Hsin (1988) is the first attempt in analyzing this issue in Taiwan. He uses the yearly macro data from 1953 to 1985, with the wage rate of private sector as the dependent variable, and the wage rate of public sector, the change rate of wholesale price index, the growth rate of labor productivity of private sector, and unemployment rate as the independent variables. He finds that the effect of public sector wage on private sector wage is positive and statistically significant, with the coefficient of the public sector wage variable being 0.1703.
Huang (1990) applies the methods of residual analysis and impulse response simulation on the yearly industrial data from 1959 to 1988. He finds that public sector wages do affect private sector wages for 11 in 16 industries, and the extent of this spillover effect ranges from 0.3-0.7.
Contrary to Hsin (1988) and Huang (1990), Lin (1992) finds that the pulling effect is not statistically significant both in the service and manufacture sectors. He analyses the relationship between public sector wages, prices and private sector wages under a complete macro model, then test the pulling effect separately on service and manufacture sector wages with the data cover from 1973 to 1991. After controlling the price, productivity, unemployment rate and the lagged term of wage, he finds that the effect is insignificant in the yearly sample models and slightly significant in the seasonal sample models, but dismiss the latter as the poor quality of seasonal data.
In a macro-econometric model with unemployment gap, Wu et al. (2002) finds that pulling effect is statistically insignificant over the data till 2001. On the other hand, Wu (1995) studies the data of period 1980-1994 and finds that it is the private sector wages affect the public sector wages.
Aside from the case of Taiwan, the main group of literature on this issue is those
8
researches which are evolved from the Scandinavian model of inflation and mainly focus around the case of Scandinavian economies (Lindquist and Vilhelmsson, 2006).
In its analysis of the interactions between domestic and international macroeconomic factors, Scandinavian model predicts that the tradable goods sector would act as the wage leader, i.e. wage increases occurred in the tradable goods sector first, and then transmitted to the non-tradable goods sectors (Lindbeck, 1979; Söderström and Viotti, 1979). Since Scandinavian countries often have small, open economies with highly unionized labor forces and relatively large public sectors, the linkages between public (non-tradable) sector and private (tradable) sector have been an issue for positive studies. Besides, in some Scandinavian countries, this model also serves as a normative model to set wages in the non-tradable public sectors, without placing
undue upward pressures on the wage formation in the private tradable sectors to harm their international competitiveness (Lindquist and Vilhelmsson, 2006).
Holmlund and Ohlsson (1992) find that while the public sector wages were always higher than private sector, from around 20% in late 1960s to 5% in early 1990s, the private sector wage changes Granger causes wage changes in central as well as local governments, which is consistent with the Scandinavian model. In fact, most studies within this context had also confirmed the wage leadership of private sector, such as Jacobson and Ohlsson (1994) and Lindquist and Vilhelmsson (2006).
As an exception, Friberg (2007) finds no evidence of a unique wage leading role for the internationally exposed manufacturing sector.
By focusing on the wage agreements of public and private sector in Canada during the 1967-1978 period, Lacroix and Dussault (1984) analyzes the possible spillover effects of the public sector over private sector, and finds that the
public-private spillover effect occurs only under the two following conditions: First, the corresponding public sector employees are white- or blue-color workers, rather
9
than teachers, nurses, fireman, policemen; Second, they are located in the same urban area as the private sector workers. Besides, the magnitude of the spillover effect is affected positively by the degree of tightness in the labor market, and negatively by the size of the urban area where the private sector workers are located and by the degree of foreign competition faced by their employer.
Mizala and Romaguera (1995) argue for the importance of institutional factors in the leading role of public sector wages in Chile. Since Chile has had a long history of state interventions in labor market, especially since the military coup in 1973, they find that the influence of public sector wages on private sector wages was quite strong before 1979, then significantly reduced with the deregulation process during
1979-1982, and finally rejected in the period of 1983-1990, in which period the state had no longer controlled the labor market.
Demekas and Kontolemis (1999) focused on the case of Greece from 1971 to 1993, and find that the increases in public sector wages did lead to increases in private sector wages, and the increases in public sector employment had no significant impact on unemployment. On the other hand, Christou (2007) find that for Romania the private sector leads in the whole sample period, and obtained a bi-directional causality from 1998 to 2006.
In the recent years, the European Central Bank (ECB) conducts a series of studies on the interaction between the public and private sector wages. Afonso and Gomes (2008) analyzes a panel data of OECD and EC country groups covering essentially the period from 1970 to 2007. They find that the effects of public sector wages and employment on private sector wages are statistically significant and positive. A 1% increase in real public sector wage growth increases private sector nominal wage growth by 0.3%, and a 1% increase in public sector employment increases private sector wage growth by close to 0.3%
10
Another ECB paper, Lamo et al. (2008), focuses the co-movement and Granger-causality between public and private sector wages in OECD and EURO countries from 1960-2006. They find that private sector wages exert mostly a stronger influence on public sector wages than the other way around, while the public sector wages exerted influence back via prices both directly and indirectly in some cases.
Pérez and Sánchez (2010) analyze the contemporaneous intra-annual causality between public and private wages for the four biggest European economies, i.e.
Germany, Spain, France and Italy, by using the quarterly data from 1980 to 2007.
They find that while there were significant linkages between public and private wages, the direction of wage leadership was depended on the choosing of countries, sample periods, and whether only the within-the year linkages were considered.
D’Adamo (2010) extends the analysis to the Central - Eastern European countries. He divides the economy into three sectors, i.e. tradable sector, market non-tradable sector and public sector, and constructed three scenarios labeled as
‘Scandinavian model’, ‘wage mark-up model’, and ‘envy-effect model,’ in which the three sectors acted as the wage leaders sequentially. By exploiting quarterly data from 2000Q1 to 2010Q2, he finds that the Scandinavian model is a good approximation only for Estonia and Slovak Republic, while the public sector wages lead in Romania, Czech Republic and Latvia. However, for remaining countries are concerned,
non-traded or public sector is not affected by the tradable sector.
Table 2.1: Summary of econometric studies
Paper Countries/Period Main findings
Hsin (1988) Taiwan, 1953-1985 Public sector pull up private sector wages
Huang (1990) Taiwan, 1959-1988 Public sector pull up private sector wages in 11 of 16 sectors
11
Lin (1992) Taiwan, 1973-1991 No pulling effect
Wu (1995) Taiwan, 1980-1994 Private sector pull public sector Wu et al. (2002) Taiwan, -2001 No pulling effect
Holmlund and
Ohlsson (1992) Sweden, 1965-1991 Private wage changes Granger caused public sector wage changes
Jacobson and
Ohlsson (1994) Sweden, 1968-1988 Private sector leads Lindquist and
Vilhelmss (2006)
Sweden, 1970-2002 Private sector leads
Friberg (2007) Sweden, 1980-2002 Central government sector leads D’Adamo
(2010)
10 Central and Eastern European Countries, 2000-2010
Tradable (private) sector leads only in Estonia and Slovak Republic
Lacroix and
Dussault (1984) Canada, 1967-1978 Public sector leads only in specific conditions
Mizala and Romaguera
(1995)
Chile, 1976-1990
Influence of public sector on private sector is strong before 1979 and fades away over time
Demekas and Kontolemis
(1999)
Greece, 1971-1993 Public sector leads
Christou (2007) Romania, 1993-2006
Private sector leads in whole sample period, and bi-directional causality in 1998-2006.
Afonso and Gomes (2008)
OECD and EU countries, 1970-2007
A 1% increase in public real wage/employment growth increases private nominal wage growth by 0.3%.
Lamo et al.
(2008)
OECD and EURO area countries, 1960-2006
1. private public closely correlated 2. Private sector lead
3. Public sector indirectly influence private sector via price
Public sector wages has leading role only in the within-the year linkages and this role have faded in the past two decades at least for Germany and Spain
12