Montoya and De Haan, 2008: Hayashida, and Hewings, 2009; Panteladis and Tsiapa, 2011; Bierbaumer-Polly, 2012). Yet others have examined the implication of asymmetries in regional business cycles for monetary integration and common currency areas (Bayoumi and Eichengreen, 1993; Krugman, 1993; Frankel and Rose, 1998; Barrios, Brülhart, Eliot and Sensier, 2003; Barrios and Lucio, 2003; Partridge and Rickman, 2005).
Montoya 和 De Haan，2008 年：Hayashida 和 Hewings，2009 年；Panteladis 和 Tsiapa，2011 年；Bierbaumer-Polly，2012 年）。还有人研究了区域商业周期的不对称对货币一体化和共同货币区的影响（Bayoumi 和 Eichengreen，1993 年；Krugman，1993 年；Frankel 和 Rose，1998 年；Barrios、Brülhart、Eliot 和 Sensier，2003 年；Barrios 和 Lucio，2003 年；Partridge 和 Rickman，2005 年）。

How far this revived interest in regional business cycles has been due to the impress of external economic circumstances is difficult to gauge, particularly as there is some debate over whether and to what extent economic fluctuations have become more or less pronounced over the past thirty years or so. Several claims were made in the early2000s that since the early-1980s the volatility of the US economy had become much attenuated, eliciting the moniker of the ‘great moderation’ (the term coined by Stock and Watson, 2002, 2003). However, such studies typically look at the year on year variability in economic growth rates, which is not the same thing as comparing the severity of successive recessions. In the case of the UK, since the early-1970s, when the so-called long post-war ‘golden age’ of relatively stable growth and mild economic cycles came to an end, the economy has been subject to a succession of sharp recessions, much deeper than those in preceding post-war period. The warning issued by Haggett in 1971 (in the quote above) turned out to be highly prescient. With the recessionary contractions of the mid-1970s, and especially those of the early-198os and the early-1990s, the economic cycle returned with a vengeance. Then the ‘long boom’ from the early-1990s into the late-2000s seemed to herald another ‘golden age’ free of major recessionary disruption, a sort of ‘great moderation’ to parallel that claimed for the US economy, and was interpreted as such by many UK economic and political observers. ${}^1$${}^1$^(1){ }^{1} However, as we now know, in the UK as in the USA, and elsewhere, the boom of the 1990s and 2000s turned out to be unsustainable and unstable, and was brought emphatically to an end in 2008 by the financial crisis which then triggered the worst cyclical contraction in output since the Great Depression of the early-1930s. This time it was Paul Krugman who proved to be prophetic: on several occasions in the mid- to late-1990s he had pointed to various warning signs of impending economic turmoil, even as the apparent ‘boom’ in Western economies was gathering pace. ${}^2$${}^2$^(2){ }^{2} For the UK, the 2008-2010 recession marked the latest in a series of contractions of historically significant amplitude.
对外部经济环境的影响在多大程度上重新引起了人们对地区商业周期的兴趣，这一点很难衡量，尤其是在过去三十多年间，人们对经济波动是否以及在多大程度上变得更加明显或不那么明显存在争议。二十世纪二十年代初，有几种说法认为，自二十世纪八十年代初以来，美国经济的波动性已大大减弱，并因此被称为 "大缓和"（Stock 和 Watson，2002 年和 2003 年创造的术语）。然而，此类研究通常关注的是经济增长率的逐年变化，这与比较连续衰退的严重程度是两码事。就英国而言，自 20 世纪 70 年代初以来，当所谓的战后长期相对稳定增长和温和经济周期的 "黄金时代 "结束时，英国经济经历了连续的急剧衰退，衰退程度远远超过了战后时期。哈格特在 1971 年发出的警告（引自上文）被证明是非常有预见性的。随着 20 世纪 70 年代中期的衰退紧缩，尤其是 80 年代初和 90 年代初的衰退紧缩，经济周期卷土重来。随后，从 20 世纪 90 年代初到 2000 年代末的 "长期繁荣 "似乎预示着另一个没有重大衰退干扰的 "黄金时代"，一种与美国经济相类似的 "大缓和"，许多英国经济和政治观察家也将其解释为 "大缓和"。 ${}^1$${}^1$^(1){ }^{1} 然而，正如我们现在所知道的，在英国、美国和其他国家，20 世纪 90 年代和 2000 年代的经济繁荣被证明是不可持续和不稳定的，并在 2008 年因金融危机而戛然而止，金融危机随后引发了自 20 世纪 30 年代初大萧条以来最严重的周期性产出萎缩。这一次，保罗-克鲁格曼（Paul Krugman）被证明是先知先觉者：在20世纪90年代中后期，他曾多次指出经济动荡迫在眉睫的各种警告信号，即使西方经济表面上的 "繁荣 "正在加速。 ${}^2$${}^2$^(2){ }^{2} 对于英国来说，2008-2010 年的经济衰退标志着一系列具有历史意义的收缩中的最新一次。

How deep recessions impact across regions (and cities) within nations is an important issue. After all, the depth of a national recession, and the speed of recovery from it, are but reflections of the contractions and recoveries in the individual regional and
深度衰退对国内各地区（和城市）的影响是一个重要问题。毕竟，全国性衰退的深度以及衰退后的复苏速度，只是各个地区和城市经济收缩和复苏的反映。

local economies of which the national economy is composed. But more than this, not only may regional variations in the depth of recessionary contraction create spatial inequalities in the negative consequences for employment, household incomes and welfare, there is the possibility that such effects may not be merely transitory, to be eliminated during the subsequent recovery phase, but prove to have longer-lasting repercussions. To use the phrase invoked by Cross, McNamara and Pokrovskii (2009), the ‘memory of recession’ (or ‘remanence’) may linger much longer in some regions than in others. Or to put it more formally, regions may have different degrees of resilience to recessions in the sense that regional economies possess different abilities for recovery, and in some cases recessions may have permanent or hysteretic effects on a region’s growth path. Thus a region that experiences a particularly severe contraction of its economy during a recession may not necessarily ‘rebound’ or return to its pre-recession growth trajectory, but emerge on a lower or inferior growth path: recessions may shape regions’ long-run growth rates, and thereby influence patterns of long run regional economic convergence or divergence (Martin, 2012, 2015; Martin and Sunley, 2015). In short, recessions and recoveries may not be events that simply take place ‘around’ autonomous long-term regional growth and development paths: they may shape and be part of those paths. Indeed, one way of viewing long-run regional growth is as a succession of phases of contraction and expansion, with turning points defined as ‘peaks’ and ‘troughs’ in activity (the terms used by Burns and Mitchell, 1946, in their classic work on business cycles). ${}^3$${}^3$^(3){ }^{3}
在这种情况下，国家经济是由地方经济组成的。不仅如此，衰退收缩深度的地区差异不仅可能在对就业、家庭收入和福利的负面影响方面造成空间上的不平等，而且这种影响还可能不仅仅是过渡性的，在随后的复苏阶段被消除，而是被证明具有更持久的影响。用 Cross、McNamara 和 Pokrovskii（2009 年）引用的一句话来说，"衰退记忆"（或 "再现"）在某些地区可能比在其他地区持续更长时间。或者更正式地说，各地区对衰退的抵御能力可能不同，因为各地区经济拥有不同的复苏能力，在某些情况下，衰退可能会对一个地区的增长路径产生永久性或滞后性影响。因此，一个地区在衰退期间经历了特别严重的经济收缩，不一定会 "反弹 "或恢复到衰退前的增长轨迹，而是会出现较低或较差的增长轨迹：衰退可能会塑造地区的长期增长率，从而影响长期的地区经济趋同或分化模式（Martin，2012，2015；Martin 和 Sunley，2015）。简而言之，衰退和复苏可能并不是 "围绕 "自主的长期区域增长和发展路径而发生的事件：它们可能塑造并成为这些路径的一部分。事实上，将长期区域增长视为连续收缩和扩张阶段的一种方法，其转折点被定义为活动的 "高峰 "和 "低谷"（Burns 和 Mitchell 1946 年在其关于商业周期的经典著作中使用的术语）。 ${}^3$${}^3$^(3){ }^{3}

Given these considerations, our aim in this paper is examine the geographies of the past four economic cycles across the principal regions of the UK. To do this we construct simple measures of regional resistance to, and recovery from, successive recessions in order to analyse whether and in what ways regions have varied in their behavior from one economic cycle to the next. Our particular focus is on the influence of industrial structure on the resilience of UK regions to these recessions. There is a view, both in studies of regional business cycles and in more recent work on regional resilience (and indeed on regional growth more generally), that industrial structure plays a formative role in explaining regional differences in economic performance. This is an intriguing issue, especially when, as here, a number of successive recessions are under investigation, since regional industrial structures themselves change and evolve over time.
鉴于上述考虑，我们在本文中的目的是研究英国主要地区过去四次经济周期的地理特征。为此，我们构建了地区抵御连续经济衰退和从衰退中复苏的简单衡量标准，以分析各地区在不同经济周期中的表现是否存在差异以及差异的方式。我们特别关注产业结构对英国地区抵御经济衰退能力的影响。无论是在地区商业周期研究中，还是在最近关于地区恢复力的研究中（甚至在更广泛的地区增长方面），都有一种观点认为，产业结构在解释地区经济表现差异方面起着形成性作用。这是一个耐人寻味的问题，尤其是在对连续多次经济衰退进行调查的情况下，因为地区产业结构本身会随着时间的推移而变化和发展。

There is no single agreed approach to analysing regional reactions to economic cycles, construed as comprising recessionary contractions or shocks and subsequent recoveries or expansions. Several different perspectives can be identified. A common
经济周期包括衰退性收缩或冲击以及随后的复苏或扩张，在分析地区对经济周期的反 应方面，目前还没有一个统一的方法。可以确定几种不同的视角。一个共同的

approach in the contemporary literature on business cycles is to isolate ‘cyclical’ movements by first detrending the data, for example using some form of band-pass filter (for an example, see Partridge and Rickman, 2005). One problem with most such studies, however, is that statistical trend-cycle decompositions make no distinction between business cycles and growth cycles, and because they often involve smoothing the time series in question they can give misleading estimates of cyclical intensities and turning points (Zarnowitz and Ozyildirim, 2006). A different approach is to assume that economic cycles present movements around some long-run ‘equilibrium’ time path, and then use time series error-correction models to estimate both the long-run ‘equilibrium’ path and the speed of adjustment to cyclical shocks that move the system away from that path (see Fingleton, Garretsen and Martin, 2012). Yet another approach is to use an explanatory structural (causal) model to produce counterfactual paths against which the impact of cyclical disturbances can be estimated (for an example, see Fingleton, Garretsen and Martin, 2015). Others have suggested that recessionary shocks can be interpreted as the occasional ‘plucking down’ of an economy from its (upward-trended) ‘maximum feasible growth ceiling’ (given by an economy’s resources and the efficiency with which they are utilized), and to which it subsequently recovers: this is essentially the approach discussed, for example, by Friedman, 1993, in his work on business fluctuations in the US macroeconomy. This obviously requires estimating what this ‘maximum feasible growth ceiling’ actually is: Friedman himself suggests the hypothetical line connecting successive cyclical peaks.
在当代有关商业周期的文献中，有一种方法是首先对数据进行去趋势化处理，例如使用某种形式的带通滤波器，从而分离出 "周期性 "运动（实例见 Partridge and Rickman, 2005）。然而，大多数此类研究的一个问题是，统计趋势周期分解并不区分商业周期和增长周期，而且由于它们通常涉及对相关时间序列进行平滑处理，因此可能会对周期强度和转折点的估计产生误导（Zarnowitz 和 Ozyildirim，2006 年）。另一种不同的方法是假设经济周期围绕某个长期 "均衡 "时间路径运动，然后使用时间序列误差修正模型来估计长期 "均衡 "路径以及对偏离该路径的周期性冲击的调整速度（见 Fingleton、Garretsen 和 Martin，2012 年）。另一种方法是使用解释性结构（因果）模型来生成反事实路径，并据此估算周期性扰动的影响（示例见 Fingleton、Garretsen 和 Martin，2015 年）。还有人认为，衰退性冲击可解释为经济从（上升趋势的）"最大可行增长上限"（由经济的资源及其利用效率决定）偶尔 "下探"，随后恢复：这基本上是弗里德曼（Friedman，1993 年）等人在其关于美国宏观经济商业波动的研究中所讨论的方法。这显然需要估算 "最大可行增长上限 "究竟是多少：弗里德曼本人提出了连接连续周期性高峰的假设线。

A common theme running through these various approaches is that recessions represent a form of economic ‘shock’ in the sense that they are (in general) unexpected and unpredictable events that disrupt the ‘normal’ growth path of an economy. Recessions are distinctly negative in nature, involving the contraction of economic activity (output), the possible closure of firms, and a fall in employment (due to a cessation of new hires and a rise in layoffs or redundancies). Given its association with the study of shocks, the notion of resilience is a potentially useful concept with which to examine how regions are affected by recessionary downturns (Martin, 2012). The basic idea of resilience is that it captures how an entity or system reacts to and recovers from an adverse disruption. Interest focuses on how long the entity or system in question takes to return to its pre-shock state, and whether in fact it returns to that state or is moved by the shock to some other (preferably more favourable) state. Although the idea of resilience, which originated in the physical sciences, has been used in ecology and certain branches of psychology for some time, it has yet to receive much attention in economics (though one or two authors have begun to explore its applicability - see Briguglio, 2004; Briguglio et al, 2009). However, in recent years it has been taken up with some enthusiasm in economic geography and regional studies, where there is now a growing literature (for example, Cambridge Journal of Regions, Economy and Society, 2010; Hassink, 2010; Bristow, 2010; Simmie and Martin, 2010; Martin, 2012; Fingleton, Garretsen, and Martin, 2012; Scott, 2013; Doran and Fingleton, 2013; Martin and Sunley, 2015; Boschma, 2015). In his discussion, Martin
贯穿这些不同方法的一个共同主题是，经济衰退是一种形式的经济 "冲击"，因为它们（一般来说）是意外和不可预测的事件，扰乱了经济的 "正常 "增长路径。衰退具有明显的负面性，涉及经济活动（产出）的萎缩、企业可能倒闭以及就业率的下降（由于停止招聘新员工以及裁员或被裁员人数的增加）。鉴于其与冲击研究的关联性，复原力概念是一个潜在的有用概念，可用于研究衰退性下滑对地区的影响（Martin，2012 年）。复原力的基本概念是，它反映了实体或系统如何对不利干扰做出反应并从中恢复。人们关注的重点是，有关实体或系统需要多长时间才能恢复到冲击前的状态，以及事实上是恢复到该状态，还是被冲击转移到其他状态（最好是更有利的状态）。恢复力这一概念起源于物理科学，在生态学和某些心理学分支中已使用了一段时间，但在经济学中尚未得到广泛关注（尽管有一两位作者已开始探讨其适用性--见 Briguglio, 2004; Briguglio et al, 2009）。不过，近年来，经济地理学和区域研究领域对这一问题的研究热情不减，文献数量不断增加（例如，《剑桥区域、经济和社会学报》，2010 年；Hassink，2010 年；Bristow，2010 年；Simmie 和 Martin，2010 年；Martin，2012 年；Fingleton、Garretsen 和 Martin，2012 年；Scott，2013 年；Doran 和 Fingleton，2013 年；Martin 和 Sunley，2015 年；Boschma，2015 年）。在讨论中，马丁
(2012) distinguished between four 'dimensions’ or aspects of regional resilience to recessionary (or other such shocks): Resistance (degree of sensitivity or depth of reaction to the shock), Recovery (speed and degree of recovery from the shock), Re-orientation (extent of reorientation and adaptability of regional economy in response to the shock) and Renewal (extent to which regional economy renews its pre-shock growth path or hysteretic shift to a new path). In their comprehensive exploration of the meaning and application of the notion of resilience in the study of regional development, Martin and Sunley (2015) revise and develop this characterization in a number of ways. First, they emphasise that resilience is a multifaceted process, not a singular, static state of affairs or fixed characteristic of a regional or local economy, and can be viewed as comprising four sequential (and recursive) steps: the risk (or vulnerability) of a region’s firms, industries, workers and institutions to shocks; the resistance of those firms, industries, workers and institutions to the impact of shocks; the ability or otherwise of the region’s firms, industries, workers and institutions to undergo the adjustments and adaptations necessary to resume core functions and performances, what we might call (adaptive) reorientation; and finally the degree and nature of recoverability from the shock (see Figure 1).
(2012) 区分了地区抵御衰退（或其他此类冲击）能力的四个 "维度 "或方面：抵御（对冲击的敏感程度或反应深度）、恢复（从冲击中恢复的速度和程度）、重新定位（区域经济为应对冲击而重新定位和适应的程度）和更新（区域经济重新走上冲击前的增长道路或歇斯底里地转向新道路的程度）。Martin 和 Sunley（2015 年）在全面探讨区域发展研究中复原力概念的含义和应用时，从多个方面对这一特征进行了修正和发展。首先，他们强调抗灾能力是一个多方面的过程，而不是一种单一的、静态的事态或区域或地方经济的固定特征，可被视为包括四个连续（和递归）的步骤：一个地区的企业、行业、工人和机构面对冲击的风险（或脆弱性）；这些企业、行业、工人和机构抵御冲击影响的能力；该地区的企业、行业、工人和机构是否有能力进行必要的调整和适应，以恢复核心功能和绩效，即我们可以称之为（适应性）重新定位；最后是从冲击中恢复的程度和性质（见图 1）。

Figure 1: Regional Resilience to Recessions
图 1：地区抵御衰退的能力

(Adapted from Martin and Sunley, 2015)
(改编自 Martin 和 Sunley，2015 年）

Second, these sequential aspects of the resilience process depend on the nature, depth and duration of the recession, on the prior growth path of a region, and on the various determinants of that growth path (regional economic structures, resources, capabilities
其次，复原力进程的这些相继方面取决于衰退的性质、深度和持续时间，取决于一个地区先前的增长道路，以及该增长道路的各种决定因素（地区经济结构、资源、能力、环境和社会）。
and competences, together with any supportive measures undertaken by local or national institutions, including government, such as welfare policies, business support programmes and the like). Those determinants shape a region’s risk and resistance (depth of reaction) to recessionary disturbances, but may themselves undergo shock-induced changes and adjustments, as the region’s economic structures and forms undergo change and adaptations caused by, or in response to, the pressures of the shock. The ability of a region’s economy - its firms, industries, technologies, working practices, institutional arrangements, etc - to undergo positive reconfiguration so as to restore core performances, such as profitability, full employment, investment and so on, defines that economy’s adaptive robustness. The notion of ‘adaptive robustness’ is used in the study of complex biological, organizational and physical systems, where, traditionally, the term was interpreted as meaning that the characteristics of a system are unaffected by perturbations: that is, where there is a consistency (stability) of structure and function (see Whitacre, 2012). But more recently the term has been reconceptualised to refer to the ability of a system to resist external and internal disturbances and disruptions if necessary by undergoing plastic change in some aspects of its structure and components in order to maintain or restore certain core performances or functionalities (Kitano, 2004). It is important to recognize that under this new interpretation, robustness is concerned with maintaining some key functions or performances of a system rather than system states or structures, and that maintaining (or regaining) performance and functionality may actually involve - even require - changes in a system’s structural components and mode of operation. Maintaining (or restoring) profitability, employment and growth can legitimately be viewed as ‘core performances and functionalities’ in an economic context, and clearly different configurations of industries, firms and skills can be commensurate with maintaining those core performances.
这些决定因素决定了一个地区的风险和抵御衰退干扰的能力（反应深度），但它们本身也会随着地区经济结构和形式的变化而发生变化和调整。这些决定因素决定了一个地区对衰退扰动的风险和抵抗力（反应深度），但其本身也可能发生由冲击引起的变化和调整，因为该地区的经济结构和形式在冲击的压力下发生了变化和适应。一个地区的经济--其企业、产业、技术、工作方式、制度安排等--能否进行积极的重新配置，以恢复核心绩效，如盈利能力、充分就业、投资等，决定了该地区经济的适应稳健性。适应稳健性 "这一概念用于研究复杂的生物、组织和物理系统，传统上，这一术语被解释为系统的特征不受扰动的影响：也就是说，系统的结构和功能具有一致性（稳定性）（见 Whitacre, 2012）。但最近，这一术语被重新定义为系统在必要时抵抗外部和内部干扰和破坏的能力，即系统结构和组成部分的某些方面发生塑性变化，以保持或恢复某些核心性能或功能（Kitano，2004 年）。 重要的是要认识到，根据这一新的解释，稳健性涉及的是维持系统的某些关键功能或性能，而不是系统的状态或结构，维持（或恢复）性能和功能实际上可能涉及--甚至需要--改变系统的结构组成部分和运行模式。保持（或恢复）盈利能力、就业和增长可以被合理地视为经济背景下的 "核心绩效和功能"，而产业、企业和技能的不同配置显然可以与保持这些核心绩效相称。

Thus, apart from the technical issues involved, from an ontological viewpoint treating recessions as shocks to an autonomous ‘trend’ or equilibrium economic path is not unproblematic. To do so is to assume that a trend or equilibrium exists independently of cyclical movements, whereas in reality the two may be inextricably interrelated. Slowmoving structural changes and shifts, for example, may alter the cyclical behaviour of an economy; and conversely, a major recessionary shock may well change the pace and direction of structural and developmental change. Separating trend and cycle may thus be misleading, however sophisticated the technique used to undertake the exercise. Our approach here is to not to assume some autonomous path or to attempt to ‘detrend’ our regional series, but to focus on how regions react to recessions defined by the national downswings from peak to trough, and recoveries defined as the national upswings from trough to peak.
因此，除了所涉及的技术问题外，从本体论的角度来看，将衰退视为对自主 "趋势 "或均衡经济路径的冲击并非没有问题。这样做是假定趋势或均衡是独立于周期运动而存在的，而实际上两者可能是密不可分的。例如，缓慢的结构变化和转变可能会改变经济的周期性行为；反之，重大的衰退冲击很可能会改变结构和发展变化的速度和方向。因此，将趋势和周期分开可能会产生误导，无论采用何种复杂的技术进行分析。我们在此采用的方法不是假定某种自主路径，也不是试图对我们的地区序列进行 "去趋势化"，而是重点关注各地区对衰退的反应，衰退的定义是全国经济从高峰到低谷的下行，而复苏的定义是全国经济从低谷到高峰的上行。

To do this, we focus specifically on measuring two of the four dimensions of resilience set out in Figure 1, namely resistance and recoverability, before turning to the role of structure and the issue of adaptability in the next section. Several different measures of resistance and recoverability can be devised. At their simplest, they could just be the straightforward proportionate falls and increases in the chosen indicator (such as regional employment or output). But a more convincing approach is to compare such movements (contractions and expansions) in relation to expected or ‘counterfactual’ falls and increases in the regions concerned. For example, in their study of the resilience of US counties to the recession of 2008-9, Hans and Goetz (2013) compute such expected falls and recoveries by projecting forward into the recession (2008-9) and recovery (2009-2011) what local county output would have been on the basis of its own recent ‘average trend growth rate’, estimated over 2005-8: in effect what would have happened in the absence of the recession. The deviation of the actual level of county output from its corresponding trend-based ‘expected’ level is then used to compute a measure of what they term ‘drop’ and ‘rebound’, and the ratio of the latter to the former is deployed to yield a measure of ‘resilience’. 4
为此，我们在下一节讨论结构的作用和适应性问题之前，将重点放在衡量图 1 所列四个抗灾能力维度中的两个维度，即抵抗力和可恢复性。抵抗力和可恢复性可以有几种不同的测量方法。最简单的是，它们可以是所选指标（如地区就业或产出）的直接比例下降和上升。但更有说服力的方法是将这些变化（收缩和扩张）与相关地区的预期或 "反事实 "下降和上升进行比较。例如，Hans 和 Goetz（2013 年）在研究美国各县对 2008-9 年经济衰退的抵御能力时，通过预测衰退期（2008-9 年）和复苏期（2009-2011 年）当地各县产出在 2005-8 年估计的近期 "平均趋势增长率 "基础上的预期下降和恢复情况：即在没有经济衰退的情况下的实际情况，来计算这种预期下降和恢复情况。然后，利用各县产出的实际水平与其相应的基于趋势的 "预期 "水平之间的偏差来计算他们所称的 "下降 "和 "反弹 "的程度，并利用后者与前者的比率来得出 "恢复力 "的程度。4

Since we are interested in how different regions (or localities or cities) are affected by a common (nation-wide) recession, a particular type of expected or ‘counterfactual’ reaction suggests itself, namely the resistance and recovery of the national economy as a whole. In other words, the expectation is that, other things being equal, each region’s employment would contract (in recessions) and expand (in recoveries) at the same rate as that nationally. Thus the expected change in, say, employment in region $r$$r$rr during a recession or recovery, say of duration $k$$k$kk periods, would be given as
由于我们感兴趣的是不同地区（或地方或城市）如何受到共同的（全国性的）经济衰退的影响，一种特殊的预期反应或 "反事实 "反应就会出现，即整个国家经济的抵抗和复苏。换句话说，在其他条件相同的情况下，每个地区的就业收缩（衰退时）和扩张（复苏时）的速度都与全国相同。因此，在经济衰退或复苏期间，例如在 $k$$k$kk 期间， $r$$r$rr 地区就业的预期变化为

where $g_N^{t+k}$$g_N^{t+k}$g_(N)^(t+k)g_{N}^{t+k} is the rate of contraction (in recession) or expansion (in recovery) of national employment, and $E_{ir}^t$$E_{ir}^t$E_(ir)^(t)E_{i r}^{t} is employment in industry $i$$i$ii in region $r$$r$rr in starting time $t$$t$tt, the base year, that is the turning point into recession or into recovery. 5 Then a measure of regional resistance can be expressed as
其中， $g_N^{t+k}$$g_N^{t+k}$g_(N)^(t+k)g_{N}^{t+k} 是全国就业的收缩率（衰退时）或扩张率（复苏时）， $E_{ir}^t$$E_{ir}^t$E_(ir)^(t)E_{i r}^{t} 是地区 $r$$r$rr 在起始时间 $t$$t$tt （即进入衰退或进入复苏的转折点）的行业 $i$$i$ii 的就业率。5 那么，地区阻力的度量可以表示为

$${\mathrm{Re}}_{\mathrm{sis}}^r=\frac{\left(\mathrm{\Delta }{E}_r^{\text{Contraction}}\right)-{\left(\mathrm{\Delta }{E}_r^{\text{Contraction}}\right)}^{\text{expected}}}{\left|{\left(\mathrm{\Delta }{E}_r^{\text{Contraction}}\right)}^{\text{expected}}\right|}$$$${\mathrm{Re}}_{\mathrm{sis}}^r=\frac{\left(\mathrm{\Delta }{E}_r^{\text{Contraction }}\right)-{\left(\mathrm{\Delta }{E}_r^{\text{Contraction }}\right)}^{\text{expected }}}{\left|{\left(\mathrm{\Delta }{E}_r^{\text{Contraction }}\right)}^{\text{expected }}\right|}$$Re_(sis)^(r)=((DeltaE_(r)^("Contraction "))-(DeltaE_(r)^("Contraction "))^("expected "))/(|(DeltaE_(r)^("Contraction "))^("expected ")|)\operatorname{Re}_{\operatorname{sis}}^{r} \text { }=\frac{\left(\Delta E_{r}^{\text {Contraction }}\right)-\left(\Delta E_{r}^{\text {Contraction }}\right)^{\text {expected }}}{\left|\left(\Delta E_{r}^{\text {Contraction }}\right)^{\text {expected }}\right|}
and a corresponding measure of regional recoverability as
相应的区域可回收性测量值为
$${\mathrm{Recov}}_r=\frac{\left(\mathrm{\Delta }{E}_r^{\text{Recovery}}\right)-{\left(\mathrm{\Delta }{E}_r^{\text{Recovery}}\right)}^{2xpected}}{{\left(\mathrm{\Delta }{E}_r^{\text{Recovery}}\right)}^{\text{expected}}}$$$${\mathrm{Recov}}_r=\frac{\left(\mathrm{\Delta }{E}_r^{\text{Recovery }}\right)-{\left(\mathrm{\Delta }{E}_r^{\text{Recovery }}\right)}^{2xpected}}{{\left(\mathrm{\Delta }{E}_r^{\text{Recovery }}\right)}^{\text{expected }}}$$Recov_(r)=((DeltaE_(r)^("Recovery "))-(DeltaE_(r)^("Recovery "))^(2xpected))/((DeltaE_(r)^("Recovery "))^("expected "))\operatorname{Recov}_{r}=\frac{\left(\Delta E_{r}^{\text {Recovery }}\right)-\left(\Delta E_{r}^{\text {Recovery }}\right)^{2 x p e c t e d ~}}{\left(\Delta E_{r}^{\text {Recovery }}\right)^{\text {expected }}}

By definition, the two measures in Equations (2) and (3) are centred around zero. Thus a positive value of Resis ${}_r$${}_r$_(r)_{r} indicates that a region is more resistant to recession (that is, less affected) than the national economy, and less resistant (more affected) for a negative value. For example, a value of ${\mathrm{Re}\mathrm{sis}}_r$${\mathrm{Re}\mathrm{sis}}_r$Resis_(r)\operatorname{Re~sis}_{r} of, say 0.5 , would indicate that the regional economy in question is 50 percent more resistant than the national economy, and a value of - 0.5 that its resistance is only half that of the nation. Similarly, a region that has a positive value of Recov has a higher recoverability compared to the national economy, and conversely if it has a negative value. This gives a 2x2 configuration of resilience possibilities (Figure 2). On a priori grounds we might expect most regions to fall in either the top right (strong resistance and strong recoverability) or bottom left (weak resistance and weak recoverability) quadrants, although regions that fall in either the top left, and particularly bottom right (weak resistance but strong recoverability) quadrants are also of interest.
根据定义，等式 (2) 和 (3) 中的两个测量值都以零为中心。因此，Resis ${}_r$${}_r$_(r)_{r} 的正值表示一个地区比全国经济更能抵御衰退（即受影响较小），负值则表示抵御衰退的能力较弱（受影响较大）。例如， ${\mathrm{Re}\mathrm{sis}}_r$${\mathrm{Re}\mathrm{sis}}_r$Resis_(r)\operatorname{Re~sis}_{r} 值为 0.5 时，表示该地区经济的抗衰退能力比全国经济强 50%，值为-0.5 时，表示该地区经济的抗衰退能力仅为全国经济的一半。同样，如果一个地区的 Recov 值为正，则其恢复能力高于全国经济，反之，如果该地区的 Recov 值为负，则其恢复能力低于全国经济。这样就形成了一个 2x2 的复原可能性配置（图 2）。根据先验理论，我们可能会认为大多数地区属于右上象限（抵抗力强，可恢复性强）或左下象限（抵抗力弱，可恢复性弱），但也有地区属于左上象限，尤其是右下（抵抗力弱，可恢复性强），这也是我们感兴趣的地方。

Figure 2: Combinations of Resistance and Recoverability
图 2：阻力和可回收性的组合

In using these measures to investigate the impact of the last recessions on the major regions of the UK, we analyse regional employment rather than output, on the grounds
在使用这些衡量标准调查最近几次经济衰退对英国主要地区的影响时，我们分析的是地区就业而不是产出，理由是
that, ultimately, it is a region’s workforce that bears the brunt of adjustment in recessionary contractions: workers laid off during a recession may or may not be rehired when the demand for a region’s products and services begins to recover, and of those who remain unemployed, some may have to move to other regions to find a job, while those who cannot move may well end up as long-term unemployed, or even drop out of the labour force. ${}^6$${}^6$^(6){ }^{6} In general, cyclical movements in employment tend to be more pronounced than those in output, although in the most recent recession the proportionate fall in employment was untypically less than that in output.
${}^6$${}^6$^(6){ }^{6} 在经济衰退收缩时，首当其冲要进行调整的最终是一个地区的劳动力：在衰退期间被解雇的工人在一个地区的产品和服务需求开始恢复时可能会被重新雇用，也可能不会被重新雇用；在那些仍然失业的人中，有些人可能不得不搬到其他地区去找工作，而那些无法搬迁的人很可能最终成为长期失业者，甚至退出劳动力队伍。 ${}^6$${}^6$^(6){ }^{6} 一般来说，就业的周期性变动往往比产出的周期性变动更为明显，尽管在最近的衰退中，就业下降的比例通常低于产出下降的比例。

As mentioned above, there have been four main recession-recovery cycles in the UK economy over the past forty years: namely, from peak to peak, 1974-79; 1979-90; 19902008; and from 2008 onwards, this last cycle being as yet incomplete. In terms of employment movements, there is a striking difference between the two downturns of 19741976 and 2008-2010 on the one hand, and those of 1979-83 and 1990-93 on the other (see Figure 3). The latter two recessions were far more severe and of far greater duration than the other two. The proportionate drop in employment in the downturns of 1979-83 and 1990-93 was around three times that in the recessions of 1974-76 and 2008-10, and employment took twice as long to recover to its pre-recession peak level.
如上所述，在过去四十年中，英国经济主要经历了四个衰退-复苏周期：即从高峰到低谷的 1974-79 年、1979-90 年、1990-2008 年和 2008 年以后，最后一个周期尚未完成。就就业变动而言，1974-1976 年和 2008-2010 年的两次衰退与 1979-83 年和 1990-93 年的两次衰退之间存在显著差异（见图 3）。后两次经济衰退比前两次严重得多，持续时间也长得多。在 1979-83 年和 1990-93 年的经济衰退中，就业人数的下降比例约为 1974-76 年和 2008-10 年经济衰退的三倍，而就业人数恢复到衰退前峰值水平所需的时间是前者的两倍。

Figure 3: UK Employment over the Last Four Recession-Recovery Cycles, 1971-2014
图 3：1971-2014 年英国过去四个衰退-复苏周期的就业情况

Source of data: Office for National Statistics
数据来源国家统计局
Note: The most recent cycle is as yet incomplete
注：最近的周期尚未完成

Even a casual inspection of the employment paths of the major UK regions, shown by the annual series in Figure 4, suggests that both the intensity of employment contraction in these four recessions and the speed and extent of the subsequent recovery expansions have varied across regions. 7 Furthermore, it is apparent that these differential responses are
图 4 中的年度数列显示了英国主要地区的就业轨迹，即使是对这些地区的就业轨迹进行粗略的考察，也可以看出，这四次经济衰退中就业收缩的强度以及随后复苏扩张的速度和程度在不同地区都是不同的。7 此外，这些不同的反应显然是

Figure 4: Employment in the UK Regions over the Last Four Recession-Recovery Cycles, 1971-2014
图 4：1971-2014 年英国各地区在过去四个衰退-复苏周期中的就业情况

Source of data: Cambridge Econometrics Regional Economic Data Base
数据来源剑桥计量经济学区域经济数据库
Note: National recessionary downturns are shown shaded
注：国家衰退期的下滑用阴影表示
inextricably bound up with quite divergent long-run employment trajectories among the regions. Thus it would appear regional employment paths diverged during both of the 1974-76 and 1979-83 recessions but noticeably less so in the subsequent downturns of 1990-93 and 2008-2010. At the same time, significant divergence took place in the
这与各地区长期就业轨迹的巨大差异密不可分。因此，在 1974-76 年和 1979-83 年两次经济衰退期间，各地区的就业轨迹似乎出现了分化，但在随后的 1990-93 年和 2008-2010 年经济衰退期间，分化程度明显降低。与此同时，在 1974-76 年和 1979-83 年的经济衰退中，各地区的就业轨迹也出现了明显的分化。

recoveries of 1976-1979 and 1983-1990 and has also occurred since 2010, but much less so in the long recovery-upswing phase of 1993-2008. The net result has been an overall marked cumulative divergence of employment growth paths amongst the regions, and especially between those in the south and those in the north of the country. There has long been discussion over the persistence, if not widening, of disparities in economic fortune between the south and east of the nation on the one hand, and the more northern areas on the other, often under the contentious and contested sobriquet of a ‘North-South Divide’ (see, for example, Martin 1988, 2004; Jackman and Savouri, 1999; Rowthorn, 2010; Gardiner, Martin, Sunley and Tyler, 2013). The employment evolutions in Figure 4 tend to support the idea of this broad divide. On the one hand, the South East, East of England, East Midlands and South West, have all expanded their employment by between 40-60 percent over the whole 1971-2014 period. Regions in the ‘northern’ group have experienced much slower employment growth, at best around 20-25 percent (in Wales and Yorkshire-Humberside), at worst barely any increase at all (in the North West and North East). Two regions complicate this ‘North-South divide’, however. Thus Northern Ireland has matched the four southern regions in terms of its rate of job growth, while up until the early-1990s recession, London, often regarded as a leading southern region, was in fact the worst performing region of all in terms of employment growth, and experienced - in common with the North East and North West regions - an overall decline in employment up to that juncture. Since then it has undergone a significant turnaround (see Gardiner, Martin, Sunley and Tyler, 2013; Martin, 2014), recording a growth rate on a par with that in regions in the southern group, though not sufficient to yet lift it into that group in terms of its long-run cumulative growth position.
1976-1979年和1983-1990年的经济复苏和2010年以来的经济复苏也出现了这种情况，但在1993-2008年的长期复苏上升阶段则要少得多。最终结果是，各地区之间，特别是该国南部和北部地区之间的就业增长路径总体上出现了明显的累积差异。长期以来，人们一直在讨论该国南部和东部地区与北部地区之间经济财富差距的持续存在（如果不是不断扩大的话），这种差距通常被称为 "南北鸿沟"（见，例如，Martin，1988 年，2004 年；Jackman 和 Savouri，1999 年；Rowthorn，2010 年；Gardiner、Martin、Sunley 和 Tyler，2013 年）。图 4 中的就业演变趋势支持了这一广泛分歧的观点。一方面，在整个 1971-2014 年期间，东南部、英格兰东部、东米德兰兹和西南部的就业人数都增加了 40% 至 60%。而 "北部 "地区的就业增长则缓慢得多，最好的时候约为 20%-25%（威尔士和约克郡-亨伯赛德郡），最差的时候几乎没有任何增长（西北部和东北部）。然而，有两个地区使这种 "南北分化 "更加复杂。因此，北爱尔兰在就业增长率方面与南部四个地区不相上下，而直到 20 世纪 90 年代初的经济衰退，通常被认为是南部领先地区的伦敦，实际上在就业增长方面是所有地区中表现最差的，而且与东北部和西北部地区一样，到那时为止，伦敦的就业率整体下降。 从那时起，它经历了一个重大转折（见 Gardiner、Martin、Sunley 和 Tyler，2013 年；Martin，2014 年），增长率与南方组的地区相当，尽管就其长期累计增长状况而言，还不足以使其进入南方组。

The resistance and corresponding recoverability indices for the UK regions, calculated using Equations (2) and (3), for the last four recession-recovery cycles, are depicted in Figure 5, where each recession-recovery cycle is portrayed in the form of the fourquadrant typology shown in Figure $2.{}^8$$2.{}^8$2.^(8)2 .{ }^{8} Several striking features are evident. The most obvious is that, apart from the cycle of recession-recovery cycle of 1990-2008, there has been a positive relationship between resistance and recoverability across regions: that is, regions that have been most resistant to recession have also enjoyed the strongest recoverability. However, this relationship is not particularly strong, with that between the recession of 1979-83 and the subsequent recovery of 1983-1990 being the most significant. The recession-recovery cycle of 1990-2008 emerges as quite different from the others. Secondly, and more notably, there are certain regions - specifically the South East, East of England and South West - that tend to recover strongly (in relation to the national economy) more or less regardless of their resistance to recession. At the same time, there other regions - the North East, Scotland and the North West - that tend to display weak
图 5 描述了使用公式 (2) 和 (3) 计算的英国各地区在过去四个衰退-复苏周期中的阻力指数和相应的可恢复性指数，其中每个衰退-复苏周期都以图 $2.{}^8$$2.{}^8$2.^(8)2 .{ }^{8} 中所示的四象限类型学的形式加以描述。最明显的是，除了 1990-2008 年这一轮衰退-复苏周期外，各地区的抗衰退能力与复苏能力之间一直存在正相关关系：即抗衰退能力最强的地区复苏能力也最强。然而，这种关系并不是特别强，1979-1983 年的衰退与随后 1983-1990 年的复苏之间的关系最为显著。1990-2008 年的衰退-复苏周期与其他周期截然不同。其次，更值得注意的是，某些地区--特别是东南部、英格兰东部和西南部--或多或少地倾向于强劲复苏（相对于全国经济而言），而不论其抵御衰退的能力如何。与此同时，其他一些地区--东北部、苏格兰和西北部--往往表现出疲软的趋势。

recoverability regardless of their resistance to recession. Thirdly, there are some regions that have undergone marked shifts in resilience from one recession-recovery cycle to the next, namely London, Northern Ireland and the West Midlands.
第三，有些地区在不同的衰退-复苏周期中的恢复能力发生了显著变化，如伦敦、北爱尔兰和西米德兰兹。第三，有一些地区，即伦敦、北爱尔兰和西米德兰兹地区，从一个衰退-复苏周期到下一个衰退-复苏周期，其复原力发生了明显变化。

A fourth characteristic is that since the early-1970s, at least until the most recent recession and recovery, differences in resilience across UK regions appear to have narrowed (compare the horizontal and vertical scales of the graphs for successive cycles), which begs the question of whether this trend itself reflects a convergence in certain structural or other economic attributes of the regions over this period. The halt to this trend and the widening of regional differences in resilience over the most recent cycle, 2008-2014, is mainly due to the particular geography of recovery since 2010. This has been emphatically led by London, and the slower recovery in much of northern Britain could well be associated with the greater dependence of employment there on the public sector, which was the target of the Coalition Government’s fiscal austerity programme. The continuation of this programme of cutting public expenditure under the Conservative Government elected in 2015 may well further slow the recoverability of this part of the UK.
第四个特点是，自 20 世纪 70 年代初以来，至少在最近的经济衰退和复苏之前，英国各地区的抵御能力差异似乎有所缩小（比较各周期图表的横坐标和纵坐标），这就引出了一个问题，即这一趋势本身是否反映了这一时期各地区某些结构或其他经济属性的趋同。在最近的周期（2008-2014 年）中，这一趋势的停止以及地区间抗灾能力差异的扩大，主要是由于 2010 年以来复苏的特殊地理位置。这主要是由伦敦主导的，而英国北部大部分地区的复苏较慢，很可能与那里的就业更依赖公共部门有关，而公共部门正是联合政府财政紧缩计划的目标。2015 年当选的保守党政府将继续实施这一削减公共开支的计划，这很可能会进一步减缓英国北部地区的复苏速度。

Figure 5: Regional Resistance and Recoverability for the Last Four Economic Cycles, UK
图 5：英国过去四个经济周期的地区阻力和可恢复性

Notes: Calculated using equations (2) and (3). The cycle for 2008-2014 is as yet incomplete (see Footnote 7).
注使用公式(2)和(3)计算。2008-2014 年的周期尚未完成（见脚注 7）。

What is clear from Figure 5, and from Tables 1 and 2, is that these findings caution against any idea that regional resilience is some fixed attribute or capacity, and hence against reading off any definite conclusions about regional resilience from an analysis of just one single recession or cycle, or from ‘averaging’ across cycles. ${}^9$${}^9$^(9){ }^{9} For one thing, recessions differ in their cause and nature, and for this reason alone are likely to impact differently across regions and localities. In the UK, the last four recessions have differed in their proximate causes: the oil crisis induced recession of 1974-76 impacted particularly on manufacturing, and the recession of 1979-82 even more so, that of 1990-93 also hit private services, while that of 2008-2010 was largely attributable to the sharp credit squeeze triggered by the financial crisis. Regional reactions to these different types of recession might be expected to have varied accordingly. But in addition, longer term processes and developments may also change regional resistance to and recovery from successive recessionary contractions. In this context, given that our study period spans 40 years, marked changes have inevitably occurred to the economies of the regions, including major shifts in industrial structures. The impact of the latter, in particular, warrants investigation, and this is the focus of the remainder of the paper.
从图 5 以及表 1 和表 2 中可以清楚地看出，这些发现告诫人们不要认为地区韧性是某种固定的属性或能力，因此也不要从对单一衰退或周期的分析中，或从各周期的 "平均值 "中得出任何有关地区韧性的明确结论。 ${}^9$${}^9$^(9){ }^{9} 首先，经济衰退的原因和性质各不相同，因此很可能对不同地区和地方产生不同的影响。在英国，过去四次经济衰退的直接原因各不相同：1974-1976 年石油危机引发的衰退对制造业的影响尤为严重，1979-1982 年的衰退影响更大；1990-1993 年的衰退也打击了私营服务业；而 2008-2010 年的衰退则主要归因于金融危机引发的信贷紧缩。预计各地区对这些不同类型衰退的反应也会相应不同。但此外，长期的进程和发展也可能改变地区对连续衰退收缩的抵御能力和复苏能力。在这种情况下，鉴于我们的研究时间跨度长达 40 年，各地区的经济不可避免地发生了显著变化，包括产业结构的重大转变。后者的影响尤其值得研究，这也是本文其余部分的重点。

Table 2: Correlations between Regional Recoverability to Different Recessions
表 2：不同经济衰退地区复苏能力之间的相关性

Note: p-values in parentheses
注：括号内为 p 值

The resilience of regional and local economies to recessions is determined by a complex array of factors (see Figure 6). As depicted in Figure 1, these factors shape the vulnerability of a region’s economy to recessionary shocks, its resistance to such shocks, its adaptability, and its recoverability. In the process, those very same characteristics and factors may themselves undergo permanent change and transformation so that regional reactions to the next recessionary shock may alter accordingly (see Martin and Sunley, 2015). It is in this context that the notions of resilience as ‘positive adaptation’ or as a new developmental pathway (Figure 1) are relevant. Many of these factors - around labour markets, financial systems and governance arrangements - might be described as ‘institutional’ in so far as they represent more durable patterns and combinations of formal and informal ways of organizing economic activity. However, most recent studies have related these institutional dimensions to long-run development rather than to responses to adverse shocks (see Gertler, 2010; Rodriguez Pose, 2013).
区域和地方经济抵御衰退的能力由一系列复杂的因素决定（见图 6）。如图 1 所示，这些因素决定了一个地区经济对衰退冲击的脆弱性、对这种冲击的抵抗力、适应力和恢复力。在此过程中，这些相同的特征和因素本身可能会发生永久性的变化和转变，从而使地区对下一次衰退冲击的反应可能会相应改变（见 Martin 和 Sunley，2015 年）。正是在这种情况下，作为 "积极适应 "或新的发展途径（图 1）的复原力概念才具有相关性。其中许多因素--围绕劳动力市场、金融体系和治理安排--可被称为 "制度性 "因素，因为它们代表了组织经济活动的正式和非正式方式的更持久模式和组合。然而，最近的大多数研究都将这些制度层面与长期发展联系起来，而不是与应对不利冲击联系起来（见 Gertler，2010 年；Rodriguez Pose，2013 年）。

Figure 6: Determinants of Regional Economic Resilience
图 6：区域经济韧性的决定因素

Isolating the contribution of the multifarious determinants shown in Figure 6 to resilience would be a major task, assuming that the requisite data were even available, and would be best accomplished by a case-study orientated research strategy. Our aim here is much more modest. Of the various factors assumed to exert a formative influence on the sensitivity of regional and local economies to cyclical shocks and perturbations, industrial structure has frequently been assigned key importance since it shapes a region’s exports and hence exposure to externally-originating demand shocks. Indeed, in the past couple of years or so, the issue of regional economic structure has attracted renewed discussion and debate. In essence, the key question is whether specialisation is ‘good or bad’ for regional development. According to some economic geographers, industrial specialisation is the major driver of regional economic growth (Storper, 2013; Storper et al, 2015). Others, however, argue that it is the diversity - the complexity - of a region’s or city’s economic structure (its industries or its products) that imbues it with higher growth and greater stability (see for example, Hausmann et al, 2013). Still others argue that what matters for regional economic growth and stability over the long run is ‘diversified specialisation’ (Farhauer and Kröll, 2011), while yet others emphasise what they call ‘related variety’ (Frenken, Van Oort and Verburg, 2007). Also, other possible determinants - such as the age structure of local firms, wage costs, workforce skills, workplace cultures, entrepreneurship and business formation rates, and the like - are to some extent themselves functions of a region’s industrial composition and its industrial history. Indeed, some of these industry-related factors may linger on, in a path dependent manner, even if the particular industries around which they developed and with which they were originally associated, have long since declined or disappeared.
如果能够获得必要的数据，那么将图 6 所示的各种决定因素与复原力联系起来将是一项艰巨的任务。我们在此的目标要温和得多。在假定会对区域和地方经济对周期性冲击和扰动的敏感性产生形成性影响的各种因素中，产业结构经常被赋予关键的重要性，因为它决定了一个地区的出口，从而使其面临来自外部的需求冲击。事实上，在过去几年左右的时间里，区域经济结构问题又引起了新的讨论和争论。从本质上讲，关键问题在于专业化对地区发展是 "好是坏"。一些经济地理学家认为，产业专业化是区域经济增长的主要驱动力（Storper，2013；Storper et al，2015）。而另一些学者则认为，一个地区或城市经济结构（产业或产品）的多样性--复杂性--才会使其获得更高的增长和更大的稳定性（参见 Hausmann 等人，2013 年）。还有人认为，从长远来看，对地区经济增长和稳定至关重要的是 "多样化的专业化"（Farhauer 和 Kröll，2011 年），还有人强调所谓的 "相关多样性"（Frenken、Van Oort 和 Verburg，2007 年）。此外，其他可能的决定因素，如当地企业的年龄结构、工资成本、劳动力技能、工作场所文化、创业精神和企业组建率等，在某种程度上本身就是地区产业构成和产业历史的函数。 事实上，其中一些与产业相关的因素可能会以路径依赖的方式持续存在，即使这些因素所围绕和最初与之相关的特定产业早已衰落或消失。

In his interesting study, Conroy (1973) demonstrates in some detail how a region’s economic structure - what he calls its ‘industrial portfolio’, the particular mix of economic activities and their interrelationships - can influence the reactions of a region’s economy to recessionary disturbances and fluctuations. Subsequent studies have likewise assigned a similar importance to regional economic structure, and to the possible implications of structural diversity (or variety) versus specialisation for cyclical stability (for example, Sherwood-Call, 1990; Siegel et at, 1994, 1995; Dissart, 2003; Ormerod, 2008). A recurring theme is that, other things being equal, a diverse or varied economic structure confers greater regional resistance to shocks than does a more specialised structure, since different industries themselves have different elasticities of demand, different export markets, different dependence on monetary conditions (exchange rates, interest rates, debtfinancing), and so on (see, Garcia-Mila and McGuire, 1993; Belke and Heine, 2004). Put another way, a diverse economic structure should allow a regional economy to ‘spread risk’. Conversely, a highly specialised regional economy - say one heavily dependent on manufacturing - is potentially more vulnerable and unstable, since if its principal industrial specialism is badly hit by a downturn it has much less scope for other, less-
康罗伊（Conroy，1973 年）在其有趣的研究中详细说明了一个地区的经济结构--他称之为 "产业组合"，即经济活动的特定组合及其相互关系--如何影响一个地区的经济对衰退干扰和波动的反应。随后的研究也同样重视区域经济结构，以及结构多样性（或多样性）与专业化对周期稳定性的可能影响（例如，Sherwood-Call，1990 年；Siegel et at，1994 年，1995 年；Dissart，2003 年；Ormerod，2008 年）。一个经常出现的主题是，在其他条件相同的情况下，多元化或多样化的经济结构比专业化结构更能增强地区抵御冲击的能力，因为不同产业本身具有不同的需求弹性、不同的出口市场、对货币条件（汇率、利率、债务融资）的不同依赖性等（见 Garcia-Mila 和 McGuire，1993 年；Belke 和 Heine，2004 年）。换句话说，多样化的经济结构应允许区域经济 "分散风险"。相反，高度专业化的地区经济--比如说严重依赖制造业的地区经济--可能会更加脆弱和不稳定，因为如果其主要的专业产业受到经济衰退的严重打击，那么其他不太专业的产业的发展空间就会小得多。
sensitive (more resistant), industries to provide some measure of buffering against the contraction. For these reasons, according to Davies and Tonts, 2010):
这些原因是，根据 Davies 和 Tonts，2010 年）："我们的产业对收缩的敏感性（更强的抵抗力）更强。由于这些原因，Davies 和 Tonts 认为）

The general contention is that those places with diverse economies are more resilient in socio-economic terms than those with a narrow economic base (p. 232).
一般认为，经济多样化的地方比经济基础狭窄的地方在社会经济方面更有弹性（第 232 页）。

Table 3: Sectoral Employment Growth Rates and ‘Cyclical Sensitivity’, UK, 1971-2014
表 3：部门就业增长率和 "周期敏感性"，英国，1971-2014 年

Note: The ‘coefficient of cyclical sensitivity’ is the regression coefficient of the percentage change in employment in an industry ( $\mathrm{\%}\mathrm{\Delta }{E}_i$$\mathrm{\%}\mathrm{\Delta }{E}_i$%DeltaE_(i)\% \Delta E_{i} ) on the percentage change in national total employment ( $\mathrm{\%}\mathrm{\Delta }{E}_N$$\mathrm{\%}\mathrm{\Delta }{E}_N$%DeltaE_(N)\% \Delta E_{N} ), calculated for the whole period, 1971-2014.
注："周期敏感性系数 "是一个行业的就业人数变化百分比（ $\mathrm{\%}\mathrm{\Delta }{E}_i$$\mathrm{\%}\mathrm{\Delta }{E}_i$%DeltaE_(i)\% \Delta E_{i} ）对全国总就业人数变化百分比（ $\mathrm{\%}\mathrm{\Delta }{E}_N$$\mathrm{\%}\mathrm{\Delta }{E}_N$%DeltaE_(N)\% \Delta E_{N} ）的回归系数，计算时间跨度为 1971-2014 年。

As Table 3 shows, not only have different UK sectors of activity had different rates of employment growth or decline over the 1971-2014 period as a whole, they have also
如表 3 所示，1971-2014 年期间，英国不同活动部门的就业增长或下降率各不相同，而且
differed in their ‘cyclical sensitivity’. ${}^{\mathbf{1}\mathbf{0}}$${}^{\mathbf{1}\mathbf{0}}$^(10){ }^{\mathbf{1 0}} For example, manufacturing and production activities tend to have high cyclicality, motor vehicles and construction being prime examples, while employment growth in services and utilities tends to much more stable over time, as evidenced particularly by electricity, gas and water, retailing, health, education, and other services. There is evidence of an inverse relationship between ‘cyclical sensitivity’ and growth: those sectors with the highest long-term average growth rates tend to be the least cyclically sensitive, indicating that buoyant growth tends to be associated with greater resistance to shocks. These attributes give some credence to the argument that within a regional setting, a diverse industrial structure might be expected to be less vulnerable and resilient to recessions. However, specialisation need not be problematic per se: much depends on the particular cyclical vulnerability or sensitivity of that specialisation. A region that is specialised in manufacturing, for instance, is likely to be much more affected by recessions, other things being equal, than is a region that has a disproportionate dependence on public services - or at least this was the case until the new era of fiscal austerity, when public sector activities have become much more vulnerable to shocks, both in and out of recession.
周期敏感性 "不同。 ${}^{\mathbf{1}\mathbf{0}}$${}^{\mathbf{1}\mathbf{0}}$^(10){ }^{\mathbf{1 0}} 例如，制造业和生产活动往往具有较高的周期性，机动车辆和建筑业就是最好的例子，而服务业和公用事业的就业增长在一段时间内往往要稳定得多，电力、天然气和水、零售、卫生、教育和其他服务业尤其如此。有证据表明，"周期敏感性 "与增长之间存在反比关系：长期平均增长率最高的部门往往对周期的敏感性最低，这表明强劲的增长往往与更强的抗冲击能力有关。这些特征在一定程度上证明了这样一种观点，即在一个地区环境中，多样化的产业结构可能会更不容易受到经济衰退的影响，并具有更强的抵御能力。然而，专业化本身并不一定有问题：很大程度上取决于该专业化的特定周期脆弱性或敏感性。例如，一个专门从事制造业的地区，在其他条件相同的情况下，受到经济衰退的影响可能要比过度依赖公共服务的地区大得多--至少在新的财政紧缩时代之前是这样，因为在新的财政紧缩时代，公共部门的活动变得更容易受到冲击，无论是在衰退中还是衰退后。

Empirically disentangling the influence of differences in industrial structure on regional responses to recessions is not straightforward. In their study of regional cyclical synchronicities across Europe, Partridge and Rickman (2005) first extract the trend from each region’s employment series to generate a regional ‘cyclical series’, to which they then apply what is essentially a shift-share decomposition to identify the contribution of industrial structure to each region’s cyclicity. For the reasons we enumerated above, however, distinguishing trend from cycles may not be ontologically meaningful, so we do not adopt that approach here. Further, since, as we saw in the previous section, a region’s reaction to and recover from recessionary shocks may well vary from one cycle to another, and since this may be due to changes in regional industrial structures over time - perhaps as a result of successive recessionary shocks themselves - we wish to identify the impact of industrial structure on regional resistance separately for each downturn and each subsequent recovery upswing.
要从经验上厘清产业结构差异对地区经济衰退反应的影响并非易事。Partridge 和 Rickman（2005 年）在研究欧洲各地区的周期同步性时，首先从各地区的就业序列中提取趋势，生成地区 "周期序列"，然后对该序列进行本质上的转移份额分解，以确定产业结构对各地区周期性的贡献。然而，由于我们在上文列举的原因，区分趋势和周期可能没有本体意义，因此我们在此不采用这种方法。此外，正如我们在上一节中所看到的，一个地区对衰退冲击的反应和从衰退冲击中的恢复在不同周期之间可能会有所不同，而这可能是由于地区产业结构随着时间的推移发生了变化--也许是连续的衰退冲击本身的结果--因此，我们希望针对每次衰退和随后的每次复苏上升分别确定产业结构对地区抵抗力的影响。

More specifically, and in order to link the analysis with the resistance and recovery indices computed in Section 2, we apply a shift-share procedure to decompose the change in a region’s employment during a recession or recovery period, $\mathrm{\Delta }{E}_r^{\text{recess}}$$\mathrm{\Delta }{E}_r^{\text{recess }}$DeltaE_(r)^("recess ")\Delta E_{r}^{\text {recess }} or $\mathrm{\Delta }{E}_r^{\text{recov}}$$\mathrm{\Delta }{E}_r^{\text{recov }}$DeltaE_(r)^("recov ")\Delta E_{r}^{\text {recov }} into three components: a national component, an industrial mix effect and a regional ‘shift’ or regional ‘competitiveness’ effect. For a given change over the time interval $t,t+k$$t,t+k$t,t+kt, t+k and for a given industry $i$$i$ii in region $r$$r$rr, the shift-share decomposition can be written as
更具体地说，为了将分析与第 2 节中计算的阻力指数和复苏指数联系起来，我们采用了转移份额程序，将衰退期或复苏期的地区就业变化 $\mathrm{\Delta }{E}_r^{\text{recess}}$$\mathrm{\Delta }{E}_r^{\text{recess }}$DeltaE_(r)^("recess ")\Delta E_{r}^{\text {recess }} 或 $\mathrm{\Delta }{E}_r^{\text{recov}}$$\mathrm{\Delta }{E}_r^{\text{recov }}$DeltaE_(r)^("recov ")\Delta E_{r}^{\text {recov }} 分解为三个部分：全国部分、产业组合效应和地区 "转移 "或地区 "竞争力 "效应。对于时间区间 $t,t+k$$t,t+k$t,t+kt, t+k 的给定变化和地区 $r$$r$rr 中的给定产业 $i$$i$ii 来说，转移份额分解可写成

$$\mathrm{\Delta }{E}_r^{t+k}=\sum _i\left(g_N^{t+k}{E}_{ir}^t\right)+\sum _i(g_{iN}^{t+k}-g_N^{t+k})E_{ir}^t+\sum _i(g_{ir}^{t+k}-g_{iN}^{t+k})E_{ir}^t$$$$\mathrm{\Delta }{E}_r^{t+k}=\sum _i \left(g_N^{t+k}{E}_{ir}^t\right)+\sum _i \left(g_{iN}^{t+k}-g_N^{t+k}\right)E_{ir}^t+\sum _i \left(g_{ir}^{t+k}-g_{iN}^{t+k}\right)E_{ir}^t$$DeltaE_(r)^(t+k)=sum_(i)(g_(N)^(t+k)E_(ir)^(t))+sum_(i)(g_(iN)^(t+k)-g_(N)^(t+k))E_(ir)^(t)+sum_(i)(g_(ir)^(t+k)-g_(iN)^(t+k))E_(ir)^(t)\Delta E_{r}^{t+k}=\sum_{i}\left(g_{N}^{t+k} E_{i r}^{t}\right)+\sum_{i}\left(g_{i N}^{t+k}-g_{N}^{t+k}\right) E_{i r}^{t}+\sum_{i}\left(g_{i r}^{t+k}-g_{i N}^{t+k}\right) E_{i r}^{t}
(a)
(b)
©
where $g_N^{t+k}$$g_N^{t+k}$g_(N)^(t+k)g_{N}^{t+k}, is the percentage rate of change of national employment over the period of recession or recovery, $g_{iN}^{t+k}$$g_{iN}^{t+k}$g_(iN)^(t+k)g_{i N}^{t+k} is the corresponding rate of change of employment in sector $i$$i$ii nationally, and $g_{ir}^{t+k}$$g_{ir}^{t+k}$g_(ir)^(t+k)g_{i r}^{t+k} the rate of change of employment in that sector in region, $r$$r$rr; and where
其中， $g_N^{t+k}$$g_N^{t+k}$g_(N)^(t+k)g_{N}^{t+k} 是衰退或复苏期间全国就业人数的百分比变化率， $g_{iN}^{t+k}$$g_{iN}^{t+k}$g_(iN)^(t+k)g_{i N}^{t+k} 是全国 $i$$i$ii 部门就业人数的相应变化率， $g_{ir}^{t+k}$$g_{ir}^{t+k}$g_(ir)^(t+k)g_{i r}^{t+k} 是地区该部门就业人数的变化率， $r$$r$rr ；以及
(a) is the national effect or component, which is the change in regional employment in a downturn (or recovery) that would be expected if employment in each of the region’s industries contracted (or increased) at the same rate as the national rate of employment change. In other words, this is identical to the ‘expected’ (counterfactual) regional response ${\left(\mathrm{\Delta }{E}_i^{t+k}\right)}^e$${\left(\mathrm{\Delta }{E}_i^{t+k}\right)}^e$(DeltaE_(i)^(t+k))^(e)\left(\Delta E_{i}^{t+k}\right)^{e} that was used in equations (2) and (3);
(a) 是全国效应或组成部分，即在经济下滑（或复苏）时，如果该地区各行业的就业以与全国就业变化率相同的速度收缩（或增加），则该地区就业的预期变化。换句话说，这与等式 (2) 和 (3) 中使用的 "预期"（反事实）地区反应 ${\left(\mathrm{\Delta }{E}_i^{t+k}\right)}^e$${\left(\mathrm{\Delta }{E}_i^{t+k}\right)}^e$(DeltaE_(i)^(t+k))^(e)\left(\Delta E_{i}^{t+k}\right)^{e} 相同；
(b) is the industrial mix effect, which is the fall (or increase) in regional employment in a recession (or recovery) allowing for the fact that some national industries may contract in recession or expand in recovery at faster rates than national employment as a whole. It thus captures that part of a region’s relative resistance to and recoverability from recession attributable to the region’s mix of (specialisation in) more and less nationally resilient industries. The more a region’s economic structure is dominated by sectors that nationally are more (or less) resilient to recessions, the more (or less) resilient will the region be to recession compared to the national economy;
(b) 产业组合效应，即在衰退（或复苏）中地区就业人数的减少（或增加），考虑到某些全国性产业在衰退中的收缩或在复苏中的扩张速度可能快于全国就业人数的整体速度。因此，它反映了一个地区对衰退的相对抵抗力和从衰退中复苏的能力，这部分能力可归因于该地区（专业化程度较高和较低的）全国性抗衰产业的组合。与全国经济相比，一个地区的经济结构越是以那些在全国范围内抵御衰退能力较强（或较弱）的行业为主，该地区抵御衰退的能力就越强（或越弱）；
© is the difference between actual regional employment change in recession or recovery and that part of regional employment change attributable to national and industrial mix effects, and captures differences in behaviour between the industries in a region and those same industries nationally. In standard shift share analysis it is usually assumed to reflect features and characteristics specific to a region (socalled ‘competitiveness effects’) that result in its industries behaving differently from their national counterparts.
是衰退或复苏时期实际地区就业变化与国家和产业组合效应导致的地区就业变化之间的差额，反映了一个地区的产业与全国相同产业之间的行为差异。在标准的转移份额分析中，通常假定它反映了一个地区特有的特征和特点（即所谓的 "竞争力效应"），这些特征和特点导致该地区的产业表现与全国同类产业不同。

By taking the national component (a) to the left-hand side of (4), that is,
将本国成分 (a) 加入 (4) 的左侧，即

we are in effect decomposing the numerator in the resistance and recovery measures in Equations (2) and (3) into an industrial structure effect and a regional ‘competitiveness’ effect. Whereas standard shift share decomposes the change in regional employment over a specified period using the initial industrial structure and holding that constant over the
实际上，我们将等式(2)和(3)中阻力和复苏措施的分子分解为产业结构效应和地区 "竞争力 "效应。标准转移份额利用初始产业结构分解特定时期内的地区就业变化，并在这一时期内保持不变。