Abstract
Although there are several mechanisms within theoretical models acknowledging that supply shocks can account for an important part of output fluctuations, even in the short-run, policy practitioners continue endorsing the idea that only demand shocks explain them. This article provides empirical evidence on several Latin American countries and the USA to show that the share of output variance explained by supply shocks in the short-run is substantial. It also offers a more agnostic implementation of the Blanchard–Quah type of structural analysis that focuses on policy evaluation. For this purpose, we propose constructing two indicators out of the historical decomposition of shocks: the goods market unbalance (GMU) and the total cyclical fluctuations (TCF). While GMU is an excess demand measurement that reveals the scope of the distortions caused by shocks, TCF, combined with GMU, helps to understand what type of shock is predominantly explaining (output and inflation) fluctuations. These two pieces of information provide a very different diagnosis than traditional output gaps and should guide monetary policy interventions more adequately. The agnosticism of this proposal has two aspects: the use of a different identification strategy and the assessment of the effects of both supply and demand shocks on output.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In the context of New Keynesian DSGE models, the output gap is defined as the difference between actual output (with nominal rigidities) and potential output (the output without nominal rigidities). However, this definition is not usually implemented by practitioners. For a more detailed discussion on how potential output and output gap should be measured see Basu and Fernald (2009).
Keating (2013) shows that empirical evidence for the USA and other countries, prior to WWI, is inconsistent with the assumption that demand shocks cannot have a permanent effect on output. His claim is based on the comparison between the theoretical assumptions imposed by the Blanchard–Quah identification and the estimated parameters of a statistical model.
For instance, a positive demand shock and a negative supply shock—simultaneously observed—cause an excess demand (GMU > 0) where firms are reducing their supply of goods while households are increasing their demand.
How to jointly interpret these indicators is illustrated in Sect. 4.
See Basu and Fernald (2009) for a more detailed discussion about the definition and estimation of potential output.
It is also the case that firms exercising market power in input markets can affect the prices of these inputs. Nonetheless, these changes in input prices should be reflected in changes in marginal costs.
For example, one could speculate that simultaneously having firms reducing their supply of goods and households increasing their demand might translate into a sizable demand for banking credit. This intuitively could depict the conditions for emergence of future financial distress with an important welfare loss.
Although one might argue that output and prices are determined by many other variables not included in the VAR, this endogeneity does not constitute a problem if we are not interested in tracking down and explaining transmission mechanisms. Instead, we are focusing on characterizing the full observed behavior of prices and output, only conditional on the existence of supply and demand shocks.
As pointed out in Fisher et al. (2015), sign restrictions do not distinguish between demand shocks with permanent and non-permanent effects. However, this is not a problem because our objective is to implement a notionally more general identification strategy.
This orthogonalization can be obtained with a Cholesky or a spectral decomposition (using eigenvectors and eigenvalues). Results tend to be robust to the selection of the orthogonalization method. Rotation matrices are obtained by applying the QR decomposition to a unitary random matrix.
See Rubio-Ramírez et al. (2010) for a detailed theoretical discussion about model identification using sign restrictions and the algorithm for identification.
The length of time for applying sign restrictions can vary depending on the existence of a priori assumption regarding the definition of shocks.
This is true, as long as \(\tau \) is sufficiently long to capture the complete dynamics of short term shocks.
In situations of nonsignificant unbalances (GMU close to zero), supply and demand shocks tend to have positive effects on output growth (TCF > 0), while inflation remains stable.
This does not mean that monetary policy should not have been countercyclical (expansionary), as it actually was. It only means that in cases like this, monetary policy responses are not likely to change the reduction in firms’ supply, unless it is believed that former contractionary monetary policy actions were the actual trigger for firms’ behavior in the first place or that softer monetary conditions can help firms improve their production.
Another relevant piece of information for policy analysis refers to the historical decomposition of inflation. Because demand and supply shocks do not impact output and inflation with the same intensity, looking at inflation components can provide a more precise idea on the effects of shocks. However, analysts are usually more interested in the behavior of inflation that is originated in the behavior of output, information already provided by GMU.
For most countries, the level of the real activity indicator was available at Central Banks’ Web sites, and the annual-log differences used for our calculations were obtained from these statistics. Two exceptions should be mentioned. Colombia only had available annual growth rates, and the level indicator was estimated using an arbitrary base. For the USA, we used the smoothed Chicago Fed National Activity Index (CFNAI-M3), which is a weighted average of 85 monthly indicators of national economic activity. Because this index already subtracts the historical trend rate of growth, we could not recover a consistent level indicator of real activity. In this case, the output gap results from applying the HP filter to the available variable.
It should be noted that, when applying long-run restrictions, historical decompositions are still computed with the matrix that factorizes the covariance matrix of reduced residuals, i.e., the matrix that contains short-run responses to structural shocks.
Brazil, for instance, could be an exception. Analysts might prefer using unemployment with a long-run identification to estimate the demand component. However, this comes at the cost of considerable mistakes in the assessment of the supply component.
References
Basu S (1995) Intermediate goods and business cycles: implications for productivity and welfare. Am Econ Rev 85(3):512–531
Basu S, Fernald JG (2009) What do we know (and not know) about potential output? Fed Reserve Bank St. Louis Rev 91 July/August 2009
Blanchard OJ, Quah D (1989) The dynamic effects of aggregate demand and supply disturbances. Am Econ Rev 79(4):655–673. https://doi.org/10.2307/1827924
Buch CM, Eickmeier S, Prieto E (2014) Macroeconomic factors and microlevel bank behavior. J Money Credit Bank 46(4):715–751. https://doi.org/10.1111/jmcb.12123
Canova F (2005) The transmission of US shocks to Latin America. J Appl Econ 20:229–251. https://doi.org/10.1002/jae.837
Canova F, Nicoló GD (2002) Monetary disturbances matter for business fluctuations in the G-7. J Monet Econ 49(6):1131–1159
Eickmeier S, Hofmann B (2013) Monetary policy, housing booms, and financial [im]balances. Macroecon Dyn 17(4):830–860. https://doi.org/10.1017/S1365100511000721
Eickmeier S, Hofmann B, Worms A (2009) Macroeconomic fluctuations and bank lending: evidence for Germany and the euro rea. Ger Econ Rev 10(2):193–223
Fernández-Villaverde J, Guerrón-Quintana P, Kuester K, Rubio-Ramírez J (2015) Fiscal volatility shocks and economic activity. Am Econ Rev 105(11):3352–3384. https://doi.org/10.1257/aer.20121236
Fisher LA, Huh HS, Pagan AR (2015) Econometric methods for modelling systems with a mixture of I (1) and I (0) variables. J Appl Econom
Fry R, Pagan A (2011) Sign restrictions in structural vector autoregressions: a critical review. J Econ Lit 49(4):938–960
Fuentes R, Gredig F, Larraín M (2008) La brecha de producto en Chile: medición y evaluación. Econ Chil 11(2):7–30
Gopinath G, Neiman B (2014) Trade adjustment and productivity in large crises. Am Econ Rev 104(3):793–831. https://doi.org/10.1257/aer.104.3.793
Halpern L, Koren M, Szeidl A (2015) Imported inputs and productivity. Am Econ Rev 105(12):3660–3703. https://doi.org/10.1257/aer.20150443
Keating JW (2013) What do we learn from Blanchard and Quah decompositions of output if aggregate demand may not be long-run neutral? J Macroecon 38(part B):203–217. https://doi.org/10.1016/j.jmacro.2013.07.007
Mumtaz H, Surico P (2009) The transmission of international shocks: a factor-augmented VAR approach. J Money Credit Bank 41(s1):71–100. https://doi.org/10.1111/j.1538-4616.2008.00199.x
Rubio-Ramírez JF, Waggoner DF, Zha TAO (2010) Structural vector autoregressions: theory of identification and algorithms for inference. Rev Econ Stud 77(2):665–696. https://doi.org/10.1111/j.1467-937X.2009.00578.x
Uhlig H (2005) What are the effects of monetary policy on output? Results from an agnostic identification procedure. J Monet Econ 52(2):381–419
Acknowledgements
This paper received the financial support of the Centro de Estudios Monetarios Latinoamericanos (Cemla) through the part-time at distance internship modality. This paper has also benefited from the comments of two anonymous referees and the participants at the seminars of Central Bank of Venezuela and Cemla, especially Daniel Barráez, Alberto Ortiz, Kólver Hernández, Horacio Aguirre, Jorge Hernández, Nora Guarata and Paul Castillo. Lorena Barreiro provided excellent research assistance.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Pagliacci, C. Are we ignoring supply shocks? A proposal for monitoring cyclical fluctuations. Empir Econ 56, 445–467 (2019). https://doi.org/10.1007/s00181-017-1371-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-017-1371-x