006 Stankeviciene Gruodis Lokutijevskij Urbaite

8/13/2019 006 Stankeviciene Gruodis Lokutijevskij Urbaite

1/22

BUILDING LI HUANIAN MACRO ECONOME RIC MODEL:FORECAS OF AVERAGE WAGES AND UNEMPLOYMEN RA E

Jelena S ANKEVIIEN, Albertas GRUODIS, Adrian LOKU IJEVSKIJ, la URBAI Vilnius Gediminas echnical University, Faculty o Business Management,

Sauletekio ave. 11, L L-10223 Vilnius, Lietuva

E-mail: [email protected]; [email protected];[email protected]; [email protected]

Abstract.Te importance o economic modelling maintains its positions and plays a rel-evant role in the analysis o behaviour o national economy due to this analysis o macroeco-nomic environment is extremely important. Moreover, analysis o regional or even global trendsis essential to policy makers. Various types o macro-econometric models have been developed

or the analysis o different economic processes and needs. Te main goal o article is to cre-ate Lithuanian macro-econometric model and to provide three orecast projections (scenarios)

o average wages and unemployment rate in Lithuania. Despite the limitations o the use omacroeconomic models during the crisis, it is possible to use such models in scenario analysis.Tis article also provides analysis o scientic literature o average wages and unemploymentrates and statistical data o Lithuanian main economical indicators. Te model is based on theIn orum philosophy, on the input-output accounting principles and identities, integrated bot-tom-up approach. Upon analysis o theoretical and practical aspects conclusions and proposalsare provided.

JEL classication:C53, C67, D57, E01, E27, E37, J39, J69.Keywords: macro-econometric model, average wages (bruto), unemployment rate, inter-

sectoral analysis, input-output models, In orum.Reikminiai odiai: makroekonometriniai modeliai, vidutinis darbo umokestis (bruto),nedarbo lygis, tarpsektorin analiz, itekli-panaudojim modeliai, In orum.

1. Introduction

Following the ongoing worldwide economic crisis it is clearly seen that the predic-tion o such events is a very tough work and requires a lot o knowledge. Integration oLithuania into the European Union (EU) had and will have a lasting inuence on cer-tain structural changes and social trans ormation in both sides. Macroeconomic mod-elling allows evaluation o the macroeconomic environment o a country in a global

ISSN 1822-8011 (print)ISSN 1822-8038 (online)

IN ELEK IN EKONOMIKAIN ELLEC UAL ECONOMICS

2012, Vol. 6, No. 1(13), p. 754775


2/22

755Building Lithuanian Macro-econometric Model: Forecast o Average Wages and Unemployment Rate

context, because changes in economic and political situation in partner countries cansignicantly inuence also economic situation in a small open country (Pos, Ozolia2007), which is also the case o Lithuania.

Furthermore, the economy o Lithuania was one o the astest growing in the

world last decade (19982008) as GDP growth rate was positive 9 years in a row. Whatis more, Lithuania was the last among the Baltic States to be hit by the economic crisisbecause its GDP growth rate in 2008 was still positive. In the third quarter o 2009,compared to the previous quarter, GDP again grew by 6.1% afer ve quarters with neg-ative numbers. Un ortunately, the rebound in Lithuanias economy in the third quarterwas the astest in the EU.

Macro-econometric modelling in the post-Soviet Lithuania became a topical area;there were a lot o discussions o macro-econometric modelling alternatives, specicityo modelling the Lithuanian economy. Moreover, very important questions are raised:

will the created empirically adequate macro-econometric model be use ul in urtherdevelopments o macro-econometric modelling, e.g., the oreign modelling experienceoverview and the analysis o Lithuanian specics were used in Mathematical model oLithuanian economy or orecasting the macroeconomic processes project documenta-tion and reports.

Tere were quite many studies based on Integration o Lithuania into the EuropeanUnion as well. More qualitative aspects o integration were analysed by Kropas (1998)and Vilpiauskas (2001) in their articles, concentrating respectively on possible chang-es in the monetary system and general issues o importance, methods and constraints

o assessment o the integration impact. Quantitative estimates on possible changes otrade ows due to customs union and the common oreign trade policy were providedin the Mikinis and egda (2001), and Ribokas and Vaitkeviius (2001) studies. Teestimates o integration caused migration size and its impact on savings; consumption,etc. were researched in Mustoet al. (2001). Tere were many other studies analysingthe impact o integration (until 2003 there were 35 related research studies initiatedonly by the European Committee under the Government o the Republic o Lithuania).However, even ew o them provided quantitative predictions and concentrated on es-timating the effects on separate sectors. Te straight orward adding-up o estimatedpartial effects could give a very biased total picture due to ignorance or doubling o theimpact. Tere ore there is a need or a general macroeconomic assessment study.

Lithuania has only ew developed macro-econometric models, which are able tomake orecasts: the one which Government is using (those models were created byOlivier Basdevant and August Leppa and it is using ministries o Finance and Economyo the Republic o Lithuania) and others, being used by several bigger banks. It is important to mentioned, that Institute o Economics with the help o others institutionswas building a medium-sized macro-econometric sectoral model o the Lithuanianeconomy called LI MOD. A central element in the model is a 12-sector input-outputtable o the Lithuanian economy acilitating analyses o structural changes.

But these models are restricted and make predictions only or the short or mediumruns. O course, maintaining a macro economic model requires a proper management


3/22

756 Jelena S ANKEVIIEN, Albertas GRUODIS, Adrian LOKU IJEVSKIJ, la URBAI

o huge amounts o in ormation, trained and highly skilled personnel able to deal withcomplex computations and persisting problems, and a good knowledge o the coun-tries macro economy. Tis is one o the main reasons why there are only ew modelsbeing used in Lithuania nowadays and one o the reasons to develop a new one, which

will provide a sophisticated approach to the uture scenarios o the economy, which canbe analyzed aferwards.Tus, macroeconomic models are widely used by international organizations, na-

tional governments and larger corporations, as well as by economics consultants andthink tanks because a macroeconomic model is an analytical tool designed to describethe operation o the economy o a country or a region. Tese models are usually de-signed to examine the dynamics o aggregate quantities such as the total amount ogoods and services produced, total income earned, the level o employment o produc-tive resources, and the level o prices. However macroeconomic models give us the

opportunity to make orecasts o the countries economy, there ore creating scenariosto analyze the uctuations in the near uture. Te main crash test or the macroeco-nomic modelling is a historical simulation, or in other words, a orecast o the pastthings and comparison with already happened. Te closer the historical tting is, thebetter the model was developed and applied, there ore making it suitable or the uture

orecasts.Macroeconomic models could be logical, mathematical, and/or computation-

al; the different types o macroeconomic models serve different purposes and goals.Macroeconomic models may be used to clari y and illustrate basic theoretical princi-

ples, they may be used to test, compare, and quanti y different macroeconomic theo-ries, may be used to produce what i scenarios (usually to evaluate the possible effectso changes in monetary, scal, or other macroeconomic policies), and they may be usedto generate economic orecasts.

None o the macroeconomic model can exist without statistical data. In Lithuaniamore or less reliable and comparable macroeconomic data have been mostly availablesince 1995 due to the change in national accounting system in 1994. Tere ore themain problem or econometric modelling, even when using quarterly data, is the shortdata sample. Another great problem is the transition state o economy. As the economystructure is changing, especially afer some larger exogenous shocks, e.g. due to Globalnancial crisis, the model parameters may be not stable, and even the most importantmodel determinants could change so rapidly that some variables might be signicantonly in certain periods. In the models o transition economies different approachesare used to deal with this problem. In general, a conclusion could be drawn that intransition countries a special emphasis should be laid on testing or possible parameterchanges and non-linearity in preliminarily specied models.

Tis article will be started by analyzing the past and present developments o mac-roeconomic models, dedicated or the creating projections o average wages and unem-ployment rates in other countries or regions.

Te main goal o article is to create Lithuanian macro-econometric model calledLithuanian Macro Model, which will give the opportunity to make orecast or the


4/22


average wages and unemployment rate in the long run, reaching year 2020. Apparentlythe orecast will not be possible without the orecast o the main economic indicatorssuch as GDP, Personal consumption expenditures, Imports, Exports and other indica-tors.

2. Developed Macroeconomic Models for Average Wages and Unemployment

Macroeconomic models have been used or ormulation o economic policy almostin every country o the world. Tese models not only provide an analytical rameworkto link the wages, unemployment rates and other economic indicator sides and theresource allocation process in an economy but also may help in reducing uctuationsand enhancing the economic growth, which are two major aspects o any economy.

Classical, Keynesian, new Classical and new Keynesian approaches have evolved overtime to analyze uctuations o domestic income, employment and price level over years(Keynes (1936), Hicks (1937), Samuelson (1939), Phillips (1958), Friedman (1968),Phelps (1968), obin (1969), Barro and Gordon (1983), Sargent (1986) Goodhart(1989), Nickell (1990), Lockwood Miller and Zhang (1998), IMF (1992)). Empirical validity o these models are tested using either macro-econometric simulations mod-els, applied multi-sectoral general equilibrium models or by using stochastic dynamicgeneral equilibrium models (Wallis (1989), MPC (1999), Pagan and Wickens (1989),Kydland and Prescott (1977)).

From the Neoclassical theory point o view, real wage changes depending onchanges in labour productivity in the long-term. According to the Phillips curve theo-ry, real wages are mainly determined by the unemployment rate or nominal wages de-pending on unemployment and ination. From theories concerning wage negotiations,the bargaining power o employees depend on unemployment and changes in labourproductivity. Tis means that in the long-term, a wage equation is:

(1)

I the Neoclassical theory is dominant, the coefficients1 and2 should equal one

and 3 should equal zero. I the Phillips curve theory is dominant,1 should be insig-nicant. Finally, i wages are not ully compensated or ination,2 should be between0.0 and 1.0.

Wage Dynamics Network macro model investigates the dynamics o aggregatewages and prices in the United States (US) and the Euro Area (EA) with a special ocuson persistence o real wages, wage and price ination. Te analysis is conducted withina structural vector error-correction model, where the structural shocks is identied us-ing the long-run properties o the theoretical model, as well as the co-integrating prop-erties o the estimated system. Overall, in the long run, wage and price ination emerge

as more persistent in the EA than in the US in the ace o import price, unemployment,or permanent productivity shocks.


5/22


Table 1. Macroeconomic models or average wages and unemployment review

No. Macro model Authors Year Main aspects1 Wage Dynamic

Network

European

CentralBank

2009 A research network consisting o econo-

mists rom the European Central Bank andthe national central banks o the EU coun-tries. Te WDN aims at studying in depththe eatures and the sources o wage andlabour cost dynamics ant their implications

or monetary policy.2 U.S.

UnemploymentMovementDynamic Factor

Framework

Heaton,Oslington

2005 Model estimates unemployment betweendifferent economic sectors, sectoral shocksare ound to account or around hal themovement in US unemployment. Shock

requencies, sectoral patterns and owsprovide some clues to the identity o someo the shocks driving unemployment.

Te model consists o a production unction, a wage setting equation, an equa-tion describing price ormation, an equation or the unemployment rate and an equa-tion or the import prices in domestic currency. Te equations contain a minimumo dynamics in order to simpli y the discussion about the long-run properties o themodel.

As regards the wage ormation, we assume that wages are determined througha bargaining process between rms and employees (or labour unions). Tis type omodels predicts that the bargaining solution will depend on the real producer wage andproductivity on the rm side, and on the real consumer wage on the workers side. Asimple log-linear orm o the wage equation corresponding to the bargaining solutioncan be written as:

(2)

where:wthe nominal wage rate;qthe producer price level; pthe consumer pricelevel;uthe unemployment rate.

According to (1), the real wage aced by rms (real producer wage) is affected by(p q), h and u. Te relative price(p q), which measures the difference between theproducer real wage and the consumer real wage, is usually re erred to as the price wedge,and plays an important role in theoretical wage bargaining models. Its coefficient, m,can be interpreted as a measure o real wage resistance, which measures the unionsability to obtain higher wages to compensate or exogenous changes in worker livingstandards (increases in p brought about, or example, by changes in indirect taxes).Te bargaining solution (1) also implies that an increase in labour productivity,h, willincrease wages, since higher productivity increases the protability o rms, makingthem more likely to accept higher wage claims rom the unions. Te unemploymentrate, u, represents the degree o tightness in the labour market, which inuences the


6/22


outcome o the bargaining process through the relative bargaining power o the labourunions and employers organizations.

U.S. Unemployment Movement in a Dynamic Factor Framework model. Te caus-es o unemployment have been a matter o longstanding dispute in economics. Many

different theories o unemployment have been proposed, and disputes over policy attimes have been acrimonious. Effective policy depends on understanding the causeso unemployment movements, and a undamental question is whether these causes aresector-specic or common to all sectors. I sectoral shocks are more important than ag-gregate shocks then it is needed micro models and policy interventions which ocuson the relevant sectors. I not then the ocus should be on aggregate macro modelsand interventions.

Aggregate models o unemployment have been the most popular according sur- veys made by Layard, Nickell and Jackman (2005), but there is no shortages o plausibl

disaggregate models which combine sectoral shocks with slow or incomplete propa-gation. Lucas and Prescotts (1974) seminal paper showed how orthogonal productdemand sectoral shocks and a search across spatially separated markets generate un-employment. Rogerson (1987) developed this urther in a two period, two sectors set-ting, and in Long and Plosser (1983) model have a sectoral shock real business cyclemodel. Ljungqvist and Sargents (1998) inuential turbulence plus skill decay accounto European unemployment is rom this amily o models. Others take a different approach to the shock generating mechanism, with demographic adjustment eaturingin Matsuyama (1992) and in ormational asymmetries in Riordan and Staiger (1993).

Robert Hall (2003; 2005) suggests a number o other possible sectoral shock modelso unemployment. Any general equilibrium trade theory o unemployment (Matusz1996, Oslington 2005, Melitz and Cunat 2006) is a sectoral model.

Some other studies have measured the contribution o sectoral shocks to US macr-oeconomic variables using a variety o methodologies. Long and Plosser (1987) used actor analysis techniques on output or sub-sectors o manu acturing rom 1948 to 1981 tassess the importance o sectoral shocks. Norrbin and Schlagenhau (1988) decomposed1954 to 1980 US output movements into aggregate, sectoral and regional components us-ing the Engle-Watson DYMIMIC actor analysis techniques. Forni and Lippi (1997) andForni and Reichlin (1998) considered very nely disaggregated US manu acturing output

or the period 1958-86 using their own dynamic actor techniques.Lithuanian Macro Model will be based on In orum (InterindustryForecasting

Project at theUniversity oMaryland) philosophy, which was ounded 40 years ago byDr. Clopper Almon, now Pro essor Emeritus o the University. In orum pioneered theconstruction o dynamic, inter-industry, macroeconomic models which portray theeconomy in a unique bottom-up ashion and input-output tables. In orum continuesto oster cooperation and development o economic knowledge and techniques withpartners around the world.


7/22


8/22


Fig. 2. Ination versus wages and salaries uctuations(Lithuanian Department o Statistics database)

In year 2000 ination ell together with wages, reaching almost 1% and 2% all accordingly (Fig. 2). Next year, both indicators rose almost 2 times compared to previous

year, but in year 2002 and 2003 ination growth was negative, as wages growth wasclearly positive. Next our years ahead, ination rose along with growth o wages, tieconomic recession hit Lithuania in 2009, ollowing with decreased both wages andination indicators. Te relationship between these two indicators is more than clear,but the growth levels may lag or one or two years and the size o growth is not propor-tional at all, wages rise or all is much more bigger, reaching two or even more timesbigger uctuations.

Another important economic indicator is unemployment, which is one o themost visible indicators o economic activity. Te rate o unemployment typically risesconsiderably during recessions then alls as the economic recovers.

From economy point o view, labour is another commodity; that is, somethingbought and sold in an open, competitive marketplace. Like any market, the demand orand supply o labour is either in equilibrium or disequilibrium. Te ormer market-clearing state exists when supply exactly equals demand, the latter whenever supplyand demand all out o kilter. Importantly, whenever external circumstances permit, in-ternal orces impel a market in disequilibrium toward equilibrium. echnically, when-ever the equilibrium point alls short o the ull-employment mark, certain amounto what Keynes called involuntary unemployment is inevitable. Even Keynesianismarch critic, the monetarist Milton Friedman, believed there was a natural rate o un-employment (Hall, 1995).


9/22


Fig. 3. Unemployment versus wages and salaries uctuations, compared to previous year

(Lithuanian Department o Statistics database)

In year 2000 unemployment rose by 2.7% compared to previous year, accord-ingly average wages all by almost 2% (Fig. 3). Following next years, annual unem-ployment rate was stably decreasing by approximately 1 to 3%till the year 2008, whenunemployment rate increased by almost 2%. Respectively average wages was increas-ing rom year 2001 till 2008; the increase rate was quite big, reaching more than20% in year 2008. Aferwards the economic recession struck Lithuania, ollowed bydrastically increased unemployment rate, reaching more than 8% and all o the av-erage wages by 5%. Average wages is highly related to the unemployment rate by theinverse relationship, meaning when one indicator is alling the other ollows with theincrease.

Fig. 4.Unemployment rate versus GDP uctuations(Lithuanian Department o Statistics database)


10/22


Matching Lithuanian GDP growth with unemployment growth we see oppositetendency. Te same like in EU, but more uctuation (Fig. 4), since 2000 due to im-proving economic condition, increase o export to Russia (post crisis period in Russia)and neighbour countries, growing domestic demand, increasing o oreign direct in-

vestment GDP is growing and unemployment is moving in inverse direction. As since2000 till 2007 we saw positive tendency (GDP was growing, unemployment rate wasdecreasing) since 2007 we see invertible negative tendency (GDP is decreasing, unem-ployment growing).

Fig. 5. FDI versus unemployment rate uctuations(Lithuanian Department o Statistics database)

Foreign Direct Investment (FDI) is one the most essential economic actors.Especially or the developing economy FDI is one the critical actor. As Lithuanianeconomy is considered as developing, extremely afer integration to EU and Schengenzone, this actor is really essential. We would like to match annual change in% o FDto unemployment rate. Tis analysis will release i FDI have any reasonable inuenceon unemployment rate (Fig. 5).

Obviously there is not any substantive relationship between FDI and unemploy-ment rate according Lithuanian Statistic Department data. Since 1998 till 2008 it wasonly positive tendency in FDI despite this unemployment dispersion was signicant. In2001 it reached peak at highest point o 16.5% afer that lowest peak in 2007 o 4.3%.

One more actor can be added to this evaluation. Since 1990 as Lithuania becameindependent privatization process has started. Due to privatization many state enter-prises were sold to oreigners. So as issue o privatization in FDI statistic it is completo recognize which unds really goes as FDI and which unds were used or buying staenterprises.

For such small country as Lithuania export is the main driver o economy. Asthe domestic market is small, almost all Lithuanian companies consider the possibil-ity to export goods or service, especially afer Lithuanias integration into the EU andSchengen zone.


11/22


Fig. 6. Export versus unemployment rate uctuations(Lithuanian Department o Statistics database)

Analysis shows that in some period there is real relationship o these actors. Asexample since 2003 till 2005 export was growing otherwise unemployment was drop-ping. But matching 20052007 there is opposite tendency as export was decreasingand unemployment was doing either. Tis mismatching could be related to boomingLithuanian economy since 20052008 and very huge domestic householder expenses(nanced by loans mainly rom Scandinavian banks). During this period consumer

index increased drastically. For many companies it was enough to sell their products,service in domestic market without thinking about export. As the result income wasbooming too (Fig. 6).

Analyzing these actors, in 2008, Lithuanian export increased by 20% comparedto 2007. Despite the act that global recession already had been started in the EU andUSA. Analyzing unemployment rate or 2008 we see pretty gures 5.8% comparing to2007 it was 4.3%. I can relate this to strong export in 2008. Due to robust economiccrisis in 2009 Lithuanian export dropped as never down. It decreased by 27%. Fromour point o view, this actor also inuenced unemployment rate.

4. Macro-Econometric Lithuanian Macro Model and its Equations

Afer the analysis o historical data o Lithuanian macro economy, the orecasto Lithuanian average wages and unemployment rate will be provided. Te work willbe done by using In orum philosophy and program called G7, but it will be modi-ed to match the Lithuanian macro economy issues. Tis program was chosen becauseit allows to build a macro-econometric model or Lithuania and to make a orecastwith minimum errors. Te next step will be to make the orecast o Lithuanian averagewages and unemployment rate.


12/22


Creation o a macroeconomic orecast o the Lithuanian economy involves manysteps. Be ore making the orecast o Lithuanian average wages and unemployment ratewe need to create a model or the whole Lithuanian economy, which includes mainmacroeconomic data: GDP, Private Consumption Expenditures and Gross Domestic

Investment, Exports, Imports and other indicators. Te process o creation this modelis explained below.Te model is based on Lithuanian historical macroeconomic data starting rom

the year 2000 till year 2009. First step is creating a Master le or the model. It is a vitapart or the model as it contains all the equations, required to calculate the main macr-oeconomic data, to calculate the behaviour ratios and make the transition o exogenous variables. Te exogenous variables are used by the model but are not determined by it;the endogenous variables are both used in the model and are determined by it.

Te main equation or this model will be the calculation o GDP, which will be an

endogenous variable and will be calculated by the ollowing ormula:GDP = C + V + FE FI + G, (3)

where:GDP Gross domestic product;C Personal consumption expenditures;V Grossprivate domestic investment;FEExports;FI Imports;GGovernment consumptionexpenditures and gross investment.

For the calculation and orecast o average wages it will be added historical valueo labour employment. Te wages equation will be a regression, depending on inationand labour productivity. Labour productivity is calculated by dividing total output by

employment. Next equation is or unemployment orecast:unr = 100 (l c emp)/emp (4)

(where: unrunemployment rate; l clabour orce; empemployment).As explained above, both GDP in real terms and in current prices are endogenous

variables, which are calculated by other endogenous and exogenous variables. Allthe calculations, except average wages, will be used both in real terms and in currentterms.

Personal consumption expenditures are an endogenous variable in this model,which is calculated by the regression which is equal to the change in exports and gov-ernment consumption expenditures and gross investment. Te same procedure will bedone with Personal consumption expenditures in real terms.

Te next endogenous variable, which is regressed in this model, is Gross privatedomestic investment. Te regression or it is equal to Private xed investment, cur-rently in nominal terms. Te same procedure goes with Gross private domestic invest-ment in real terms.

Another variable which is regressed and is required or the calculation o GDP isimports. Te imports are equal to regression with Exports, Gross domestic investment

and Personal consumption expenditures. Te same procedure is with the Imports re-gression in real terms.


13/22


Since there is no possibility to orecast the Exports o the country, unless a multi-national model has been built, the Exports will be taken as a projection. Te same willbe done or Government consumption expenditures and gross investment, as this is

ully enough or the current phase o the created model.

Also GDP Deator is introduced, which is endogenous indicator in this model.Te main purpose o it is to calculate the dynamics o average wages and unemploy-ment rate.

Now all the mentioned above can be summarized in a single Master ( able 2) le.

Table 2. Te cong Master le or G7

Master# Lithuanian Macro Model Master le# 7 exogenous (not counting time)# 3 behavioural ratios or eR, lpcheckdup y

ex time = time# Personal consumption expenditures equationadd cR.sav r c = e, gadd c.sav r cR = eR, gR # Gross private domestic investment equationadd vR.sav r vR = v R add v.sav r v = v # Imports equationadd R.`sav r R = eR, vR, cR add .sav r = e, v, c# GDP Real

gdpR=cR+vR+ eR-R+gR # GDP Nominal

gdp = c + v + e - + g# GDP deator

gdpD=gdp/gdpR # Behaviour relationship or exports

ex eBR = eR/(cR+vR-R) eR = eBR*(cR+vR-R) e = eR*gdpD

# Labour productivity behaviour ratioex lpBR = gdpR/emp emp = gdpR/lpBR


14/22


# Calculation o average wagesadd awag.sav check gdpR 1.0# Labour Force behavioural ratio

ex l cBR = l c/tpop l c = l cBR * tpop# Calculation o Unemployment Rate

unr = 100. * (l c-emp)/empcheck gdpR 1.0end

Lines that begin with# are just comments and are ignored by the computer. Linesthat begin with an orm the variable on the lef by the expression on the right. Teend command signals the end o the master le or the model building program. Linesbelow it do not go into the model but have another unction.

Every equation in Master le ( able 2) is an identity, so this approach has earnedits label as identity-centred modelling. Essentially, identities to replace variables thatare hard to think about are replaced by others that are easier to grasp intuitively. Gooduse o identities is essential or good modelling. Tis recognition o the central role oidentities in modelling is in stark contrast to the way that they are usually dismissed ineconometric texts with the comment that an identity can be used to eliminate a vari-able. O course it can, and then one is lef with a variable that is hard to think aboutwithout the handlethe identitythat gives an easy way to think about it.

5. Output of Lithuanian Macro Model

Te Lithuanian Macro Model have been success ully built and run, providing withthe orecast o average wages and unemployment rate in Lithuania. Now ollows steby step explanation what is the output o the model or Lithuanian average wages andunemployment rate.

Moderate scenario. Analyzing the historical simulation o average wages in moder-

ate scenario (Fig. 7) it can be seen that the reproduction really t, only the year 2009was orecasted to decrease a little bit more, than it was actually. From the orecast it canbe stated that the average wages will decrease in year 2010 to the value o 1939.92 L Land aferwards will stabilize in year 2011 and will start to increase urther. Te rise atthe beginning will be quite slow, as the economy is recovering rom the worldwide cri-sis, also the employers will not be rushing with the increase o wages or their employees, but in year 2015 we can see a signicant increase to 2107.39 L LL (bruto), whichmay lead to the act that Lithuanian economy may ully recover rom the recession aneconomically will be strong to introduce EURO as a main currency. Aferwards the

increase will stabilize again and will ollow the trend o the ination, reaching averagwages o 2469.2 L LL (bruto) in year 2020.


15/22


Fig. 7. Forecast projection o average wages (bruto), L LL

Pessimistic scenario. In a pessimistic scenario the labour productivity was decreasedby 5% compared to the moderate scenario. Te decrease o 5% inLabour productivity had a signicant impact on the average wages trend (Fig. 7). Te orecast o the year2010 shows that the level o average wages will be slightly higher than in the year 2007,aferwards ollowing a slow increase. Te breaking point will remain the samethe

year 2015as in the moderate scenario, leading to conclude that 2015 will be a ullrecovery o the recession, leading to a aster urther increase. In year 2018 the level oaverage wages will overcome level in best economic times o 2008 and will continueto increase reaching average wage o 2350.10 L LL (bruto) in year 2020. So to sum upthe decrease o labour productivity leads to a signicant decrease in average wages, icompared to numerical expression, the decrease o 5% in labour productivity has led toa decrease o 4.8% in average wages.

Optimistic scenario. Labour productivity increased by 5% compared to moderatescenario. So the increase in 5% o labour productivity resulted in signicant inuence

on the average wages (Fig. 7). Starting rom the orecast o year 2010, the level o aver-age wages slowly increases till the year 2014, and aferwards ollows with even biggerincrease rom year 2015. Te optimistic scenario shows that the level o year 2014 willreach the level which was in the year o the highest average wages in 2008 and the nextyear will ollow with even higher increase leading to a biggest level o average wagesthan it was ever be ore. Te average wages orecast will reach a value o 2588.30 L LL(bruto) in the year 2020, which states that a high ination may set in or Lithuania aver-age living standards will increase signicantly during the 10 years period rom now. oconclude, the increase in 5% o labour productivity in optimistic scenario resulted in a very signicant increase on the level o average wages (bruto), and in numerical expres-sion led to 4.8% increase in total or the analyzed period. Meaning that average wagesdepends mostly on labour productivity ratio and partly on the ination.


16/22


Fig. 8. Forecast o unemployment rate, %

Moderate Scenario. For unemployment rate historical simulation and orecast projection (Fig. 8), we are using the gures o executed projections o employment andlabour orce. Having this data, we can run unemployment equation. Te discrepancyin historical simulation is not substantial. In addition, according to the macro modelthe unemployment rate in 2010 will reach the peak; aferward start to decrease reach-ing 7.36% in 2020.

Pessimistic Scenario. Analyzing the data we can see substantial growth o unem-ployment comparing to moderate scenario. For example, in moderate scenario in 2020,unemployment rate is 7.36% and in pessimistic scenario, it reached 8.83%. Comparingpessimistic orecast with historical data it can be said what the level o 2005 unemployment rate o pessimistic scenario will reach in 2020. So it is determined what this isreally pessimistic scenario or Lithuanian economy. We can explain this by a decreasein export that has caused a decrease in GDP and other economic actors. Obviouslynot only export made such an inuence on unemployment rate. By changing exportbehavioural ratio all variables in this ratio has changed too. Labour orce remains thesame as in the moderate scenario.

Optimistic Scenario. Tis scenario provides nice numerical as we see on gure 42o unemployment rate. Comparing to moderate it is decrease by 1.4% in 2020. As pes-simistic scenario provides increase in unemployment by 1.5%. In such an economicprojection, we are getting a real booming economy. However, this unemployment rateis not something impossible or unrealistic. Going back to historical data, we see thatsuch gures Lithuania already has had in 2007 when unemployment rate was evenlower 4.3%, and in 2008 it was quite the same 5.8%.


17/22


6. Further Development of Lithuanian macroeconomic model

Te main goal o urther development was development and improvement oLithuanian macroeconomic model using G7 program. At current stage it is made quar-

terly historical simulation and orecast o main economic variables. Model is orecastedunemployment rate and average wages by quarter till 2020 year.Te underground and basis or urther development is QUES Quarterly eco-

nomic structural model o the US economy.

Fig. 9. Forecast o quarterly unemployment rate, L LL

By analyzing two lines it is obviously seen quite close tting o statistical data andorecasted projection. O course there is some discrepancy, however in general there

is close tting. Te lowest rate o unemployment was in period since 2005 till 2008.Aferward, due to a worldwide economic crisis unemployment rate started to grow. Onlysince 2010s second quarter did it start to decrease. Observing the orecast we see a uc-tuation o unemployment rate. Since 2012 a small decrease o unemployment rate is seen;however in 2013 -2014 a very slow decrease or even increase in unemployment rate isseen. Aferward there is substantial decrease till second quarter o 2016. And since 2016it is seen increase o unemployment rate and some uctuation with decrease trend.

Te highest unemployment rate, according to statistical data, in 2010 second quarter,was 18.3%. According to the orecast in 2010 second quarter, unemployment rate was12.98% it is seen some discrepancy. However already in ourth quarter o 2010 it is al-most the same numbers in orecast it is 16.71% and statistical data provided 17.10% closetting. Te orecast provided in 2020 ourth quarter 11.63% unemployment rate similarlevel was in 2004. As is seen, the orecast does not provide an optimistic scenario. It isquite moderate, taking into account the unstable situation in worldwide economy, such ascenario could be real and should be estimated as one o the possibilities.


18/22


19/22


None o the orecast is available without proper analysis o countries macroeconomicindicators. Also correct statistical data have to be collected and correctly understood.

Short review o model creation process is provided. Also, article describes modeloutcomes o three various scenarios o average wages and unemployment ratemod-

erate, pessimistic and optimistic.Te orecast have concluded that the average wages uctuate dependently with thelabour productivity ratio, which is an important actor o every country.

Te level o average wages which were seen in Lithuania in the year 2008 will beapproximately reached in the year 2012-2014depending on the urther behaviour othe economy.

Te orecast o unemployment is interrelated to productivity ratio, export, labourorce; these actors are crucial or the unemployment orecast in this model.

Te level o unemployment will reach the level be ore the recession in the year

20182020.Te orecast shows that the economy o Lithuania will recover in the year 2016,leading to a quite rapid increase o average wages and a decrease in unemploymentrate.

Regarding the orecast, the peak o unemployment will be in 2010; aferward, in all3 scenarios, the unemployment will go down. Te overall economic climate will changetowards a positive tendency.

References 1. Bhattarai, K. R. (2005).Keynesian Models or Analysis o Macroeconomic Policy , p. 26. 2. Celov, D.; Vilkas, E.; Grinderslev, D.; Andersen, F. M. (2005). Macro-econometric model

or Lithuania. In:Economic Modelling22(4), July: 707719. 3. Clopper, A. (1999).Real Effects o a Fall in the Stock Market.USA. Maryland. 4. Clopper, A. (2008).Te Craf o Economic Modelling . Fifh edition. Maryland. 5. CSB Database. (2009). [interactive]. [accessed 10 January 2012], . 6. Database o the Bank o Latvia. (2009). [interactive]. [accessed 21 February 2012], .

7. Duffy, F. (2009).Unemployment & Underemployment. New York. 8. EBRD. (2009).Lithuanian economic overview. 9. Estrada, A.; Fernandez, J. L.; Moral, E.; Regil, A. V. 2004. A Quarterly macroeconometric

model o the Spanish economy.Banco de Espana Documentos de rabajo [Working Paperso the Bank o Spain]: 0413.

10. European Commission Eurostat. [interactive]. [accessed 10 January 2012], .

11. European Commission. (2009).Labour Market. Europe in guresEurostat yearbook 2009.Luxembourg: Office or Official Publications o the European Communities, ISBN 978-92-79-11625-4, pp. 265288.

12. Eurostat Database 2011. (2011). [interactive]. [accessed 21 February 2012], .


20/22


13. Friedman M. (1970).Te Counter-Revolution in Monetary Teory .14. Gali, J. (2009).Te Return o the Wage Philips Curve.15. Hicks, J. R. (1937). Mr. Keynes and the Classics.16. Lehmus, M. (2007).Empirical Macroeconomic Model o the Finnish Economy (EMMA).

Discussion Papers o Labour Institute or Economic Research. Helsinki: Labour Instituteor Economic.

17. Lbermanis, G. 2003. Starptautiskie ekonomiskie sakari un Latvija [International economicrelations and Latvia]. Riga: Kamene.

18. Lithuanian Ministry o Finance. (2010).Convergence Programme o Lithuania o 2009 tothe European Commission.Vilnius.

19. Lithuanian Statistic department. (2012).Statistics pre-dened tables, Statistics Databases[interactive]. [accessed 10 January 2012], .

20. Lundberg, J. (2009).Te classical macroeconomic model .21. Mankiw, N. G.; Romer, D. (1991).New Keynesian Economics.22. McQuinn, K.; ODonnell, N.; Ryan, M. (2005). Macro-econometric model or Ireland.

In: Research echnical Paper o Central Bank and Financial Services Authority o Ireland December: 146172.

23. Merikll, J. (2004). Macroeconometric modelling o the Estonian economy. Modelling theeconomies o the Baltic Sea region. artu: artu University Press.

24. Ozolina, V. (2006).Forecast o Population And Employment In Latvian Macro- And Multisectoral Models. Latvia. Riga.

25. Pocs, R.; Auzina, A. (2007). Base-scenario orecasts by Latvian INFORUM model: resuland problems. Latvia. Riga.

26. Pos, R.; Ozolia, V. (2007). Makroekonomisko procesu modelana[Macroeconomic mod-elling]. Riga: R U Izdevniecba, Latvia.

27. Russian Federal State Statistics Service database. 2009. [interactive]. [accessed 21 February2012], .

28. Saeed, K.; Radzicki, M. J. (1993). A Post Keynesian Model o Macroeconomic Growth,Instability, and Income Distribution.

29. Sakiene, H. (2009).Unemployment Regulation Policy Analysis in Lithuania. Klaipda,pp. 270277.

30. University o Maryland. Economic Department. (2009).Exploring G7.USA. Maryland31. University o Maryland. Economic Department. (2009).Inter-industry Forecasting Project.

USA. Maryland.32. University o Maryland. Economic Department. (2009).Te Build program. USA.

Maryland.33. University o Maryland. Economic Department. (2009).Te G7 user guide and re erence

or G7.3.USA. Maryland.34. Weyerstrass, K.; Neck, R. (2007). SLOPOL6: Macroeconomic model or Slovenia

International Business & Economics Research Journal 6(11): 8194.35. World Economic Outlook. Database o International Monetary Fund. (2009). [interactive].

[accessed 10 January 2012], .36. Zle, H. (2003).Latvijas rjie ekonomiskie sakari [Latvian external economic relations].

Riga, uriba: Biznesa augstskola press.37. : 20082012 . ( ). 2009. [Prognoz

indikatorov ekonomiki RF: 20082012 gg. (bazovyj scenarij)] [interactive]. [accessed 21February 2012], .


21/22


LIETUVOS MAKROEKONOMETRINIO MODELIO KRIMAS:VIDUTINIO DARBO UMOKESIO IR NEDARBO LYGIO PROGNOZ

Jelena S ANKEVIIEN, Albertas GRUODIS, Adrian LOKU IJEVSKIJ, la URBAI

Santrauka. Ekonominio modeliavimo svarba ilaiko savo pozicijas ir vaidina svarb vai-dmen tiriant nacionalins ekonomikos pokyius ir makroekonomines tendencijas. Regioniniaiir globaliniai makroekonomini rodikli pokyiai takoja priimamus ekonomins politikossprendimus. oki sprendim primim palengvina vairs makroekonometriniai modeliaiskirti ekonomini proces analizei. Straipsnio tikslas sukurti Lietuvos makro-ekonometri-n model ir pateikti tris scenarijus vidutinio darbo umokesio ir nedarbo lygio prognozsLietuvoje. yrimo rezultatai parod, kad nepaisant makroekonomini modeli naudojimo ap-ribojim krizs metu, galima naudoti tokius modelius vairi makroekonomini scenarij ana-lizei. Straipsnyje analizuojami vidutinio darbo umokesio ir nedarbo lygio kitimo tendencijos

Lietuvoje, kartu su pagrindiniais makroekonominiais rodikliais. Silomas makroekonominismodelis grindiamas In orum losoja, itekli-panaudojim analizs principais ir tapatumu,integruotu i apaios vir metodu. Ianalizavus teorinius ir praktinius aspektus, pateikia-mos ivados ir pasilymai.

Jelena Stankeviien. Assoc. Pro ., Ph.D. at the Department o Finance Engineering,Vilnius Gediminas echnical University (Lithuania), the Dean o the Faculty o BusinessManagement. Research interests: assets and liability management, regulation o nancial in-stitution, nancial management o value creation, value engineering, inter-sectoral analyses,input-output models.

Jelena Stankeviien. Vilniaus Gedimino technikos universiteto, Finans ininerijos ka-tedros docent, socialini (ekonomika 04S) moksl daktar. Mokslini interes sritys: turto irsipareigojim valdymas, nans valdymas kuriant vert, verts ininerija, tarpakinio balansoanaliz.

Albertas Gruodis. MSc in Business rom Vilnius Gediminas echnical University(Lithuania), the Faculty o Business Management. Research interests: macroeconomic model-ling, inter-sectoral analyses, input-output models.

Albertas Gruodis.Vilniaus Gedimino technikos universiteto, Verslo vadybos akultetomagistras. Mokslini interes sritys: makroekonominis modeliavimas, tarpakinio balanso ty-rimai, itekli-panaudojim modeliai.

Adrian Lokutijevskij.MSc in Business rom Vilnius Gediminas echnical University(Lithuania), the Faculty o Business Management. Research interests: macroeconomic model-ling, intersectoral analyses, input-output models.

Adrian Lokutijevskij. Vilniaus Gedimino technikos universiteto, Verslo vadybos akultetomagistras. Mokslini interes sritys: makroekonominis modeliavimas, tarpakinio balanso ty-rimai, itekli-panaudojim modeliai.


22/22


la Urbait. PhD (economics) student at Vilnius Gediminas echnical University(Lithuania), the Faculty o Business Management. Research interests macroeconomic model-ling, intersectoral analyses, input-output models.

la Urbait. Vilniaus Gedimino technikos universiteto, Verslo vadybos akulteto dokto-rant. Mokslini interes sritys: makroekonominis modeliavimas, tarpakinio balanso tyrimai,itekli-panaudojim modeliai.

Documents

006 Stankeviciene Gruodis Lokutijevskij Urbaite