هزینه بی ثباتی مالی، ترازنامه بخش خانگی و مصرف

Extant work on costs of financial instability focuses on fiscal costs and declines in aggregate GDP following banking crises. We estimate effects of banking and currency crises on consumption in 19 OECD countries, showing consumption plays an important role in the adjustment following a crisis, and effects are not captured solely by the impact of crises on standard consumption determinants, income and wealth. Additional effects, attributable to factors such as time-varying confidence, uncertainty and credit rationing, are aggravated by high and rising leverage, despite financial liberalisation easing liquidity constraints. High leverage implies that banking crises taking place now could have greater incidence than in the past.

Whereas there is an extensive and growing literature which focuses on predicting the incidence of financial instability (in respect of both banking and currency crises), the literature on the costs of financial instability – the key reason for concern in respect of financial turbulence – is relatively sparse. Most of the literature that does exist is focused either on fiscal costs or aggregate measures of losses in GDP from banking crises, rather than the subcomponents of GDP or including currency crises. It is also commonly focused on a global sample, or solely on emerging market economies but rarely on OECD countries alone. Analysis of subcomponents such as investment and consumption is highly relevant for assessing the incidence of crises and devising policy measures to neutralise their adverse impact. Hence, for example, any tax or other fiscal changes designed to cushion the fall in GDP would need to be appropriately directed to the relevant type of expenditure. Clearly, which subcomponents of GDP crises affect will depend partly on the impact of the crisis itself on their determinants, which for consumption are real incomes and net financial wealth. But there are also likely to be effects arising from factors such as uncertainty, confidence and credit availability whose impact is likely to go beyond the “normal” response of consumption to these variables. Impacts of crises on consumption may also vary by type of country and over time. If there are such differences, they may be explicable by structural factors that have evolved over time or vary across countries, e.g. depending on size. The impacts of crises on expenditure components may change over time as financing behaviour and expectations about liquidity respond to financial deregulation. The effect may also vary with balance sheet structure, where growing leverage has been a marked feature of household sectors in recent years. In this context, this article seeks to assess address these issues by measuring the costs of financial instability – in terms of both banking and currency crises – for 19 OECD countries in respect of consumption, the largest component of GDP. The article is structured as follows. First we summarise some recent estimates of the costs of financial instability. We then estimate baseline consumption functions for all 19 countries, and then break the 19 countries we study into G-7, small open economies and the Scandinavians to see if there are differences across countries. We also investigate whether the financially more liberalised 1990s look different from the preceding period. Using these groups throughout, we divide effects of crises between those due to standard consumption determinants and those left unexplained by our equations. We also assess whether the scope of these unexplained effects varies across types of crisis and has increased with rising household leverage.

The literature on costs of financial instability tends to focus on fiscal costs and the impact on GDP. In this paper, we have analysed the effect of a crisis on consumption both in terms of its impact on real personal disposable income and net financial wealth and unexplained effects over and above that. We have shown that consumption plays a key role in the macroeconomic adjustment following a financial crisis. Furthermore, the effect12 of a crisis is aggravated by high and rising leverage, notably as shown by the effect of a high debt–income ratio, despite the accompanying benefits of financial liberalisation in easing liquidity constraints, which in the light of high current levels of this ratio (Debelle, 2004) implies that past estimates of the impact of a banking crisis may be on the low side.13 The impact is also greater in a small open economy than in the G-7. Viewed in the light of growing leverage in recent years, the results imply that a banking crisis taking place now could have a much greater incidence on consumption than in the past, especially if macroeconomic policy14 is unable to respond, as for a small country in EMU.