RDP 2026-02: Shifts in Australian Price-setting Behaviour around Large Shocks 2. Literature Review

Modern macroeconomic models of the business cycle typically incorporate price rigidity to explain how fluctuations in nominal economic variables affect the real economy over time. A widely used framework is the Calvo pricing model, which is popular due to its tractability. This framework assumes that only a fixed proportion of firms can adjust their prices in each period (Calvo 1983). This frequency of price changes – often referred to as the ‘Calvo parameter’ – is usually treated as constant. This results in ‘time-dependent’ pricing, where firms decide to adjust prices on the basis of time rather than economic conditions. Consequently, nominal rigidity is also constant over time, implying a linear relationship between the size of a shock and its inflationary impact. ^[1]

This paper contributes to a substantial empirical literature that has emerged to assess price-setting dynamics. Early studies using CPI and scanner data found that prices were more flexible than assumed in standard macroeconomic models (e.g. Nakamura and Steinsson 2008; Klenow and Malin 2010). More recent work has leveraged web-scraped data, which provides advantages such as higher frequency, broader item coverage, and fewer measurement distortions than official statistics. Pioneering studies by Cavallo and Rigobon (2016) and Cavallo (2018) explore these benefits and show that web-scraped data can yield more accurate estimates of price-setting behaviour.

A general finding across this literature is that the frequency of price changes varies with economic conditions, challenging the fixed rigidity assumption in the Calvo framework. This evidence has motivated the development of models with state-dependent pricing, which better match observed behaviour. These models often imply nonlinear dynamics, where large shocks prompt more frequent price adjustments and therefore have larger and quicker effects on inflation.

Many of these models generate state-dependence by assuming that firms face fixed ‘menu’ costs when changing prices, so they adjust only when the benefit of doing so exceeds the cost (e.g. Alvarez, Le Bihan and Lippi 2016). Other approaches introduce costs of reviewing prices (e.g. Blanco et al 2024c) or costs of acquiring the information needed to choose optimal prices (e.g. Turen 2023). In all of these settings, shocks that push a firm's optimal price further from its current price strengthen the incentive to re-optimise. Firms therefore update their prices more frequently when economic conditions warrant it, giving rise to the state-dependent price setting observed in the data.

Recent papers have explored the importance of pricing dynamics in state-dependent models, particularly in the context of the pandemic (Dedola et al 2023; Auclert et al 2024; Cavallo, Lippi and Miyahara 2024; Blanco et al 2024a, 2024b, 2024c; Gautier et al 2026). These studies find that price rigidity tends to decline during large shocks, resulting in stronger inflation responses than predicted by time-dependent models. For example, Auclert et al (2024) show that while time- and state-dependent models yield similar inflation dynamics for small shocks in the euro area, state-dependent models produce markedly stronger responses when shocks are large.

The incorporation of state-dependent pricing into models also has important implications for monetary policy trade-offs. Karadi et al (2024), for instance, show that the trade-off between lowering inflation and preserving real economic activity in the United States is reduced during large shocks, as falling price rigidity steepens the Phillips curve. This implies that central banks should respond more aggressively to inflationary supply shocks than they would if price rigidity did not decrease.

Our paper contributes to the growing literature using prices microdata to examine price-setting shifts during and immediately after the COVID-19 period. These include Rudolf and Seiler (2022) for Switzerland; Montag and Villar (2025) for the United States, Henkel et al (2023) for euro area countries; Klein, Strömberg and Tysklind (2024) for Sweden; and Bilyk, Khan and Kostyshyna (2024) for Canada. Across these studies, the early pandemic (before inflation started rising significantly) generally saw relatively limited changes in the frequency of price adjustments. Studies covering the subsequent high-inflation period consistently document a strong increase in the frequency of price changes, driven mainly by more frequent price increases. This pattern suggests a shift toward more flexible pricing in response to large shocks, consistent with state-dependent pricing behaviour.^[2]

Finally, our work builds on the limited literature examining price-setting dynamics in Australia. Using RBA firm survey data from 2000 to 2006, Cagliarini, Robinson and Tran (2010) find that prices are less rigid than typically assumed in structural macroeconomic models. Sutton (2017) uses CPI microdata from 2001 to 2017 and identifies moderate variation in price stickiness over time, with only modest increases in flexibility around the global financial crisis. Our paper extends these analyses, using a different data source – web-scraped prices – and focusing on a more recent period of heightened economic volatility with an emphasis on policy implications.

Footnotes

Rotemberg pricing is another form of non-state-dependent pricing commonly used in macro models, where rigidity arises nonlinearly because firms face quadratic price adjustment costs (Rotemberg 1982). In log-linearised models, it results in identical predictions of pass-through as Calvo models, despite the different pricing mechanism. Nonlinear Phillips curves may arise through other mechanisms, such as the specific form of the demand function (Harding, Lindé and Trabandt 2022), and via labour market frictions (Benigno and Eggertsson 2023; Schmitt-Grohé and Uribe 2025). [1]

While these papers document a change in price-setting rigidity, they tend not to test which mechanism is driving firms to shift their behaviour (i.e. menu costs versus changing attention and frequency of price reviews). [2]