A Comparison Of Replacement Rate Between Different Age, Gender, Region, Household Composition And Level Of Education

This paper analyses financial incentive across different age, gender, region, household composition and level of education. The most important contribution is to provide an empirical test of whether replacement rates vary across different age, gender, region, household composition and level of education. This paper using correlation supported the strong positive relationship between replacement rates and unemployment rates, indicates higher replacement rates of one group of individuals with have higher unemployment rates (lower financial incentive to work). In addition, this paper through simulation techniquences to calculate the counterfactual income to estimate the replacement rates. The finding support hypotheses on the financial incentive vary across different sub groups.

 

Key Words: Replacement rates;incentive; level of education; age; family; gender; region

 

 

1. Introduction

From the middle of the 1990s to 2007 the growing economic environment in Ireland resulted in the Irish government reducing income tax and rising social welfare at the same time. This paper empirically analyse the impact of social benefit on different type of family incentive to work. The Irish government needs to design tax threshold and welfare carefully to balance cost and tax (IMF, 2014) after the fiscal crisis in 2008. Since 2008, Ireland has experienced unemployment level up to 14.8% (SUR, July 2012) and from early 1980s to mid-1990s high unemployment rates between 13% to 17%. One of the factors which result in this weak employment rate may be tax or social welfare(TSG, 12/11). This paper use various techniques to quantify the incentive to work and presents how incentives change across different individuals. The main measure in this paper is the replacement rate shows which sorts of population has weaker or stronger financial incentive to work. The replacement rate has been widely used in policy debate(e.g. NESC, 2011) not only the nationally also internationally(e.g. OECD, 2014). Consequently, the replacement rate be focused for this paper.

 

Originating in the research of Brewer(2005), Callen(2014) and Crilly(2012), the replacement rate is the most important representation of the financial incentive to work. Also, the marginal effective tax rate may whether influence the individual’s incentive to work(Callen, 2012). One of the goals of this paper is to test the replacement rate has a positive relationship with the unemployment rate. Some of literature also address this test, Savage(2014) and Callen(2012) explains that the replacement rate keeps in line with Microeconomic Theory(Duncan and Giles, 1997), which can explain the relationship between the replacement rate and unemployment rate(this is discussed further in section 2). This paper also test replacement rate whether be differential across different age, gender, region, household composition and level of education.

 

In Ireland, O’Donoghue(2011) and Mitchell(2010) compares the replacement rate across different age, gender, region, household composition and level of education. Form the existing literature in general female have higher replacement rates than males’; rural have higher replacement rate than urban; elder people have higher replacement rate; female with children with have highest replacement rate than others family composition and individuals have higher level of education will have lower replacement rate. Callen(2012) using OLS tests whether replacement rate differs in different individuals.  Also, O’Donoghue(2011) describes a simulation model which calculate the replacement rate to measure the monetary incentive for different individuals to work. However, replacement rate is seen as the most important tools to measure individual incentive to work. Comparison replacement rate across different type of group can help government to publish more precise policy.

 

This paper examines (i)the relationship between replacement rates and unemployment rates and (ii) comparison the replacement rate in different sub group: age, gender, region, household composition and level of education. The method used is ordinary least square(OLS). By performing OLS with the data^[1] can be easier understood, it is possible to analyze the replacement rate in different sub group. This paper used equation to calculate the counterfactual income to estimate the replacement rates(Engen et al., 1999; Scholz et al., 2004) argument whether replacement rate be differential in different age, gender, region, household composition and level of education.

 

The structured of this paper as follow. Section two presents a review of the exist literature, a comparison of the replacement rates across different individuals, a discussion replacement rates to measure stronger or weaker financial incentive for individuals to work. The next section is an outline of the data, describing the details of the used data. And methodology and empirical results in section 3, describing the details of the counterfactual income and estimates the replacement rates in different sub groups, also see the correlation between replacement rate and unemployment rate during observed time. The final section concludes and outlines possible research area in the future.

 

2. literature review

The unemployment rates in Ireland has a changed since the onset of the fiscal crisis in 2008. This paper analyse the changes in the financial incentive for citizen to work taking into account private earnings and social welfare. The main measure of work incentive is the replacement rate. However, some literatures indicates the participation tax rate can measure the incentive to work at all(Gregg, 1999). In addition, the marginal efficiency tax rate can measure the incentive to work for citizens(Zarutskic, 2003; Hubbard and Gentry, 2014). This section of the dissertation shows the main measure for incentive to work(2.1), introduce each sub groups be observed(2.2) and how to measure the replacement rate(2.3).

 

The incentive for an individual to work takes into account net income and hourly paid work income, versus cost of work. There should be two situations for measure the individual’s incentive to work. One is incentive to work at all, another one is incentive for individuals who are skilled or unskilled. Adam, Brewer and Shephard(2005) quantitative the incentive to work at all using the replacement rate and participation tax rate and use efficiency tax rate to measure the incentive for individuals who may have the efficiency wages. Marginal effective tax rate can measure the incentive for individuals, though working longer hours or have greater skills or effort which keep in line with standard human capital model. For instance, high tenure employees can get better paid because long time work makes them acquired general skills and raise their productivity. And marginal effective tax rate influence the intensive margin, define people who in work how many hours to work(Callen, 2013). Again, Marginal effective tax rate refers to both the impact on an individual’s earning of income tax, withdrawal social welfare benefits and others income (Tim and Niamh, 2012). However, in this paper I only focus on using replacement rate test the financial incentives for individuals to work. Test the relationship between the replacement rates and the unemployment rate, points out replacement rates can measure the individuals incentives to join in the workforce. This paper will also compare the relationship of replacement rate across different age, gender, region, household composition and level of education in this paper.

 

2.1 Replacement rate

The replacement rate is the most commonly used tool to measure financial incentive for individuals to work. It depends on a family’s disposable income when out of work versus disposable income when in work. The replacement rates is the ratio of out of work income to in work income(Adam and Browne, 2013). Replacement can always be seen as two perspectives. The first perspective is: replacement rate shows the financial incentive to work. Another perspective is: replacement reflects the level of unemployment welfare to support individuals out of work. For instance, individuals can purchase daily goods by received social welfare therefore they have no willing to join in the labor force.  Government must design the policy takes into account of all perspectives, balancing the cost and tax to keep individuals incentives to work. Also, the replacement rate keeps in line with the Standard Microeconomic Theory(Duncan and Giles, 1997), with two different effects by the wage rate increase, which is income effect and substitute effect. Again, depends on Microeconomic Theory with two distinct effect on individual incentive to work.

 

• Income Effect:  When wages increase, individuals spend less time in work but still get the same net income. It’s a negative effect for individual incentive to work.
• Substitute Effect: When individual get a higher net income, they may get more when they work additional hours. It’s positive effect for individual incentive to work.

 

The replacement rate keeps in line with this theory. It predicts that a net wage increase from working (that is to say, when the replacement rates measures decrease) individuals can get more net income, individuals more willing to work. Keeping in line with theory predicts there has positive effect on labour supply(substitute effect). While when the replacement rate increases, more net income will be replaced, people will have no incentives to work, as the same as microeconomic theory dominate an negative effect on labour supply(income effect). Again when the non-employment income rise(e.g. jobseeker benefit), replacement rate increase. Theory predicts the rises replacement rates will give a negative effect on labour supply(T.Callen. 2014). However, replacement rate have been widely used in policy debate(e.g. NESC, 2011) not only the nationally also internationally(e.g. OECD,2014). Consequently, replacement rate be focused for this paper, depends on Standard Microeconomic Theory, increase in replacement rate would leads in a decrease in individual financial incentive to work.

The replacement rate generally involves single calculation of incentive to exist in work. It calculates the amount of in-work wage which might be replaced or retained when out of work, for example by payments of jobseekers. Replacements rates are in broader context involve both in theoretical models and empirical researches of labour market as well as they are broadly involved in policy argument, both locally and globally (Scruggs, 2013).

According to Van Duijn et al (2013), both locally and globally, for the time being disposable income is most general measure involved to evaluate financial work incentives to work, for instance, the data of OECD on replacement rates and those fashioned by Fiscal Studies Institutes. Greater number of unemployed individuals did not have kids. Stress on disposable income appearing from contributions of social insurance and pay net of tax, shared with the benefits of cash which might comprise of benefits related to housing and also welfare benefits related to children and person. It is broadly known that various other elements, importantly costs of travel to work and childcare may also influence net benefit from employment. This issue has been examined by Ferrarini et al (2013) thoroughly, involving data on childcare costs’ pattern from Survey on Living and Income Conditions to evaluate the expected influence of this supplementary element on the allocation of replacement rates. It is founded by them that while definite costs of work intensified the financial work incentive to engage in employment for individuals seeking job and are unemployed, the affect was very smaller.

There is a significant proof that spells out of labour market or spells of unemployment have, typically, a negative effect on incomes which can be directed in labour market. In viewing at replacement rates, it is necessary to understand two distinct perspectives. From one aspect, a measure is provided by replacement rate of financial work incentive to work. Replacement rate is inversely proportional to incentive to work (Wang and Van Vliet, 2014). Higher the replacement rate, lower will be the incentive to work and on the other hand, lower the replacement rate, greater will be the incentive to work, however, other things will remain equal. If viewed from the viewpoint of income support goal, while, greater replacement rate is viewed as enhancing the degree of support and level of insurance allowed to those who lost their employment. It is required for the policy to be designed in a way that it strikes equilibrium between these viewpoints, in a perspective in which a significant variation is there both in potential incomes and in needs. The requirement to stable the potential disputes between work incentive and income support goals represents that careful observation of both work incentive and income support results is required (Choon and Tsui, 2012).     

Shin et al (2014) claimed that fiscal and social policy tools encounter a basic trade-off. A tool which carries out well from stipend preservation perspective might result in unintentional behavioural effects. For instance, if there is dependency of process of tax-benefit tools on the features which can be affected by people than this connection will have a significant effect on the behaviour of people. In number of cases, altering the behaviour of people is an intentional effect of the strategy and is, therefore, enviable. For instance, fines and taxes on practices, which cause negative impact on others, the tax-debit ability of unemployment or charitable donations, provide advantage in empowering job seekers to carry out a more comprehensive search operation. 

Normally, while, the incentives that are being raised as a result of such tools are unintentional. For example, high minimal tax rates raised as a result of either withdrawal of advantages or tax system minimise the net revenue from supplementary income and hence will make endeavours to maximize incomes less attractive. In the same way, taxes or benefits which rely on specific position of recipient/taxpayer have an influence on prestige of leaving or entering this position (Wu et al, 2013).

Rates of replacement are a measure of level to which living standard of individuals during work is kept during the time span of employment. The replacement rate of household is directly proportion to the effect of losing work income. Higher the replacement rates, higher they will be save from the effect of losing work income. Simultaneously, while, high rate of replacement may result in minimising endeavors of people in securing their employment (Promberger and Marteau, 2013).

According to Ferrarini et al (2013) these influences are at the core of concerns the benefits apart from work may result in inadequate endeavours to avoid slipping or escape unemployment into it. The chances which are being faced by unemployed people such that accepting the offered jobs to them wouldn’t cause any financial gain for them. Hence, they got lock into unemployment – this instance usually referred as the trap of unemployment – with viewpoint of discovering a job weakening with the passage of time. In the same way, those worked who are currently employed might not result in losing much by getting unemployed. In this way, there can be relevancy in the rate of replacement even if, as usually is the case, benefits are not provided in case of employment termination relationship is founded to be ‘voluntary’. There may have been the use of unemployment insurance by the employers to sort out demand sequences by stopping people when there is weakness of demand and re-employing them when there is strength in business.               

Replacement rates are significant tools used by people and analysts of policy for making plans about retirement and assessing the sufficiency of social security advantages and overall income of retirement. There is a differences in meaning and measurement of replacement rates on the basis of definition of preretirement earnings. The concept of replacement rates are used by both individuals and makers of policy. Replacement rates are used by them for expressing retirement income as a portion of preretirement earnings. These rates are used by individuals as a rule of thumb in retirement planning. The measures of replacement rates are used by policy makers for analysing benefit of social security under recent schedule of benefit versus that might be given under different policies (Breen, 2005).

However, there exists confusion with respect to the utilisation of replacement rates. Particularly, it is easier to isolate the numerate in equation of replacement rate, either periodic social security payments or total income of retirement. There is no consensus given by scholars regarding the appropriate way of measuring preretirement earnings.

As a consequence, individual planning and discussions of policy often integrate various measures of preretirement earnings that can results in wrong calculations regarding current or future rates of replacement. Particularly, it is commonly accepted that 70 percent is usually considered to be an adequate rate for retirement income from various sources. In addition to this, 40 percent replacement rate consists of social security benefits. The replacement rate is measured by financial advisors in relation to earnings rapidly preceding retirement. In contrast to this social security replacement are measured in relation with average of wage-indexed lifetime earnings. The financial advisors cannot draw valid conclusions on the basis of replacement rates that have been calculated with the help of various denominators. When there is a variation in denominators of replacement rate equation, then true and error free results cannot be ensured (Rajevska, 2016).                   

The rate of replacement rate of individuals relies heavily on the situation of individual. It is obvious that prospective or current wage degree is involved is playing a significant role. For the provided degree of benefits, lowers incomes will usually match up to greater rates of replacement. Generally, there will also be various other wages, given to the same individual or to any other person of his family, which have an impact on comprehensive income of family, and therefore, the rate of replacement. Moreover, costs which have to be experienced in situation of one labour market but no other would have a significant impact on disposable wage for example, job search costs, costs of commuting to work, union costs, costs of giving care in working hours to dependents etc. While, considering control of policy makers, every income component is less eagerly attainable than benefits and taxes. Partially for this rationale and due to more control of policy-makers over transfer policy and tax as compared to over decisions of household, the impact of tax-benefit schemes has been the center of attention (Savage et al, 2014).

Previous researchers have evaluated the degree of unemployment benefits as a portion of employment wage in separation from other schemes of tax-benefit. Since, it is known that the exclusion of benefits and taxes apart from benefit of unemployment creates over-simplified outcomes which can be badly ambiguous. There is more concern of people regarding their gross incomes and it is required for the measurement concepts to exhibit this (Kendal et al, 2016). Ongoing income taxes on incomes attached with a tax-free or favourable position of benefit payments reflect that the rates of replacement before taxes are noticeably lower as compared to the alleged net rates of replacement which are calculated net of contribution payments and tax. Similarly, the significance is of the fact that there has been founded more extreme values of rates of replacement for few individuals are commonly because of very composite interdependencies between portions of tax-benefit scheme which have been established at various times, with multiple objectives or are operated by various agencies or authorities. Unless every appropriate portions of tax-benefit scheme are considered, these ‘irregularities’, which might result in having very severe allegations for employment incentives, will not be displayed in resulting rate of replacement measures (Rajevska, 2016).           

 

 

Replacement Rate( RR)=  100 

 

For example, If a family disposable income is EUR 200 in one week when out of work and they have EUR 400 for disposable income when in work, then for this situation the replacement rate for them is 50%. Again, when the net income is EUR200 in work and EUR200 out of work, means the income be totally replaced, it will leads individual have lower incentive to work. When the replacement is 70% will be seen as excessive.

 

• There is Hypothesis 1: H₁: Replacement rate have relationship with unemployment rate.

 

2.2 Different replacement rates on different individuals

Where replacement rate always be used to measure the incentive for individuals to work. Different age, region, gender, household composition and level of education individuals present different replacement rates under different policy in the labour market. This paper will focus on different individuals to see what sort of population meet stronger incentive to work or weaker incentive to work.

 

(i)Household Composition:

Some researcher(eg; Adam,S; Brewer M, 2005; Callen, kelly& Savage, 2015) point out women in couples with children face highest replacement rate than other type of population and lone parent have higher replacement rate than other population, and single adults have lowest replacement rate than any other population.

 

(ii)Age:

In addition, groups of different level of age on replacement rates should have different performance. Replacement rate always be measure how elder individuals can maintain their pre-retirement level of consumption once they out of work(Munnell and Soto, 2005). There are some different approaches to analysis the replacement rate across different level of age. Some researches(e.g. Engen et al., 1999; ?Scholz et al., 2004) using simulation technique to calculate the counterfactual income to estimate the replacement rates. O’Donoghue and Li(2011) represents an obvious peak over aged 65. Also, OECD(2011) shows the minimum payment rate for 18 to 21 was €100 and for age 20-24 was €150 and nearly 24% individuals will be affected by this reduction in minimum payment rate. Significantly, the reduction in potential entitlement makes those age individuals less likely to get the payment.  Correspondingly, the replacement rates facing an decreased for individuals aged under 25(ERSI, 2012). The result points out highest replacement rates in the group of aged over 65 and lowest replacement rate in the group of aged below 25.

 

(iii)Gender:

While, gender also have effects on the replacement rate across the population. Adams, Brewer and Shephard(2012) using Family Resource Survey(FRS) 2002-2003 and TAXBEN under 2005 tax and benefit system to calculate the replacement rate of different type of groups shows gender and marriage status always have different replacement rate. Female have higher replacement rate than male, especially female with children(e.g. Adam, Brewer&Shephard, 2012; Callen, kelly & Savage, 2015). Using abundant of data sample to compute the the effects of change in the sample of household and inference the effect for whole population, also shows the female have higher replacement rate than male(C.Giles 1995). Female out of labour market more than male in recent years(Larry. D and Michael. C).

 

(V)Region:

However, some research using the simple Harris-Todaro model divided into an urban sector and a rural sector(e.g. Corde W.M&Findlay, R, 1975; Josepg E.S, 1974). And unemployment is consistent with equilibrium in this model, points out unemployment rate in urban is decreased. As the unemployment rate decreased, replacement rate decreased as well. The replacement rates for urban individuals lower than rural individuals.

 

(VI)Level of education:

Moreover, the level of education also one of the important factors affects the replacement rates. Higher skilled individuals will have lower replacement rate, when employee have higher education means employee would becomes the skilled employee suitable for the labour market. Higher unemployment is association with the poor education standard. Comparing with the different unemployment rate across level of education, about 10% unemployment rate for low education, 5% unemployment rate with upper second degree and 4% unemployment rate with tertiary education. The research shows higher unemployment in the lower education level (e.g Stephen Nickell,1997;  Daniel Oesch,2010). Some research analysis the replacement rate across different level of education individuals by the unskilled or high skilled employee influence on the unemployment rate(Richard,F & Ronald.S,2001; Richard.B 2005)

 

• There is hypothesis two: Different Replacement rates will be deserved across different household composition, age, gender, level of education and region.

 

2.3 Approaches and data conducting the replacement rates in previous research

Three approaches are used to measured replacement rate in Ireland,

1. According to Hughes Walsh(1983) and Blackwell(1986), the replacement rate can be measured by social welfare payment rate and average industrial earning in cass. Again, calculated by Department of Finance e.g.NESC(1993) While, depends on those cass cannot reflect the real situation in actual circumstance, may also lead individual would like to unemployed.(Atkinson and Micklewright, 1985). However, since O’mahony(1983) shows there actually have correlation from those two factors, Nolan(1987) ^[2]derive the Walsh theory, tracked the time changing and points out the hypothesis cass less focus on the time-series changing, fails to reflect the replacement rate in actual circumstance.
2. The second approach calculate average replacement rate from unemployment benefit (UB) and unemployment assistance(UA), data from a massive of expenditure and claimant numbers, then compare with per employee average income. This measurement put forward by Browne and McGettigan(1993) but this approach only focus on the overall trend among the time change.
3. The third one using the large sample of household and involves the modelling of in-work and out-of work income to estimates the current replacement rate for those unemployment or employment. And for 1987 ESRI survey on income, poverty and use of state services give a basis for the tax and benefit microsimulation model(Callen, 1991). Basic this model, Callan, O’Donoghue and O'Neill(1994) using the data from 1987 survey can get fully evaluate the change for distribution of replacement rate. Overall, using this model can see uprating sample of change in the tax, welfare system and  in average earnings over time.

 

A microstimulation study of financial incentive to work should be built of the tax/benefit model. This paper focus on using data from CSO, the replacement rate calculated by the tax-benefit model. The results dominate the different individuals will have different distribution in replacement rate. Using the third approach can actually reflect the change in distribution of replacement rates over time. To see the relationship of unemployment rate with replacement rate over time, if reliance on the microsimulation model^[3](Callen,1991) there has correlation with replacement rate and unemployment rate, shows is there has negative relationship or positive relationship. However, the distribution in replacement rate can reflect financial incentive for individual, if there has lower replacement rate(e.g.20%) means individuals who has incentive to work. On the other hand, if there has higher replacement rate that means more individuals earning will be replaced then people will have less incentive to work(Duncan and Giles, 1997). For example, if there has 100 percent replacement rate shows the earning of individual will be totally replaced, that is to say, individual will better off not work and rely on social benefit. Also, using counterfactual income to estimates the replacement rate. It is possible to comparison the replacement rate across different age, gender, region, household composition and level of education.

2.4. Financial Incentives to Work

Financial incentives to work can be caught via measures which represent the proportion of supplementary income taxed away through abstraction of social insurance offerings, benefit withdrawal or direct taxation (Kostøl and Mogstad, 2014). A difference is generally made between financial incentives to work at extensive or intensive boundary of worker supply. The measures at the exhaustive borderline reveal the incentives to maximise worker supply. Work incentive along wide-ranging borderline involve with the decision to work or not, and therefore to shift between inactivity or unemployment and employment. In order to place it another way, the intensive and extensive borderlines correspond, correspondingly, to rate of employment and to the time of employment (Savage et al, 2014).

It is confirmed by Laroque and Salanié (2014) that both of the borderlines matter in describing alterations in total supply of labour and worked hours. While, the connection between extensive and intensive borderlines and work incentives are not linear and can be impacted by elements external to the tax-benefit scheme. For instance, financial incentives are analysed by Shaw and Gupta (2015) to operate in perspective of composite tax-benefit improvements and wage maximises in UK. It has been shown by them that in spite of diminishing tax rates work incentive at extensive borderline can be destabilised till people give their reaction to disposable income level comprehensively of work. This creates an exchange between general economic, benefit adequacy and tax schedule situation in market. It also requires carrying investigation of work incentives at extensive and intensive borderlines mutually, and also for analysis of involvement of various elements to financial work incentives (Promberger and Marteau, 2013). Two wider practices of investigating financial work incentives can be differentiated further: through theoretical model-type family measurements or with applied micro-simulation on representative facts and figures. The model-type family judges specific scenarios of labour market for selected households. Measures of work incentives can be obtained for the selected estimations of gross incomes. This forms the measurements and outcomes easy to understand. Model family-based measures of low income, traps of unemployment and inactivity are frequently published by Organization for Economic Cooperation and Development (OECD) in collaboration with European Commission (Fulbeck, 2014).           

The institutions of labour market, specifically benefits of welfare, have been recognised as giving to the dynamics of unemployment in the economy market. For instance, it is founded by Giles et al (2014) that larger than one half of the emergence in unemployment in Western European from 1960s to initial half of 1990s is described by alterations in institutions, specifically benefits of welfare, unions costs, labour taxes and employment protection. It is argued alternatively by Glasziou et al (2012) that increase in the rate of unemployment is described not by institutions themselves, however rather by communications between shocks and institutions.

The significance of financial disincentives for unemployed and employed individuals is evaluated. Depending upon the panel survey linked with the managerial registers, they calculated the financial incentives for the participants of Danish labour between employments and being on the benefits of unemployment, however, they also clarify for the immovable costs of the work like getting across and the care costs for the child (Brown et al, 2014). From the results, we came to know that in the year 1996, there were 6 percent men and 13 percent women who had efficient rates even more than 100%, which they comrade with the work disincentives. However they also have several attitude measures which are included into the relapses, the main point that the financial measures have got the most vigorous effect on the risk of being in trap of unemployment.                                  

3. Date and Methodology

3.1. Introduction

A research needs to be conducted depending on the best and appropriate methodology. This is significant because of the fact that the major part is played by it in the development of required results of research (Kumar and Phrommathed, 2005). The researcher has explained the methodology which has been used for the research and has also given the justification for it. Before making the selection of specific research method, the researcher has evaluated the choices which he has, and then methodology has been selected which is more effective in terms of acquirement of research objectives.

The research has been conducted for the interpretation of effect of education on the organizational performance. Such methods and procedures have been used which are beneficial for achieving the targets and objectives of research. With the help of research methodology, collection of data becomes easier.

3.2. Research Methodologies Available

Selection of proper research methodology is dependent over various problems that are faced while conducting the research. There are different choices available, but those processes have been chosen by the researcher who ae linked with the nature of issue. It is a process that ensures the progress of research and provides a proper shaping to the research (Mackey and Gass, 2015). In the research introduction, the problem statement is explained and design of the research questions is made. The research questions give the explanation about the research method. However, critical thinking must be used. According to Flick (2015) it is dependent over skills, abilities and knowledge. As it has been given that both of secondary and primary data are used in the research and through these processes the information has been collected in an effective manner. Household individuals were the major primary source of information; journals, books and articles are the secondary source of information. In addition to this, positivism philosophy has been utilized in this research. This is the basic methodology of research and it is more rigorous for study. In addition to this, in the further headings the other processes and explanations related to methodology are given.

 For conducting the research, two research methods can be utilized. Qualitative and quantitative methods are the instances of these two research methods. When quantitative research methods are used, the studies are dependent on figures, facts and calculations. Such researches which are based on calculations and evaluations should be dealt through quantitative data (Neuman and Robson, 2012). In contrast with this, qualitative data is used for exploring the subjective problems in research. For getting individual feedback of respondents, qualitative method of research is used. In case of less availability of data, qualitative research method is used (Pickard, 2012). Qualitative method involves high dependency over problems of subjectivity therefore specialists mainly criticize it. Because of the subjectivity, the researcher can be get more biased for analysis and collection of data (Blumberg et al, 2014). Due to this, reliability and trustworthiness of research can be decreased. Because of these reasons, quantitative research method has been used in this research. The major reason behind using this procedure is that it assists in accomplishment of the research in valid and reliable way (Kumar and Phrommathed, 2005). It was assumed at the time of selection of method that researcher won’t be able to refer to the subjectivity problem at this step. Therefore, it was quite difficult to decrease the biasness of individuals that is possessed b them in qualitative research method. Because of the reason, that recent study has been conducted through qualitative research methods. By using the procedures of quantitative research, the data analysis is done and because of this the research can be made more reliable. In this research, quantitative research process is used. In accordance with this procedure, the outcomes have been acquired relevant to the facts and figures. For acquiring the required outcomes, the questionnaire has helped in the collection of data. This assisted the research to acquire the valid and reliable outcomes (Billig and Waterman, 2014). Through quantitative research methods, the data was collected from household individuals to analyse the important of replacement rates and analysing the replacement rates across the different education level, age, gender, region and family type of individuals by calculate different predict wages for each independent variables.

3.3. Research philosophy

Ten of the philosophies can be used for research; these philosophies have been defined by the research onion linked with the uses and implication in researches. Though, only few research techniques have been used but for the researcher it is important to consider the other choices too. This can prove to be helpful for him in selecting the one most suitable for conducting the research and for getting the beneficial data for study (Green et al, 2012).

In this research, positivism philosophy has been used. This philosophy proves to be helpful in developing the knowledge regarding the particular cases. It assists in identifying the view of researcher regarding the solution of problem or issue and particular condition with the help of implementation of some particular procedure. This also gives an idea to the researcher regarding his conduct during conduction of research (Williams et al, 2011). Through positivism research philosophy, generalization has been taken into account by the researcher. In addition to this, the research has conducted the research in a more systematic manner. The positivism helps in acquiring the objectives of research regarding the analysis of importance of replacement rates for different types of individuals. This philosophy proves to be helpful for conducting the research in terms of interest fields.

3.5 Sampling Method and Population Sample

As given, the survey research technique is utilized for obtaining the targets and objectives of research and for answering the research questions. For performing the survey, the design of questionnaire has been used. The sample was selected dependent on the random sampling. This sampling method was selected because of the reason that it takes less cost and time (Pickard, 2012).

Most research uses only one year of unemployment rate and replacement rate to measure the relationship between these two variables. This paper focus on 2005 to 2012 data to provide a more detailed analysis of unemployment rate and replacement rate and explore what relationship exist between them. Also, it analysis if different types of individuals will have different replacement rates and the relationship between the replacement rate and different level of education, age, region, gender and household type.

 

The data used is of EU Income and Living Conditions in Ireland (SILC) from 2005 to 2012, which is a survey covering an abundance of  issues relate to living condition and income, each survey includes large households and a mass of interviewed individuals. It is an official resource used by the CSO information about the issues living household income across different types of households in Ireland. This allows me to identify sample individuals and subgroups. Information on household annual net earnings and household annual benefit from government welfare system are collected. “Household composition”, “household annual employee income” and “household unemployment benefit”.

3.6 Method of data collection

In recent research, the strategy of survey has been used for obtaining the targets of research and for making answers to the questions of research. For deriving valid results, researcher has made the use of primary data. For collecting the primary data, both of web-based and self-administered questionnaire has been used. The researcher suffers from too much complexity during the collection of data as researchers seem to be concern for their identity. The researcher has used hand and online delivery survey and email for the collection of data through participants. Some of the participants were given questionnaires in hand and the participants who could not be approached were provided with the questionnaire through online survey and email.

Penalty is also expected due to expose of secret and important information of respondents. So the researcher needs to ensure that identity of participants will not be revealed and that there will be the strict maintenance of the confidentiality of participants. The researcher used the data only for the purpose of research.

These variables are collects from SILC and used to measure replacement rate.

• “household composition”, “education”, “Rub_rur”, “Age” and “gender” can help to identify different sub group of the sample. This allows an analysis of whether the replacement varies by household type, age, levels of education, region and gender phrasing.

• “household annual employee income” is used to calculate the predicted wage for unemployment individuals. Using the predict wage and welfare to identify the replacement rates. Using the regression of the replacement rate can see the differ changing in different subgroups.
• “household annual unemployment benefit” from SILC is numeric values is used to identify the welfare for unemployment individuals. Welfare equals to the benefit they received. With data “Household annual employee income” using the SWITCH model can calculate replacement.

• “Replacement rate” be calculated using the SILC data. The changing replacement rates across  different subgroups of the population can be beneficial in understanding which part of population face weak or strong financial incentive to work. In addition, new policies to boost individual work incentives can be developed.

3.7. Research Instrument - Questionnaire

In current research, the collection of data has been done through questionnaire. Questionnaire proved to be helpful in collection of data regarding the importance of replacement rates for household individuals. Flick (2015) has stated that most important thing is to target the administration and design of questionnaire. There are some of the alternatives for the collection of data but it is suitable as it helps in collection of data in cost effective and quicker way. Questionnaire involves the higher response rate. In accordance with the recommendation of John Kuada (2012) literature review is utilized for making the design of questionnaire.

For conduction of research, questionnaire was used being a research instrument. This instrument proves to be helpful in making measurement of the variables to be studied in research. There are three sections of questionnaire. The first section refers to the demographic information of participants. The second section refers to the analysis questions for each independent and dependent variable. The design of questionnaire has bene made depending on the previous studies. For designing the questionnaire in the most effective way, it is significant to select the right scale. The extent to which the feedback given by individual is accurate is based on the scale used. Along with the items regarding variables, questions regarding demographic variables were added for comparing the participants depending on the demographic information.

3.8. Piloting

Before distributing questionnaire to all of the participants, it is distributors to some household individuals. After collecting the data, the analysis of the feedback has been done. Along with this, the questionnaires have been tested for their reliability through analysis of reliability. After testing the questionnaire, some of the alterations have been made in some items of questionnaire.

3.8. Data Analysis

For deriving the results from the collected data, it is significant to do the analysis of data. The analysis of data can be done through particular strategies and tools. In recent research, SPSS software has been used by the researcher for the analysis of data gathered through participants. The usage of descriptive analysis has been made for the analysis of feedback acquired through participants for every item present in the questionnaire (Billig and Waterman, 2014). During the usage of SPSS software, correlation and regression analysis were utilized for testing of the link in between dependent variables and independent variables. Regression analysis informs about the limit of link in between variables, on the other hand correlation analysis informs about the power of link in between these variables (Pickard, 2012). In addition to this, analysis of reliability is used for making test of the consistency of questionnaire with the research problem.

3.9. Data reliability and validity

The validity and reliability of the questionnaire has been tested by the researcher before the distribution of the sample. The testing of validity of questionnaire was done by consulting to experts and professionals of replacement rates. This assisted in acquiring the specialist insight in questionnaire. Depending on the perception, needed alterations were made in the questionnaire. Along with this, the questionnaire was checked for being reliable through testing its consistency with problem of research.

3.9. Ethical Issues

For ensuring that the research has been conducted in fair way, it is significant to comply with ethical standards and norms of research. The research should be conducted depending on the ethical exercises. In present research, researcher has ensured that ethical standards, norms and practices have been fulfilled. For the collection of primary data, the researcher ensured that participants have been identified and not dependent on the personal capacity. The confidentiality of participants is maintained by the research. Furthermore, there was no enforcement over participants to fill up the questionnaire. Participants were provided with ample time to fill up the questionnaire. This proves to be helpful for researcher in acquiring unbiased data.

3.10. Summary

Methodology has been given according to the problem statement in this chapter. The most important factors of methodology of research include the research procedure which involves strategy, approach and philosophy of research for conducting the research. In addition to this, the justification of research procedures has also been given.

 

Overall, data from SILC can be beneficial in understanding how many percentage each groups take in the total population. Average for annual employment income and average annual unemployment benefit reflect the statistical average of annual employment income and annual unemployment benefit.  Table 1 shows an increase in annual unemployment benefit besides 2009 year, from €1,265.995 to €2,809.335 and totally low in 2009 is €881.010 and average annual employee income have a downward trend, from€31,532.47 to €26,900.94, also in 2009 there has a peak in €11,292.90. Table 2 shows how many percentage each different household types engaged in this survey in each year. In general, from 2005 to 2012 one adult with children always take accounts for very low proportion on average 5.95% and two adults without children or with children take almost similarly large proportion on average separately 22% and 27%. Also, we can see the household composition changing from 2005 to 2012, one adult with children shows a trend of increase, as well two adults with children. While, one adult without children, two adults without children, three adults without children and other household with children represent a trend of decrease. Family composition three adults without children shows a extremely downwards from 20.62% to 11.14%. Table 3 also shows the changing for percentage of different level of education individuals during 2005 to 2012. Across the population, individuals in no primary education level take a largest proportion from 2005 to 2010, almost 30% percentage. Individuals who get the upper secondary also get higher proportion than others subgroups, nearly 23%. Anyway, the percentage of individuals who not get no education decrease from 29.52% to 20.10%, as well lower secondary and post leaving cert. Individuals getting upper secondary level of education almost keep constant changing only 0.04%. But the percentage of individuals who get third level of education no degree and have degree or above all shows rises trend. Overall, this table shows individuals all enhance their level of education in ireland. Changing for age subgroups take how many percentage in the total population shows in table 4. Aged from 25 to 49 takes largest proportion in the total population and rises from 29.33% to 32.33% during 2005 year to 2012. Also aged 0 to 14 increased from 20.97% to 24.74%. However, group individuals aged 15 to 24, 50 to 64 and over 65 all decrease. In addition, table 5 shows different gender take how many percentage in the total population. Female always have higher proportion than male, but the distinct not obvious. Only nearly 3% difference between female and male. Whereas, table 6 shows the urban or rural individuals takes how many percentage in the total population. During 2005 year to 2012, urban individuals decreased from 64.25% to 58.83% but rural people raised from 35.75% to 41.16% in total population.

 

Source: EILC  2014   table1

                                  Annual employee income and Annual unemployment benefit from 2005-2012

	2005	2006	2007	2008	2009	2010	2011	2012
Annual employee income Average	32,246.50666	34,348.77872	36,492.42782	35,997.69197	11,292.90778	25,152.232268	25,863.4883565	26,900.94293401
Annual unemployee welfare  Average	1,204.422056	1,308.160109	1,809.68487	1,855.104547	881.010220873	2,764.316161623	2,806.604934	2,809.335036828

 

 

 

 

 

Source:SILC  2014       Table2         

                                           Each household Groups take how many percentage in Total Population

 

Type	2005	2006	2007	2008	2009	2010	2011	2012
1 adult Without  children	12.12%	12.66%	13.02%	12.55%	12.08%	12.05%	11.06%	10.35%
2 adults without children	20.96%	22.18%	23.67%	24.47%	22.35%	22.57%	19.42%	18.70%
3 adults without children	20.62%	21.36%	14.74%	13.08%	12.32%	16.86%	10.44%	11.14%
1 adult,1+ children	3.80%	3.60%	5.32%	6.40%	7.15%	5.24%	7.75%	8.40%
2 adults, 1-3 children	22.53%	20.92%	26.32%	26.63%	28.46%	26.44%	33.75%	33.65%
other household with children	19.75%	19.27%	16.90%	16.84%	17.62%	16.80%	17.54%	17.20%

 

Source: SILC 2014   Table 3

                                           Each Education Groups take how many percent in Total Population

 

Education	2005	2006	2007	2008	2009	2010	2011	2012
No formal/ primary education	29.52%	29.59%	29.17%	29.24%	25.59%	24%	21.88%	20.10%
Lower Secondary	19.38%	19.34%	19.35%	18.94%	17.58%	17.73%	16.68%	15.95%
Upper Secondary	22.96%	21.97%	22.28%	21.96%	23.41%	23.24%	22.37%	22.92%
Post Leaving Cert	7.35%	7.32%	7.03%	7.35%	7.07%	6.61%	6.46%	6.82%
Third level-no degree	7.56%	5.19%	7.91%	7.70%	6.20%	3.53%	15.28%	16.51%
Third level degree or above	13.20%	16.55%	14.23%	14.70%	20.11%	24.77%	17.30%	17.68%

 

Source: SILC 2014   Table 4

                                           Each Age Groups take how many percent in Total Population

 

AGe2	2005	2006	2007	2008	2009	2010	2011	2012
0-14	20.97%	20.02%	18.87%	18.13%	20.45%	22.60%	23.95%	24.74%
15-24	13.47%	13.03%	12.51%	11.95%	11.29%	10.90%	10.45%	10.51%
25-49	29.33%	28.13%	27.76%	26.76%	29.08%	31.22%	32.24%	32.33%
50-64	18.02%	19.13%	19.74%	21.34%	19.91%	17.60%	16.53%	16.45%
65+	18.19%	19.67%	21.11%	21.81%	19.24%	17.65%	16.80%	15.95%

 

Source: SILC 2014   Table 5

                                           Each Gender Groups take how many percent in Total Population

 

SEX	2005	2006	2007	2008	2009	2010	2011	2012
Male	48.90%	48.94%	48.30%	48.46%	48.52%	49%	48.60%	48.20%
Female	51.10%	51.06%	51.70%	51.54%	52.48%	51%	51.40%	51.79%

 

Source: SILC 2014   Table 6

                                           Each Region Groups take how many percent in Total Population

 

REGION	2005	2006	2007	2008	2009	2010	2011	2012
Urban	64.25%	63.13%	62.78%	61.79%	62.63%	59%	60.03%	58.83%
Rural	35.75%	36.87%	37.22%	38.21%	37.37%	41%	39.96%	41.16%

 

Measure the financial incentive for individual to work still needs unemployment rate data, compare those data with replacement rate can prove: weak or strong incentive correspond to lower or higher  replacement rate and reflect what type of sample would fact the weak or strong financial incentive to work. The data is used is 2005-2012 “Seasonally Adjusted Annual Averages standardised unemployment rates by state and year” (CSO,2014). This survey is a large-scale and the whole country survey of household in Ireland, also it be designed to produce the data about the unemployment and employment in the state. All data about unemployment among ten years consists with the data for replacement rate from CSO, which can compare with each years replacement rate and see the changes during ten years and get result to see if the higher replacement rate will have higher unemployment rate. If there actual has high unemployment rate with high replacement rate, which means the results are same as the exist theory. Also, using replacement rate data from OECD “Unemployment Trap, 2001-2012) in order to compare with unemployment rate, this report only using the data from 2005 to 2012. In this survey replacement be identify by Net Replacement Rate(NRR) and exclude social assistance(SA) and housing benefit(HB), which covering two rning level, three family situation and 60 months of unemployment. Although this survey includes several countries, this report only focus on Ireland replacement rate.

• “Seasonally Adjusted Annual Averages standardised unemployment rates by state and year” is numeric value reflect different unemployment rate among 2005-2014 year, it is useful for compare with replacement rate to prove the relationship between these two variables.

 

 

 

 

 

Source: CSO 2014  table 7

                                           Annual Averages Standardised Unemployment Rates(%)

year	2005	2006	2007	2008	2009	2010	2011	2012
state	4.4	4.5	4.7	6.4	12.0	13.8	14.6	14.7

 

 

• “NRR summary measure of benefit entitlements (excluding both SA and HB), 2001-2012” also shows the numeric value and chose data from 2005 to 2012 correspondent to unemployment then get the compare with these two variables.

 

 

Source: OECD, Tax-Benefit Model               Table  8         

                                                         Replacement Rate Exclude SA and HB(%)

year	2005	2006	2007	2008	2009	2010	2011	2012
Ireland	50	52	54	58	63	60	59	58

 

Overall, table 7 shows an extremely upwards trends for annual average unemployment rate in Ireland from 4.5% to 14.7% between 2005 to 2012 year. Finally, table 8 also shows a total increase trends in replacement rate in Ireland from 50% to 58% between 2005 to 2012 year and the peak at 63% in 2009.

 

 

4. Methodology & Result

4.1 Generating Counterfactual Wages

This paper aims to analyses the important of replacement rates and analysing the replacement rates across the different education level, age, gender, region and family type of individuals by calculate different predict wages for each independent variables to get an multiple regression model for replacement rate in different age, education, region, gender and family type. Finally can summarize what effects on replacement rate by the different type of household and analysis which type of individuals have lower or higher incentive to work. The higher the replacement rate the lower incentives to work(T. Callen, 2014). The wages equation shows the effects of age, education, region, gender and household type on employed person. It presents the relationship between different independent variables and wage. This multiple regression model is displayed as equation (1).

 

W_i=₀+₁ Age_i +₂ Education_i +₃ Region_i+₄ Gender_i +₅ Household Type_i  (1)

 

Where W_i which is dependent variable indicates the wage of person i associated with the independent variables: age, education, region, gender and household type. The predict wage for unemployed individuals depends on different group of age; different education level; urban or rural region; female or male and different household type. Also the dummy variable age, education, region, gender and household type are the independent variables in this equation. Predict wage depends upon these variable. Where i as observation type number goes from 1 to N. Thus, age_i; education_i; region_i; gender_i and household type_i as dummy variable indicated the ith observation of independent variables age, education, region, gender and household type. And ₀ is constant or intercept term in the linear regression model which means the value of predict wage W_i when all the explanatory variables age_i, education_i; region_i; gender_i and household type_i are equal to zero. Age is a series of dummy variables while indicates (i)individuals from 0 to 14 years old; (ii)individuals from 15 to 24 years old; (iii)individuals from 25 to 49 years old; (iv)individuals from 50 to 64 years old; (v) individuals older than 65 years old. In addition, i in categories variable households type from 1 to 6 is dummy variables: (i)one adult without children; (ii)two adults without children; (iii)three adults without children; (iv) one adult with children; (v)two adults without one to three children; (vi)other households with children. And i in education variable also from 1 to 6: (i)no normal/primary education; (ii)lower secondary education; (iii)upper secondary education; (iv) post leaving cert; (v)third level but no degree; (vi) third level with degree or above. _1,dominate the effects on the predict wage W_i of a one unit increase in Age when holding constant the education, region, gender and household type, as well ₂,₃, ₄ and ₅. Anyway, age_i indicate a vector of i variables may have influence on predict wage i as well the others dependent variable education_i; region_i; gender_i  and household type_i. Using this predict wage regression model can calculate the coefficient for each explanatory variables, then I can get the replacement rate regression model.

Moreover, analyses the relationship between replacement rate with all the explanatory factors: age, education, level of education, gender, region and household type by the estimate replacement rate equation 2:

 

RR_i=₀+₁ Age_i +₂ Education_i +₃ Region_i+₄ Gender_i +₅ Household Type_i (2)

 

RR means continuous variables replacement rate dominate the change in associate with age, education, region, gender and household type. In which case, replacement rate is dependent variable effects by different age, gender, region, level of education and household type. In addition, age, education, region, gender and household type are independent variables. While, all the observation groups same as the predicted wage regression model. From this equation, get the constant value and different coefficient for different observation groups to analysis the relationship with replacement rate across different observations. Once identify the replacement rate equal to welfare divided by predicted wage, getting each observations replacement rate can analysis the relationship with all the explanatory variables. After the test, if the result shows there has relationship with household type means the result same as the hypothesis two.

 

Overall, this report wants to analysis the replacement rate across different types of individuals to see which type of individuals have higher or lower incentive to work. The relationship between replacement rate and unemployment rate is important. It shows a link between the replacement rate and the incentive to work. Using correlation to measure the strength of the relationship with replacement rate and unemployment rate. If there is a strong positive relationship this means provide a support for hypothesis one in section two.

 

This can summaryse the analyses as following

1. Estimate the effects of demographics factors on the wages of  those in work.
2. Using the coefficient from (1) to generate the predicted wages of those out of work.
3. Calculate the replacement rate using the hypothesis in work wage and actual unemployment benefit.
4. Run a regression where the replacement rate is the dependent variable and use demographics factors to explain the replacement rates.

 

4.2 Counterfactual Results

SPSS Tables appendix describe the spss result,

In general, we can see for all counterfactual wage estimates are statistically significant the regression  table 9 to table 16, shows the significant value for the F test is zero. This means the equation for predict wage can accept. In addition, the value of R square is approximately which means how much of the variation of predict wage can be explained by family type, region, gender, age and the level of education.We can see the individuals coefficients in this equation result for each independent variables. If the sig value is below 0.1, it means the valuable statistically significantly. While, if the sig value is below 0.01 mens it significant by 99% level, but sig value not below 0.01 but less than 0.05, it indicate this valuable is significant at 95% level, below 0.1 sig at 10% level. Also, the table can describe the constant variable which should be final output when all the explanatory variables be zero. Table 17 to Table 25 indicate the result of different type of household type, region, gender, age and level of education effects on predict wage. In general, for all years, the results are significant. Overall, from this test and getting the coefficient for each independent variables in predict wage can calculate the predicted wage for each groups. The replacement rate can be calculated by the predicted wage and welfare. According to the estimates replacement rate equation:

 

Replacement Rate( RR)=  100 

Using welfare as the out of work family disposable income and predicted wage as in work family disposable income the above equation calculate the replacement rate. The regression model of the replacement rate can help to analysis the relationship between replacement rate and different observations.

 

4.3 Replacement Rate Result

This section discusses the impact of different variables on individuals replacement rates. The result are presented in the appendix from table 25 to table 32. Thus, in general, all years from 2005 to 2012 the sig value for the F test are all equal to zero, we can see from table 41 to table 48. The replacement rate can be explained by the independent variables: different categories of age, gender, region, household type and level of education. For instance, for Table 35^[4] shows the R square is equal to 0.138 indicate there has 13.8% replacement rate be explained by the different categories of age, region, gender, family type and level of education. In addition, all region, education, household type and age sig values are zero indicating those variables are significant at 99% level. But, age₂^[5] shows sig value as 0.030 dominates aged from 14 to 25 years old significant at 95% level. From Table 35 to Table 40, for all years from 2005 to 2012, the R square all above zero and almost average 0.134. In general, sig value for T test amounts shows different education levels and ages have effects on replacement rate from 2005 to 2012. Exceptions to this one individuals who at (i)lower secondary and at age from 15 to 24 years old in 2008; (ii)age range from 25 to 49 years old in 2009 year; (iii)age range from 25 to 64 years old in 2010 year; (iv)aged from 25 years old in 2011 year; (v)aged from 25 to 49 years old and at lower secondary education, as well post leaving cert in 2012 year. In addition, gender significant in 2009 and 2010 year (as Table 29 and Table 30), means only two years different gender would be effects on the replacement rate. As well, during the 8 years observation, individuals who live in rural or urban would influences replacement rate in only three years which is 2005, 2007 and 2008 year(as Table 25, Table 27, Table 28). The details of region be discussed in the next paragraph. While, households type especially two adults without children and three adults without children during eight years observes almost all have effects on replacement rate besides in 2009 year. The replacement rate shows negative in family two adults without children and three adults without children. The result reflect different family type have different replacement rate. But for one adult without children, one adult with children, two adults with children and other households with children have no difference on replacement rate in 2005, 2009, 2010, 2011 and 2012 year. Consequently, it appears that gender is only significant in determining replacement rate in two years. And across family type, only two adults without children and three adults without children is significant factors in determining replacement rate. Moreover, individuals who live in urban or rural area is not an significant factor in four years after all, education and age is significant in determining replacement rate. Education and age are the only two consistent factors explaining the replacement rates.

 

In general, the result of the multiple regression model of replacement rate indicate higher education would have lower replacement rate. The higher level of education, the lower coefficient for education in which cases indicant lower replacement rate. For instance, the coefficient is -0.025 for individuals who at lower secondary education level means replacement rate lower than individuals who at primary education. With higher education coefficient is -0.046, -0.084 and -0.098 separately in 2007 which dominate higher education level would have lower replacement rate during the reference period. It indicate education _i is negative, higher education level of individual who would have lower replacement rate. Also, age_i is a dummy variable shows positive, as individuals age increase, replacement rate also increase. For example, in 2008 year the coefficient for aged from 25 to 49 is 0.034, represent compare to the aged from 15 to 24, the replacement rate increase. When analysis the households type focus on the replacement rate of one adults without children and two adults without children. Where the replacement rate of families have two adults without children always lower than families have one adult without children. In 2008 year, household composition one adults without children coefficient is -0.031 means the replacement rate decrease, represent different family composition have different replacement rate. While, in 2008 individuals in different region will also have different replacement rate. According to dummy variables region, 1 present individuals in urban region and 2 present individuals in rural region. The coefficient is 0.016 points out the replacement rate for individuals who live in rural region higher than live in urban. In addition, for example, in 2009 year gender has effects on replacement rates. Depends on the coefficient for gender is -0.20 explaining the replacement rate for male higher than female.

 

Overall, in general, age and level of education are significant during the eight years observation. Age shows a positive relatives with replacement rates and levels of education shows a negative relatives with replacement rates. While, the results indicates region, gender and household composition are inconclusive. Those three variables only significant in few years during the observed time. I can summaryse the general result as the follow table:

 

Variables	Sign
age	+
level of education	_
region	INC
gender	INC
household composition	INC

inc: inconclusive

 

4.4 The replacement rate and the unemployment rate level.

Moreover, discuss the strength of the relationship of unemployment rate and replacement rate in the fitted line and scatter points. The correlation table dominate the correlation between unemployment rate and replacement rate are 0.786. It’s strong positive linear correlation between unemployment rate and replacement rate, tells us somebody who have higher unemployment rate will have chance and higher correlation higher unemployment rate. And statter points shows actual the correlation looks like in the appendix accept the hypothesis one. This is support the hypothesis one in section 2.

 

 

5.Conclusion

This paper have tested two hypotheses: one is replacement rates have positive relationship with the unemployment rate; another one is whether replacement rates differential across different age, gender, region, household composition and level of education. Aimed to analysed the importance of  individual’s financial incentive to work across different age, gender, region, household composition and level of education. For this paper the key methodological is using simulation technique to calculate the counterfactual income to estimate the replacement rates(Engen et al., 1999; Scholz et al., 2004).

 

The results points out the replacement rates have strong positive relationship with the unemployment rate. The strong positive relationship keeps in line with the Standard Microeconomic Theory(Duncan and Giles, 1997), the higher replacement rates measures the lower incentive work, results in higher likelihood unemployment. Also, the coefficient results indicates replacement rates represents vary across different age, gender, region, household composition and level of education. And Adam,S; Brewer M(2005) and Callen, kelly & Savage(2015) provide a supports for replacement rates across different family types. Replacement rates should be vary across family composition, but for the result only two adults without children and three adults without children always explaining the replacement rates. Adam & Brewer & Shephard(2012), Callen & kelly & Savage, (2015) and C.Giles(1995) offering female have higher replacement rates than male. But the results gender only significant two years in observed 8 years. The error may due to when I analyze the family composition ignore separate the gender to test. From exist study indicates female with children have highest replacement rate than all the other family composition. Also, Corde & Findlay(1975) and Josepg(1974) provide exist study supports rural replacement rates higher than urban area. From the observatory 8 years, there have 4 years indicates rural and urban area explaining the replacement rates. While, The results shows only age and levels of education can always explaining the replacement rates. One is Munnell & Soto(2005), Engen(1999) and Scholz(2004) provide a support for age factor explaining the replacement rates. Elder individuals have higher replacement rates and younger individuals have lower replacement rates same as the results. In addition, Stephen Nickell(1997), Daniel Oesch(2010), Richard,F & Ronald(2001) and Richard(2005) support levels of education have effect on replacement rates. The results same as the exist study, replacement rates represents negatives, higher education level would have lower replacement rate.

 

For the results support government to focus on the employees to promote the skills what enterprise needs. For instance, springboard which was first launched as the government’s job. It give unemployment individuals and self-employment individuals a chance to upskill or re skill in areas  where identified the skill shortage in the labour market. Also, the ICT skills conversion programme also provide chance for graduate jobseeker free to participate in the programme. Through the tools to address the lower education individuals unemployment issue.

 

The paper raise age and levels of education should be focused in the future. From the results indicates replacement rates can measure the financial incentive for individuals. Higher replacement rates at lower education level individuals points out lower education individuals have lower incentive to work. Unemployment rates and replacement rates from OECD can test the hypothesis one, but the data from SILC cannot all test the hypotheses two. Gender, region and household composition during the reference time may not be usually linked to the replacement rates. Overall, levels of education and age is worth considering in the future policy publish.

 

 

References

Adam, S., M. Brewer and A. Shephard, 2006. "Financial Work Incentives in Britain:

Comparisons Over Time and Between Family Types", Institute for Fiscal

Studies, Working Paper WP06/20

 

Adsera, A. (2005). Vanishing children: From high unemployment to low fertility in developed countries. American Economic Review,189-193.

 

Andersen, Torben M. and Svarer, Michael, The Role of Workfare in Striking a

Balance between  Incentives and Insurance in the Labour Market (April 2008).

CESifo Working Paper Series No. 2267; University of Aarhus Economics

Working Paper No. 2008-5. Available at SSRN: http://ssrn.com/abstract=1114766

 

Arulampalam, W. (2001). Is Unemployment Really Scarring? Effects of Unemployment Experiences on Wages. Economic Journal, 111(475), pp.585-606.

 

Blanchard, O. and Wolfers, J. (2000). The Role of Shocks and Institutions in the Rise of European Unemployment: the    Aggregate Evidence. Economic Journal, 110(462), pp.1-33.

 

Bradley, D. H., & Stephens, J. D. (2007). Employment Performance in OECD Countries A Test of Neoliberal and Institutionalist Hypotheses. Comparative Political Studies, 40(12), 1486-1510.

 

Breen, R. (2005). Explaining cross-national variation in youth unemployment market and institutional factors. European sociological review, 21(2), 125-134.

 

Baum, S., & Mitchell, W. F. (2010). People, space and place: A multidimensional analysis of unemployment in metropolitan labour markets.Geographical Research, 48(1), 13-23.

Brown, J.R., Sox, H.C. and Goodman, D.C., (2014). Financial incentives to improve quality: skating to the puck or avoiding the penalty box?. JAMA,311(10),1009-1010.

Choon, C.N. and Tsui, A., (2012). Adequacy of Singapore’s Central Provident Fund Payouts: Income Replacement Rates of Entrant Workers. Singapore: Department of Economics, National University of Singapore.

 

Callan, T., Nolan, B. and O'Donoghue, C. (1996). What has happened to replacement rates?. Dublin: Economic and Social Research Institute.

 

Callen, T. Crilly, N. Keane, C. and J.R Walsh (2012). tax, welfare and work incentives.

Whitaker Square, Sir Rogerson's Quay, Dublin 2: The Economic and Social Research

 Institute.

 

Callan, T., C. Keane, M. Savage, J.R. Walsh, K. Timoney, 2012. “Work Incentives:

New Evidence for Ireland” in T. Callan (ed.) Budget Perspectives 2013, Dublin:

The Economic and Social Research Institute.

 

Callan, T., C. Keane, M. Savage and J.R. Walsh,.2013."Taxes on Income: Ireland

in Comparative Perspective",Budget Perspectives 2014, Dublin: The Economic

and Social Research Institute

 

Citizensinformation.ie, (2014)Universal Social Charge (USC). [online] Avarilable at: http://www.citizensinformation.ie/en/money_and_tax/tax/income_tax/universal_social_charge.html#l0b797 [Accessed 16 Nov. 2014].

Crilly, N., Pentecost, A. and Richard, S. (2012). The cost of working in Ireland. Whitaker Square, Sir Rogerson's Quay, Dublin 2: The Economic and Social Research Institute,   pp.4, 10-12.

 

Corden, W. M., & Findlay, R. (1975). Urban unemployment, intersectoral capital mobility and development policy. Economica, 59-78.

 

CSO, 2004-2012. Survey on Income and Living Conditions (SILC) 2014, Dublin: Stationery

             Office

 

Dahl, S. A., Nilsen IV, O. A., & Vaage, K. (2000). Work or retirement? Exit routes for Norwegian elderly. Applied Economics, 32(14), 1865-1876.

           

Discover.ukdataservice.ac.uk, (2012). UK Data Service Discover » Family Resources

 Survey. [online] Available at: http://discover.ukdataservice.ac.uk/series/?sn=200017

 [Accessed 25 Feb. 2015].

 

Discover.ukdataservice.ac.uk, (2012). UK Data Service Discover » Family Expenditure

Survey. [online] Available at: http://discover.ukdataservice.ac.uk/series/?sn=200016

[Accessed 25 Feb. 2015].

 

Forfas, (2013). labour market competitiveness. Dublin.

 

Freeman, R., & Schettkat, R. (2001). Skill compression, wage differentials, and employment: Germany vs the US. Oxford Economic   Papers, 53(3), 582-603.

Ferrarini, T., Nelson, K., Korpi, W. and Palme, J., (2013). Social citizenship rights and social insurance replacement rate validity: pitfalls and possibilities. Journal of European Public Policy, 20(9),1251-1266.

Fulbeck, E.S., (2014). Teacher Mobility and Financial Incentives A Descriptive Analysis of Denver’s ProComp. Educational Evaluation and Policy Analysis, 36(1), 67-82.

Giles, E.L., Robalino, S., McColl, E., Sniehotta, F.F. and Adams, J., (2014). The effectiveness of financial incentives for health behaviour change: systematic review and meta-analysis. PloS one, 9(3), e90347.

Glasziou, P., Buchan, H., Mar, C.D., Doust, J., Harris, M., Knight, R., Scott, A., Scott, I.A. and Stockwell, A., (2012). When financial incentives do more good than harm: a checklist. Bmj, 5047, 345-350.

 

 

Johnson, J. L. (1983). Sex differentials in unemployment rates:A case for no concern. The Journal of Political Economy, 293-303.

 

Giles. C and McCrae. J, TAXBEN(1995).: the IFS microsimulation tax and benefit model -

Institute For Fiscal Studies - IFS. [online] Ifs.org.uk.

Available at: http://www.ifs.org.uk/publications/572 [Accessed 5 Feb. 2015].

Kendal, A.R., Prieto-Alhambra, D., Arden, N.K., Carr, A. and Judge, A., (2016). Mortality rates at 10 years after metal-on-metal hip resurfacing compared with total hip replacement in England. British journal of sports medicine, 50(3), 187-187.

Kostøl, A.R. and Mogstad, M., (2014). How financial incentives induce disability insurance recipients to return to work. The American Economic Review, 104(2), pp.624-655.

Laroque, G. and Salanié, B., (2014). Identifying the response of fertility to financial incentives. Journal of Applied Econometrics, 29(2), pp.314-332.

 

 

Layte, R., & Callan, T. (2001). Unemployment, welfare benefits and the financial incentive to work. Economic and Social Review, 32(2), 103-130.

 

Li, J., & O'Donoghue, C. (2011). Incentives of retirement transition for elderly workers: An analysis of actual and simulated replacement rates in Ireland.

 

Lynch, L. M. (1989). The youth labor market in the eighties: determinants of re-employment probabilities for young men and women. The Review of Economics and Statistics, 37-45.

 

Irish Times, (2015). Welfare versus work: Does it pay to take up a low-paid job?. [online] Available at: http://www.irishtimes.com/news/social-affairs/welfare-versus-work-does-it-pay-to-take-up-a-low-paid-job-1.1944722?page=3 [Accessed 18 Nov. 2014].

 

Irish Times, (2015). Just 3% of dole recipients ‘better off’ without a job. [online] Available at: http://www.irishtimes.com/news/social-affairs/just-3-of-dole-recipients-better-off-without-a-job-1.1568724 [Accessed 18 Nov. 2014].

 

Niemi, B. (1974). The female-male differential in unemployment rates.Industrial and Labor Relations Review, 331-350.

 

Nickell, S. (1997). Unemployment and labor market rigidities: Europe versus North America. The Journal of Economic Perspectives, 55-74.

 

Nickell, S., Nunziata, L., & Ochel, W. (2005). Unemployment in the OECD Since the 1960s. What Do We Know?*. The Economic Journal, 115(500), 1-27.

Promberger, M. and Marteau, T.M., (2013). When do financial incentives reduce intrinsic motivation? comparing behaviors studied in psychological and economic literatures. Health Psychology, 32(9), p.950.

Rajevska, O., (2016). Theoretical Old-Age Pension Benefits and Replacement Rates in the Baltic States: A Retrospective Simulation. Economics and Business, 28(1), pp.13-19.

Savage, M., Callan, T. and Walsh, J.R., (2014). A Profile of Financial Incentives to Work in Ireland. Journal of the Statistical & Social Inquiry Society of Ireland, 44.

Scruggs, L., (2013). Measuring and validating social program replacement rates. Journal of European Public Policy, 20(9), pp.1267-1284.

Shaw, J.D. and Gupta, N., (2015). Let the evidence speak again! Financial incentives are more effective than we thought. Human Resource Management Journal, 25(3), pp.281-293.

Shin, S., Son, H. and Lee, H., (2014). The estimation of lifetime income replacement rates. Journal of the Korean Data and Information Science Society, 25(6), pp.1315-1331.

 

 

Ruane, f. (2013). Message from the Director. [online] Esri.ie.

Available at: http://www.esri.ie/about_us/message_from_the_director/ [Accessed 5 Feb. 2015].

 

Studenmund, A. (2011). Using econometrics. 6th ed. Boston, Mass.: Addison Wesley.

 

Savage, M., Callen, T., Keane, C., Kelly, E. and walsh, J. (2015). Welfare Targeting

and work incentives. Whitaker Square, Sir Rogerson's Quay, Dublin 2: The Economic and Social Research Institute, pp.2-11.

 

Oesch, D. (2010). What explains high unemployment among low-skilled workers? Evidence from 21 OECD countries. European Journal of Industrial Relations, 16(1), 39-55.

 

OECD, 2004-2012. Unemployment Traps 2001-2012, paris, OECD publishing. [Online] Available

             at: http://www.oecd.org/els/benefitsandwagesstatistics.htm

Van Duijn, M., Mastrogiacomo, M., Lindeboom, M. and Lundborg, P., (2013). Expected and actual `replacement rates in the pension system of the Netherlands: how and why do they differ?. Journal of Pension Economics and Finance, 12(02), pp.168-189.

Wang, J. and Van Vliet, O., (2014). Social assistance and minimum income benefits: Benefit levels, replacement rates and policies across 33 countries, 1990-2009. J. Wang and O. van Vliet (2014),'Social assistance and minimum income benefits: Benefit levels, replacement rates and policies across, 33, pp.1990-2009.

Wu, A.Y., Karamcheva, N.S., Munnell, A.H. and Purcell, P.J., (2013). How Do Trends in Women's Labor Force Activity and Marriage Patterns Affect Social Security Replacement Rates?. Social Security Bulletin, 73(4), pp.1-24.

 

 

 

 

 

 

 

 

 

 

 

 

Appendix:

 

Table 1 counterfactual income equation summary in 2005

 

Table 2 counterfactual income equation summary in 2006

 

Table 3 counterfactual income equation summary in 2007

 

Table 4 counterfactual income equation summary in 2008

 

Table 5 counterfactual income equation summary in 2009

 

Table 6 counterfactual income equation summary in 2010

 

Table 7 counterfactual income equation summary in 2011

 

 

Table 8 counterfactual income equation summary in 2012

 

 

Table 9 F test for counterfactual income equation in 2005

 

Table 10 F test for counterfactual income equation in 2006

 

Table 11 F test for counterfactual income equation in 2007

 

Table 12 F test for counterfactual income equation in 2008

 

Table 13 F test for counterfactual income equation in 2009

 

Table 14 F test for counterfactual income equation in 2010

 

Table 15 F test for counterfactual income equation in 2011

 

Table 16 F test for counterfactual income equation in 2012

 

 

 

 

 

 

 

 

 

Table 17 Coefficients for each independent variables for counterfactual income in 2005

 

Table 18 Coefficients for each independent variables for counterfactual income in 2006

 

 

 

 

 

 

 

 

 

Table 19 Coefficients for each independent variables for counterfactual income in 2007

 

 

 

 

 

 

 

Table 20 Coefficients for each independent variables for counterfactual income in 2008

 

 

 

 

 

Table 21 Coefficients for each independent variables for counterfactual income in 2009

 

 

 

 

Table 22 Coefficients for each independent variables for counterfactual income in 2010

 

 

 

 

 

Table 23 Coefficients for each independent variables for counterfactual income in 2011

 

 

 

 

Table 24 Coefficients for each independent variables for counterfactual income in 2012

 

Table 25 coefficient for replacement rates across each explanatory variables in 2005

 

 

 

 

 

 

 

Table 26 coefficient for replacement rates across each explanatory variables in 2006

 

Table 27 coefficient for replacement rates across each explanatory variables in 2007

 

 

 

 

 

 

 

Table 28 coefficient for replacement rates across each explanatory variables in 2008

 

Table 29 coefficient for replacement rates across each explanatory variables in 2009

 

 

 

 

 

 

 

Table 30 coefficient for replacement rates across each explanatory variables in 2010

 

 

Table 31 coefficient for replacement rates across each explanatory variables in 2011

 

 

 

 

 

 

 

 

Table 32 coefficient for replacement rates across each explanatory variables in 2012

 

Table 33 R sqr for  estimated replacement rate equation in 2005

 

Table 34 R sqr for  estimated replacement rate equation in 2006

 

Table 35 R sqr for  estimated replacement rate equation in 2007

 

Table 36 R sqr for  estimated replacement rate equation in 2008

 

Table 37 R sqr for  estimated replacement rate equation in 2009

 

 

Table 38 R sqr for  estimated replacement rate equation in 2010

 

Table 39 R sqr for  estimated replacement rate equation in 2011

 

 

Table 40 R sqr for  estimated replacement rate equation in 2012

 

Table 41 F test for counterfactual income equation in 2005

 

 

Table 42 F test for counterfactual income equation in 2006

 

Table 43 F test for counterfactual income equation in 2007

 

Table 44 F test for counterfactual income equation in 2008

 

Table 45 F test for counterfactual income equation in 2009

 

Table 46 F test for counterfactual income equation in 2010

 

Table 47 F test for counterfactual income equation in 2011

 

 

Table 48 F test for counterfactual income equation in 2012