Why are the high-income countries not mired in deflation? This is the puzzle today, not the absence of the hyperinflation that hysterics have wrongly expected. It is weird that inflation has remained so stable, despite huge shortfalls in output, relative to pre-crisis trends, and prolonged high unemployment. Understanding why this is the case is important because the answer determines the correct policy action. Fortunately, the news is good. The stability of inflation seems to be a reward for the credibility of inflation targeting. That gives policy makers room to risk expansionary policies. Ironically, the success of inflation targeting has revitalised Keynesian macroeconomic stabilisation. A chapter in the latest World Economic Outlook from the International Monetary Fund presents the case for this encouraging conclusion, as has already been noted by, among others, Gavyn Davies and Paul Krugman. Its starting point is with the stickiness of inflation, despite the lengthy period of very high joblessness. Thus, states the IMF, “we find a dog that did not bark”. A possible explanation for this phenomenon is structural. It is argued by many, for example, that the workers who lost their jobs in construction and other bubble-era activities have the wrong skills or are located in the wrong places to take up the new jobs that might now – or soon – be on offer. Again, if unemployment is allowed to remain high for long, what starts as temporary joblessness is likely to become permanent, as workers lose the skills and networks that make it relatively easy to find jobs. So the duration of the Great Recession has put long-term joblessness near record levels. All this tends to make competition in labour markets weak. An alternative explanation is more encouraging. It is that inflation targeting has anchored expectations and so labour market behaviour. Moreover, these targets are close to zero. We know that workers are resistant to cuts in nominal wages. This has remained true throughout the Great Recession: it is indeed one of the reasons why the adjustment in the eurozone is so painful. So, for this reason too, inflation would be sticky, at least downwards. The chapter’s preliminary analysis of these alternatives reaches three main conclusions. The first is that “expectations are strongly anchored to the central banks’ inflation targets rather than being particularly affected by current inflation levels”. Second, the anchoring of expected inflation has increased over time, while the impact of current inflation on expected inflation has diminished. Finally, the relationship between inflation and current unemployment has correspondingly diminished. It became nearly non-existent after 1995, which saw a long period of stable inflation in line with the inflation targets of central banks. The detailed econometric work supports this preliminary analysis, but adds two further points. The most important is that there is substantial cyclical unemployment, at present. A less important one is that the impact of global inflation on inflation in individual countries has not shown any clear trend. An analysis of the US reveals the import of these changes: if the relationship between the cycle and inflation were now as it was in the 1970s, the US price level would already be falling. Fortunately, that has not been the case: otherwise, real rates of interest would now be strongly positive and balance-sheet deflation far more threatening to US stability than it has been. Encouragingly, experience during the period of economic boom, before the financial crisis, suggests the stickiness of inflation does not only work in one direction. Inflation remained in line with targets then, too. This was notably true in Spain and the UK. (See charts.) A final interesting conclusion emerges from the contrast between the performance of the US and Germany in the 1970s, which was when the Bundesbank cemented its reputation. It did not achieve this success because it never made mistakes, but because people believed it would do what was necessary to hit the target. Inflation targeting could indeed be flexible, provided it remained credible. This is important analytical work with big implications for policy. First, mistakes in estimating the degree of economic slack, which are inevitable, may not matter that much, provided people remain sure that the central banks are committed to their targets. This is one of the big benefits of a flat “Phillips curve” (the relationship between cyclical unemployment and inflation). Second, given the uncertainty about the degree of slack and the failure of inflation to respond to huge recessions, it is imperative that central banks not limit their aims to hitting their inflation targets. On the contrary, their job in a deep recession and, in some countries, alas, a worsening one, is to aim for the highest levels of activity consistent with stable inflation. Past success gives them not only the opportunity, but the obligation, to risk expansion of demand in contractionary times. Those running the European Central Bank, please note: low inflation is not enough. Third, while central banks should retain inflation targets at the core of their objectives, experience has also demonstrated that this is not enough. The idea that it is easier to clean up after a financial mess than limit a credit bubble has proved wrong. The only question is how to act. It is clearly important to build a more resilient financial system, through higher capital standards and aggressive macroprudential policies. None of this is going to be easy. A chapter in the companion Global Financial Stability Report brings out, for example, the potential drawbacks of the unconventional policies that central banks were driven to use once interest rates became close to zero. While changing inflation targets would be extremely risky, what has happened suggests that somewhat higher inflation might have been helpful. Experience certainly indicates that monetary policy is not all that effective, on its own, during a balance-sheet recession. It must be complemented by fast reconstruction of the financial system, accelerated private sector deleveraging and a willingness to employ the fiscal balance sheet to support demand, wherever feasible. Yet, for all the complacency of the pre-crisis period, it is good to know that its one clear success – that of cementing inflation expectations – has given policy makers needed policy flexibility. They must use it. 为什么高收入国家没有陷入通货紧缩的泥潭？这才是今天的待解之谜，而不是为什么恶性通货膨胀没有出现——一些歇斯底里的人曾错误地预测会出现恶性通胀。奇怪的是，尽管相对于危机之前的趋势水平，经济产出已大幅下降，失业率也长期居高不下，通胀却一直保持稳定。理解为什么会出现这种情况非常重要，因为答案决定了能否采取正确的政策行动。幸运的是，我们有好消息。稳定的通胀似乎是对通胀目标可信度的一种嘉奖。于是政策制定者就有了采取扩张性政策的冒险空间。具有讽刺意味的是，通胀目标的成功赋予了凯恩斯主义宏观经济稳定政策新的活力。 国际货币基金组织(IMF)最新发布的《世界经济展望》(World Economic Outlook)的一个章节就分析了这个鼓舞人心的结论，加文•戴维斯(Gavyn Davies)和保罗•克鲁格曼(Paul Krugman)也都指出过这点。其出发点就是，尽管高失业率长期持续，但通胀具有粘性。因此，IMF指出，“我们找到了一条不叫的狗。” 这个现象可以从结构的角度来解释。比如说，很多人都认为，建筑行业或者其他泡沫行业的失业工人，由于缺乏合适的技能或者身在错误的地方，而无法从事现在或者不远的将来可能出现的新工作。同样，如果放任失业率长期居高不下，最初的临时失业则可能会转变为永久失业，原因是工人会失去技能和关系网，使得找工作相对较难。因此，“大衰退”(Great Recession)的持续已经使得长期失业逼近破纪录的水平。所有这些都可能会使劳动力市场的竞争弱化。 另外一个解释更鼓舞人心：通胀目标使人们的预期固定下来，也稳定了劳动力市场的表现。此外，这些通胀目标都接近于零。我们知道，工人们反感名义工资的降低，整个“大衰退”期间都是如此：这也是欧元区调整如此痛苦的原因之一。因此，也正是因为这个原因，通胀将具有粘性，至少是向下的粘性。 该章节初步分析了这些可能的解释之后，得出了三个主要结论。第一，“央行的通胀目标对预期有很强的锚定作用，而当前的通胀水平对预期没有特别大的影响。”第二，通胀目标对通胀预期的锚定作用随着时间的推移越来越大，而当前通胀率对通胀预期的影响逐渐减小。最后，通胀率和当前失业率之间的联系也相应减小。1995年之后，这种联系基本上已经不存在，而同时，通胀与央行通胀目标之间长期保持稳定的一致。 更为细致的计量经济学研究工作支持这一初步的分析，但补充了两点。较为重要的一点是，当前存在明显的周期性失业。次要的一点是，全球通胀对各国通胀的影响还没有任何明显的趋势。 对美国的分析揭示了这些变化的重要性：如果现在的周期和通胀之间的关系与上世纪70年代类似，那么美国的物价应该已经下降。幸运的是，事实并不是这样：否则的话，实际利率就会远高于零，资产负债表收缩对美国稳定性的威胁也会远大于现在。令人鼓舞的是，金融危机之前的经济繁荣期间的经验表明，通胀的粘性不仅是朝着一个方向。那个时候通胀率与通胀目标也是一致的。这点在西班牙和英国表现尤为明显。（见图表。） 最后一个有趣的结论是通过对上世纪70年代美国和德国的对比得出的，正是在那个时候德国央行(Bundesbank)确立了声望。它不是因为从不犯错而获得这种成功，而是因为人们相信它会采取必要措施来实现目标。只要保持可信度，通胀目标就可以是灵活变动的。 这是重要的分析工作，对政策有着重要意义。 首先，评估经济放缓程度时即使犯错（这是不可避免的），可能也没那么要紧，只要人们依然确信央行致力于实现制定的目标。这是平坦的菲利普斯曲线（Phillips curve，描述周期性失业和通胀之间的关系）的较大优势之一。 第二，鉴于经济放缓程度的不确定性和通胀应对严重衰退的失败，央行必须不将目标局限于实现通胀目标。与此相反，当经济深陷衰退时，或者在某些国家，当经济衰退加剧时，它们有责任以实现与稳定的通胀水平相一致的最高经济活动水平为目标。过去的成功不仅赋予了它们这种机会，也赋予了它们一种职责，在收缩时期试着扩大需求。欧洲央行(ECB)的管理者们请注意：低通胀是不够的。 第三，尽管央行需要将维持通胀目标作为核心目标之一，经验也表明这是不够的。在金融动荡之后“收拾摊子”比抑制信贷泡沫要容易的观点，已被证明是错误的。唯一的问题在于如何行动。通过更高的资本标准和进取性的宏观审慎政策，建立一个更强韧的金融体系，显然非常重要。 所有这些都不会很容易。比如，IMF同时发布的《全球金融稳定报告》(Global Financial Stability Report)中的一章，就指出了在利率接近于零的时候，央行倾向于使用的非传统政策的潜在缺陷。改变通胀目标确实非常危险，但事实表明较高的通胀或许是有益的。经验证明，在资产负债表衰退期间，单单货币政策并不会产生多大效果，必须得到如下方面的补充：金融体系快速重组，私营领域加速去杠杆化，政府有意愿在任何可行的领域通过动用财政资产负债表手段来支持需求。 另外，尽管危机前的繁荣导致了人们的自满，它却有一个明显的成功之处是我们需要知道的——它强化了人们的通胀预期，这给政策制定者提供了必要的政策灵活性。他们必须用好这种灵活性。
