 25

The effect of minimum wage on unemployment in European countries from 1999-2015