WORKING PAPERS
Climate Change, Sovereign Debt Sustainability, and Fiscal Vulnerability (2026), joint with Abdulla, E.
This paper was presented at:
European Economics and Finance Society Annual Conference (Presented by coauthor, June 2026, Sevilla, Spain);
57th Annual Conference of the Money, Macro and Finance Society (Presented by coauthor, Sept 2026, Lancaster, UK);
Government Investment vs. Consumption: Fiscal Multipliers under Financial Frictions, joint with Wan, Q. ( Online Appendix)
Does the composition of government spending matter when financial intermediaries are constrained? In U.S. data, state-dependent local projections show that government- investment multipliers deteriorate sharply under tight credit conditions while government- consumption multipliers do not; a Kitagawa–Blinder–Oaxaca decomposition attributes this gap to credit conditions rather than to the business cycle. A DSGE model with leverage-constrained intermediaries, rule-of-thumb households, and public capital explains why: the two spending types generate nearly identical impact multipliers but sharply different cumulative multipliers (0.90 versus 0.66 at five years), because bond- financed public investment both builds productive capital and relaxes intermediary balance sheets. Financial frictions thus trade impact for persistence, and government investment’s cumulative advantage survives at every level of intermediary fragility
This paper was presented at:
12th Hong Kong Economic Association Biennial Conference (Presented by coauthor, June 2026, University of Macau, China);
28th INFER ANNUAL CONFERENCE & FI BA XXIII INTERNATIONAL CONFERENCE ON FINANCE AND BANKING, FI BA, (Bucharest, Romania); (*Shortlisted for the Young Economist Best Paper)
State-Dependent Labour Mobility Responses to Macroeconomic Shocks: Does the Type of Recession Matter? (2026). (Online Appendix)
Does the type of recession shape how macroeconomic shocks move workers across sectors? Using sectoral mobility data spanning 1947–2026 and the four-state decomposition of Jo and Zubairy (2025), I show that it does. The central result is a cross-shock regularity: across a wide set of independently identified shocks, distinguishinglow-inflation from high-inflation slack states—demand recessions from stagflation, a distinction drawn by realised inflation rather than by identifying the source of each downturn—reveals state dependence that the conventional recession-versus-expansion split averages away, and the mobility response is the larger in the high-inflation slack state for essentially all of the cleanly identified shocks, spanning demand, supply, andphysical-disaster shocks alike. This regularity is not an artefact of any single historical episode: the influential stagflation quarter differs across shocks, and the ordering survives removing the 1980–1982 episode for nine of the ten identified shocks. Fiscal news is the sharpest individual case—it has no effect at all in a standard unconditional local projection, yet the four-state decomposition shows it raising mobility modestly but persistently in low-inflation demand recessions and substantially more in high-inflationstagflation—but it is also the most leverage-sensitive, its peak magnitude resting heavily on a single quarter in the short 1975–1985 stagflation sample, so I report the di-rection of its contrast rather than its size. Whether the larger stagflation response reflects productive “cleansing” or unproductive “sullying” reallocation cannot be settled with aggregate data, since the mobility index measures the volume of reallocation, not its quality; but a first look at linked worker-flow data, in which movers’ destination earnings fall further when inflation is high, is consistent with sullying. The findings imply that stabilisation policy should condition not only on whether the economy is in recession, but on which kind.
[SUERF Policy Brief Mar 2026, No 1382]
This paper was presented at:
10th International Ioannina Meeting on Applied Economics and Finance, (June 2026, Syros, Greece).
13th Meeting of the Nordic Econometric Network (June 2026, Helsinki, Finland).
32nd International Conference on Computing in Economics and Finance, CEF (July 2026, Venice, Italy)
Oil Price Shocks and US House Prices: The Role of Credit Conditions (2026), joint with Li, Q. (Online Appendix)
We study how oil price shocks transmit to US housing markets. In a structural VAR identified with oil supply news shocks, a 10% oil price increase lowers US house prices by 1–1.2%. State-dependent local projections show that credit conditions govern the persistence of the decline rather than its impact: under loose credit conditions house prices fall by 1 . 9% and recover within eighteen months, whereas under elevated spreads the trough is three times shallower and never reverses. A two-sector New Keynesian DSGE model with oil as an intermediate input rationalises this: oil shocks raise non- housing prices by more than house prices, so housing cheapens in relative terms, while bank balance-sheet and borrower collateral frictions shape the recovery. Tightening loan- to-value ratios shifts the strain onto borrower loan pricing and tightening bank leverage onto capital spreads, leaving the house-price response unchanged: macroprudential policy reallocates rather than eliminates the financial footprint of energy shocks.
This paper was presented at:
18th Annual Meeting of The Portuguese Economic Journal, (July 2025, Lisbon, Portugal).
XXVII Applied Economics Meeting (June 2025, Murcia, Spain).
XXV Conference on International Economics and XII Meeting on International Economics, (Presented by coauthor, June 2024, Alicante, Spain);
The 2024 RCEA International Conference in Economics, Econometrics, and Finance, (May 2024, London, UK).
24th Scottish Economic Society Annual Conference, (April 2024, Glasgow, UK).
17th South-Eastern European Economic Research Workshop of the Bank of Albania, (Presented by coauthor, December 2023, Tirana, Albania).
Why services wage stickiness drives U.S. inflation (2026)*, (Online Appendix)
Aggregate US inflation hides wide differences across sectors, and those differences are lopsided: inflation rates and how often firms change prices vary across product categories by more than a factor of ten, while how often wages are reset barely varies at all. In an eleven-sector New Keynesian model estimated on US data from 2006 to 2019, one sector’s wage shocks account for roughly 87 percent of the unpredictable movement in aggregate inflation. That sector is Other Services — housing, education, medical care and personal services, about half of what households spend. Its wages are the slowest in the economy to be renegotiated, and because the sector is so large, that slowness governs how quickly aggregate inflation can move. What a single-slope aggregate Phillips curve averages over is therefore not differences in price flexibility but one sector’s wage margin. The horizon on which monetary policy reaches inflation is set by how often services wages are renegotiated, not by the faster economy-wide average that a model with one labor market implies.
*Previously circulated under the name "Sectoral Shocks, Labor Mobility and Heterogeneity in Price/Wage Stickiness".
This paper was presented at:
Bank of Albania Friday Seminar Series, (May 2025, Tirana, Albania).
54th Annual Conference of the Money, Macro and Finance Society, (September 2023, Portsmouth, UK).
Society for Computational Economics 29th International Conference on Computing in Economics and Finance, (July 2023, Nice, France).
27th International Conference on Macroeconomic Analysis and International Finance, (May 2023, Crete, Greece).
47th Simposio de la Asociación Española de Economía-Spanish Economic Association (SAEe), (December 2022, Valencia, Spain).
25th Central Bank of Colombia Macroeconomic Modelling Workshop (November 2022, Virtual).
2nd Bank of Lithuania Invited Lecture Series and Conference (September 2022, Vilnius, Lithuania).
EcoMod2022 International Conference on Economic Modelling and Data Science (September 2022, Slovenia/Online).
Warsaw International Economic Meeting 17th Conference (June 2022, Warsaw, Poland).
Economics Research Students Annual Conference (June 2021, University of Exeter).
Graduate Workshop (February 2021, University of Exeter).
The effects of government spending under trend inflation: theory and empirics (2026). (Reject & Resubmit EER) Old version. (Online Appendix)
This paper presents empirical evidence that long-run inflation is important in explaining cross-country differences in the response of private consumption to a government spending shock. Contributing to the debate on the size of fiscal multipliers, I motivate my analysis by documenting, in a quarterly dataset of OECD countries, that countries with high long-run inflation display a relatively higher response of private consumption to an increase in government spending. It is on this basis that I develop a small-scale DSGE model with positive trend inflation and show that the higher the trend inflation in an economy is, the higher the response of private consumption to a government spending shock. If we interpret positive trend inflation as the long-run inflation target, I show, convincingly, that the monetary stance of the central banks has important implications for the effectiveness of short-run fiscal policy interventions. Finally, I calculate consumption multipliers. I find that the consumption multipliers in countries with low trend inflation are below one, while under high trend inflation are higher than 2. These multipliers are consistent with the empirical evidence, which I provide in the paper.
This paper was presented at:
2nd Baltic Central Banks’ Lecture Series and Workshop (May 2025, Riga, Latvia).
The 6th Baltic Economic Conference, (Tallinn, Estonia, June 2024).
Warwick Macro and International Workshop, (Coventry, UK, June 2024).
16th Annual Meeting of The Portuguese Economic Journal, Braga, Portugal, (July 2023).
EcoMod2023 International Conference on Economic Modeling and Data Science (July 2023, Prague).
19th Euroframe Conference on Economic Policy Issues in Europe, Sciences Po, Paris, France, (June 2023).
The Royal Economic Society’s 2021 Annual Conference (April 2021, Online).
The Italian Economic Association 60th Annual Scientific Conference (October 2019, Palermo).
Economics Research Students Annual Conference (June 2019, University of Exeter).
Graduate Workshop (November 2018, University of Exeter).
Publications
Sabaj, E (2024). How do sectoral elasticities affect the transmission of monetary shocks? Journal of Economic Studies (ABS 2), doi:
https://doi.org/10.1108/JES-05-2024-0317
[SUERF Policy Brief, Nov 2024, No 1026]
Troug, H and Sabaj, E (2024). Monetary Policy in a Small Open Economy with Non-Separable Government Spending. Journal of Economic Studies (ABS 2), doi: https://doi.org/10.1108/JES-10-2022-0513
Sabaj, E, Sbia, R., and Troug, H (2023). Does it matter where and how governments spend? Economics Letters (ABS 3), doi:
https://doi.org/10.1016/j.econlet.2023.111158
[SUERF Policy Brief, Jun 2023, No 606]
PRE-PHD RESEARCH WORK
Kahveci, Mustafa and Sabaj, Ernil (2017). Determinant of Housing Rents in Urban Albania: An Empirical Hedonic Price Application with NSA Survey Data. Eurasian Journal of Economics and Finance, Vol. 5(2), p. 51-65, 2017.
Available at: https://eurasianpublications.com/wp-content/uploads/2021/02/EJEF-5.2.4.pdf
Sabaj, Ernil (2017): Cyclical Behaviour of Fiscal Policy in the Western Balkans. (Dormant)
Available at: https://mpra.ub.uni-muenchen.de/84279/
Sabaj, Ernil and Kahveci, Mustafa (2016): Forecasting tax revenues in an emerging economy: The case of Albania. (Dormant)
Available at: https://mpra.ub.uni-muenchen.de/84404/