10751
Views
413
Downloads |
Do increases in public sector wages affect inflation?
Ozana Nadoveza* 
Ozana Nadoveza
Affiliation: Croatian National Bank, Zagreb, Croatia ; University of Zagreb, Faculty of Economics and Business, Zagreb, Croatia
 0000-0002-3651-7795
Correspondence
onadoveza@efzg.hr
Article | Year: 2025 | Pages: 1 - 44 | Volume: 49 | Issue: 1 Received: June 1, 2024 | Accepted: October 27, 2024 | Published online: March 10, 2025
|
FULL ARTICLE
FIGURES & DATA
REFERENCES
CROSSMARK POLICY
METRICS
LICENCING
PDF
Note: Prices are measured by the Harmonized Index of Consumer Prices (HICP). Wages aremeasured by compensation of employees per employee from national accounts statistics, productivity
is calculated as the ratio of real GDP to the number of employees from national accounts. Source: Eurostat.
|
Shocks/Variables
|
Demand
|
Supply
|
Labour supply
|
Wage mark-up
|
Other
|
|
GVA (GDP)
|
+
|
+
|
-
|
-
|
?
|
|
HICP
|
+
|
-
|
+
|
+
|
?
|
|
W
|
+
|
+
|
+
|
+
|
?
|
|
PROD
|
+
|
+
|
? (-)
|
? (+)
|
?
|
|
EMP (UR)
|
+ (-)
|
+ (-)
|
- (-)
|
- (+)
|
?
|
Note: The historical decomposition, used to calculate the relative importance of individual shocks
in wage dynamics (see formulas A3 and A4 in appendix), is presented in appendix (graph A24)
for both the public sector and the overall economy. Source: Author.
Note: The total effect was calculated using the formula: HIPCttotal = HIPCtdirect + HIPCtindirect, where the total potential indirect effect is given by: HIPCtindirect = αADHIPCtADindirect + αWHIPCtW_indirect. This was calculated under different, mutually exclusive assumptions related to αAD. Source: Author.
Note: The graph displays the unconditional correlation coefficients between public sector wages in period t and private sector wages in period t-l, where l represents a quarter and ranges from l = [-12, 12], corresponding to ±3 years. Source: CBS, author.
|
Variable
|
Data
|
Source
|
|
Wages
|
Average wages based on regular
monthly surveys of net and gross wages, classified according to the National
Classification of Activities (NKD)
|
Croatian Bureau of Statistics (CBS),
quarterly average, seasonally adjusted using X-12-ARIMA
|
|
Public sector wages
|
Average wages for sectors O, P, and Q
based on regular monthly surveys of net and gross wages, classified according
to the National Classification of Activities (NKD)
|
CBS, quarterly average, seasonally
adjusted using X-12-ARIMA
|
|
HICP
|
Monthly Harmonized Index of Consumer
Prices, seasonally adjusted using X-12-ARIMA, quarterly average
|
Eurostat, [prc_hicp_midx]
|
|
Public sector services HICP
|
See table A2 and graph A1
|
Eurostat, Cai and Vandyck (2020)
|
|
Gross domestic product (GDP)
|
Gross domestic product, constant
prices, seasonally and calendar adjusted
|
Eurostat, [namq_10_a10]
|
|
Gross value added (GVA)
|
Gross value added, constant prices,
seasonally and calendar adjusted
|
Eurostat, [namq_10_a10]
|
|
Gross value added (GVA) of public
sector
|
Gross value added (GVA) of O, P, and
Q sectors, constant prices, seasonally and calendar adjusted
|
Eurostat, [namq_10_a10]
|
|
Employment
|
Employed individuals based on regular
monthly surveys of employment by activity
|
CBS, seasonally adjusted using
X-12-ARIMA
|
|
Employment in public sector
|
Employed individuals in of O, P, and
Q sectors based on regular monthly surveys of employment by activity
|
CBS, seasonally adjusted using
X-12-ARIMA
|
|
Unemployment rate
|
Unemployment rate, ages 15-74,
percentage of the labour force, trend-cycle data
|
Eurostat,
[une_rt_q_h]
|
|
Productivity
|
GDP/employment, GVA/employment
|
Eurostat, [namq_10_a10] i CBS
|
|
Productivity in public sector
|
GVA of O-Q sectors/employment of O-Q
sectors
|
Eurostat, [namq_10_a10] i CBS
|
|
Million
euro, current prices
|
Out-patient
services (CP062)
|
Hospital
services (CP063)
|
|
Human health services (CPA_Q86)
|
386.5
|
80.4
|
Source: Eurostat, Cai and Vandyck (2020), author's calculation.
Source:  Author.
Graph 1Wages, productivity and inflation DISPLAY Graph
Graph 2Annual wage growth and HICP inflation in the public sector and the overall economy (%) DISPLAY Graph
Table 1Identification restrictions DISPLAY Table
Graph 3Shock dependent cumulative wage-to-price multipliers DISPLAY Graph
Graph 4Relative importance of shocks in wage dynamics (2-year moving average) DISPLAY Graph
Graph 5Time-varying correlation between wage growth and inflation (2-year moving average) DISPLAY Graph
Graph 6Cumulative response of public sector services HICP to a one-standard-deviation wage mark-up shock in public sector DISPLAY Graph
Graph 7Cumulative response of wages to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph 8Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph 9Range of estimated effects across different economy-wide model specifications (pp) DISPLAY Graph
Graph 10Unconditional correlation between public and private sector wage growth DISPLAY Graph
Table A1Data and data sources, 2004-2023 DISPLAY Table
Table A2Part of the bridging matrices related to the human health services DISPLAY Table
Graph A1HICP for human health services and associated subcomponents DISPLAY Graph
Graph A2Impulse response functions DISPLAY Graph
Graph A3Impulse response functions DISPLAY Graph
Graph A4Impulse response functions DISPLAY Graph
Graph A5Shock dependent wage-to-price multipliers DISPLAY Graph
Graph A6Cumulative response of public sector prices to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph A7Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph A8Impulse response functions DISPLAY Graph
Graph A9Shock dependent wage-to-price multipliers DISPLAY Graph
Graph A10Cumulative response of public sector prices to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph A11Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph A12Impulse response functions DISPLAY Graph
Graph A13Shock-dependent wage-to-price multipliers DISPLAY Graph
Graph A14Cumulative response of public sector prices to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph A15Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph A16Graph A16 Impulse response functions DISPLAY Graph
Graph A17Shock-dependent wage-to-price multipliers DISPLAY Graph
Graph A18Cumulative response of public sector prices to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph A19Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph A20Impulse response functions DISPLAY Graph
Graph A21Shock-dependent wage-to-price multipliers DISPLAY Graph
Graph A22Cumulative response of public sector prices to a one-standard-deviation wage mark-up shock DISPLAY Graph
Graph A23Estimated potential effect of public sector wage growth on headline inflation in 2024 (pp) DISPLAY Graph
Graph A24Historical decomposition of nominal wages DISPLAY Graph
* The author would like to thank two anonymous reviewers for their valuable comments, suggestions, and
insights, which contributed to shaping this paper. Special thanks to Davor Kunovac for his ideas behind the
methodology for calculating the effects of total wage growth on inflation, which served as the foundation
for the approach used in this paper to estimate the impact of public sector wage growth on overall inflation.
Additionally, the author acknowledges the use of ChatGPT (version 4.0 mini) to enhance the translation,
readability, and overall language quality of the paper. The views and opinions expressed in this article are
those of the author and do not necessarily reflect the official policy or position of the Croatian National Bank.
The article was judged the best regular article in the 2024 annual competition of the Hanžeković Foundation.
1 For instance Index argued that in 2022 labour costs (nominal wages) were increasing but that simultaneously
real wages were decreasing (see Perković, 2022). From the context of the argumentation it is clearly implied
that substantial nominal wage growth at the time (7.5% yearly) was not enough to offset inflation, the suggestion
being that more nominal wage growth is required.
2 During the periods of elevated inflation, particularly in the first half of 2023, concerns arose about the potential for a wage-price spiral. For instance, the Croatian Employers’ Association cautioned that robust wage growth “raises the risk of a wage and retail price spiral” (---,  2024). This risk was taken seriously amid strong price increases, as reflected in remarks by key central bank figures. Notably, ECB President Christine Lagarde addressed the issue in a 2023 speech (ECB, ref13615#2023), and CNB Governor Boris Vujčić noted in 2024 that inflationary pressures had been contained before a wage-price spiral could take hold (CNB, 2024a).
3 For instance in Lider (---, 2024b) and HRT (see: Kovaček, 2024).
4 Which includes public administration and defence services; compulsory social security services (O), education services (P), and human health services (Q).
5 See Deutsche Bundesbank ( 2019). It is important to note that consumers (and central banks) are primarily interested in the potential pass-through of wage growth to consumer prices measured by the Harmonized Index of Consumer Prices (HICP). In contrast, the conclusions of theoretical models generally refer to prices measured by the GDP deflator, which captures the prices of goods produced within the domestic economy. Since consumers also purchase imported goods and since a significant share of imported intermediates is used in the domestic production of consumer goods, the pass-through of wage growth to consumer prices will not necessarily correspond to the pass-through of wage growth to the prices of goods produced within the domestic economy.
6 Markets are often not perfectly competitive, allowing companies to set prices above marginal costs, meaning they can charge markups. These markups do not necessarily remain constant, so companies can adjust them to mitigate (or amplify) the pass-through of higher wages to prices by reducing (or increasing) their markups. 7 In describing the model, we abstract from the effect of so-called aspiration wages, which Blanchard and Bernanke ( 2023) consider in their highly influential paper. Specifically, while standard WS-PS models typically do not include aspiration wages (see, for example, Carlin and Soskice, 2014), integrating the concept of aspiration wages explicitly into the methodological approach underlying the empirical estimation would be challenging.
7 In describing the model, we abstract from the effect of so-called aspiration wages, which Blanchard and Bernanke ( 2023) consider in their highly influential paper. Specifically, while standard WS-PS models typically do not include aspiration wages (see, for example, Carlin and Soskice, 2014), integrating the concept of aspiration wages explicitly into the methodological approach underlying the empirical estimation would be challenging.
8 In the empirical model, expectations are not explicitly modelled because the time series of short-term and long-term inflation expectations published by Consensus Economics for Croatia is too short and only available at a semi-annual frequency.
9 For instance, Gumiel and Elke ( 2018), Bobeica, Ciccarelli and Vansteenkiste ( 2019, 2021).
10 The potential output is assumed to be the level of income achieved at the natural rate of unemployment, that is, the unemployment rate that does not accelerate inflation (Non-Accelerating Inflation Rate of Unemployment, NAIRU).
11 The idea that the pass-through of wages to prices can vary depending on the shocks affecting an economy originates from the literature on exchange rate pass-through to inflation, as discussed in the papers of Kunovac and Communale ( 2017), and Forbes, Hjortsoe and Nenova ( 2018).
12 The authors adopt the idea of identifying the effects of specific shocks using a hypothetical scenario from the literature related to isolating the role of confidence in the transmission of government spending shocks (Bachmann and Sims, 2012) and isolating the role of the credit channel in the transmission of monetary shocks (Ciccarelli, Maddaloni and Peydró, 2015).
13 The identification scheme relies on sign restrictions on impulse response functions proposed by Canova and De Nicolò ( 2022), and Uhlig ( 2005), and further refined by Rubio-Ramírez, Waggoner and Zha ( 2010), and Arias, Rubio- Ramirez and Waggoner ( 2014).
14 Data, data sources and their modifications are available in table A1.
15 According to the statistical classification of products by activity (CPA) section O includes public administration and defence services; compulsory social security services, section P includes education services, and section Q includes human health and social work services.
16 Harmonised index of consumer prices.
17 The HICP for public sector services is calculated as the share of consumption of a specific product by activity (CPA) in the public sector which is linked in the conversion tables to the corresponding category of consumption by purpose (COICOP) and the corresponding HICP index by purpose of consumption (COICOP). See explanation for table A2 and graph A1 for details.
18 Calculated using Croatian Bureau of Statistics (CBS) data collected through regular monthly surveys on net and gross wages according to the classification of economic activities in the European community (NACE).
19 Calculated as the ratio of GVA (gross value added of the O-Q sections) and the number of employees (in the O-Q sections).
20 The correlation coefficient is 0.56.
21 For example, structural reforms that lead to changes in wages regardless of the prevailing economic conditions (such as the 2024 public sector reform) or alter workers’ bargaining power serve as good examples.
22 It is important to note that this approach to analysing the effects of wages on inflation differs from cases where wage changes are the result of endogenous wage reactions to various economic shocks. In those situations, it is necessary to isolate the labour cost channel to evaluate its contribution to inflation at a specific point in time, considering the mix of shocks affecting the economy at that moment. It that case the focus is on the role of wages in amplifying the effects of economic shocks on inflation. The concept of studying “amplification” in a VAR model by constructing a hypothetical scenario – where wages do not respond to a specific shock – was employed by Ivanac, Kunovac and Nadoveza ( 2024), drawing on the methodology of Bachmann and Sims ( 2012) and Ciccarelli, Maddaloni and Peydró ( 2015).
23 The price response is measured by the impulse response function (IRF) of the public sector services HICP to a one-standard-deviation wage mark-up shock over a given horizon, h.
24 In the case of wage spillover effects, the indirect effect can be calculated by using the impulse response function of inflation to wage mark-up shock and the estimated economy-wide wage mark-up shock.
25 I.e., the relationship between wages and prices under different shocks, or the conditional correlations between wage growth and inflation.
26 As explained in the theoretical section, wages are endogenous and primarily driven by inflation in previous periods (wages catch-up to maintain real purchasing power) and labour productivity (if wage growth is offset by productivity gains it does not create additional production costs). To calculate the wage shock, the wage increase related to reform must be adjusted to account for last year’s inflation and the anticipated labour productivity growth in the public sector. The expected annual productivity growth rate is 0.65%, which reflects the average of the annual growth rates derived from the Hodrick-Prescott trend of public sector labour productivity, based on the ratio of gross value added to employment in the O-Q sector.
27 See: Government of RC ( 2024).
28
29 See: Government of RC ( 2024).
30 
31 
32 Although it may appear counterintuitive, it’s important to note that any cost shock, by definition, negatively impacts production and leads to a reduction in employment (or an increase in unemployment). As a result, the pass-through of higher corporate wage bill to prices is moderated by the negative effects of reduced production and lower employment. In contrast, during demand shocks, the increase in costs due to wage growth further exacerbates price increases driven by strong demand.
|
