2016년 2월 16일 화요일

[B. Christophers's Banking Across Boundaries] 3. Enclosing the Unproductive (2013)




※ 발췌 (excerpt):

3. Enclosing the Unproductive  (p. 103)

In the 1920s, and then with an even greater intensity of force and effect in the 1930s, the world of international money and banking shrunk back from its previously open and integrated financial islands. It would remain thus through to at least the end of the 1950s. Only from the beginning of the 1960s did the accordion begin to de-compress, and global finance and banking begin the long and uneven process of disassembling the territorial borders that had been erected during the preceding 30 to 40 years.

  The period of assembly and defense of these financial borders provides the immediate historical backdrop to this chapter. Its objective, however, is not simply to offer a narrative of key events, or, more ambitiously, to offer explanations as to why things happened the way they did. Plenty of fine existing studies cover both of these bases, and I draw on them extensively here. My task, rather, is to trace the evolving relationship between developments in financial boundary construction and developments in economic ideas about such boundary─particularly insofar as those ideas also pertained to another type of boundary, namely that which separated economic "productiveness" from unproductiveness. This relationship, I argue, became increasingly close and increasingly material during the period in question; and it is a relationship to which the existing literature on historical geographies of finance and banking does not do justice.

  The essential thrust of the argument can be summarized as follows. When the financial world fractured spatially beginning in World War I, it did so reactively and pragmatically; this was a retrenchment driven by immediate, pressing political-economic exigencies, not one occurring in response to the formulation, mobilization, and application of economic ideas. Yet, critically, in the 1930s and 1940s, various socio-technical apparatuses of economic representation came to suggest that this protectionism in finance and banking, if not originally based on economic theory, was nonetheless economically right and proper; and they did so specially through recuperation of 19th-century political economy's productive/unproductive dichotomy, albeit with "labour" epithet now conspicuously dropped. This recuperation took two forms. First, and as previously noted by a handful of economic historians, there arose a conceptual distinction between productive and unproductive flows of capital. This distinction, moreover, was invoked not only to justify a posteriori the inter-state capital controls that already existed in the inter-war period, but also─and arguably more importantly─to rationalize a priori the selective maintenance of such controls in the post-war period. The second form of recuperation of the productive-unproductive distinction, meanwhile, occurred in relation to economically-active institutions and the products or services they offered. The primary setting for this recuperation upon envisioning of a "production boundary" dividing productive activities from unproductive ones. The common determination, in the early days of national accounts, that banking was economically unproductive, fitted with the prevailing protectionist world even if it did not actively underwrites it. And in all of this, I show, the figure of John Maynard Keynes loomed especially large. For it was in his writing, his policy intervention, and his shaping of emergent systems of statistical representation that a melding of the twin arguments about productiveness─of capital flows and of banks─both with one another, and in turn with arguments about financial internationalization, began to achieve their clearest expression.

  Given the various interconnected dimensions of this argument, I lay it out here in a series of stages. The first part of the chapter looks, in overview, at the raw facts of what happened in international finance and banking from the 1920s through to the 1950s. The basic story, as indicated, is one of national banking and monetary systems turning in on themselves. There was, however, a stuttering nature to this process of enclosure─assorted capital controls imposed during or immediately after World War I were in many cases lifted in the mid- to late 1920s, before being reinstalled in the following decade─and the types of boundaries implanted varied considerably across both time and space. Relying on the extensive secondary literature on such matters, I draw out the main developments and the explanations that have been posited for these.

  The second section examines the first of the ways in which a distinction between economic productiveness and unproductiveness was pictured and, more pointedly, was yoked to the question of financial protectionism. ( ... ... )

  A fuller analysis follows in the third part of the chapter, whose subject is the birth, in the US and various parts of Europe in the 1930s and 1940s, of systematic, government-sponsored attempts to create economic accounts for nation-states. I show that a division between productive and unproductive economic activities, which had been central to classical political economy but theoretically dismissed by ascendant neoclassicism of the early 20th century, was fundamental to the generation of these accounts─and especially to the one method of calculating national economic output (the "output" or "product" method) that allowed the relative contributions of different industry sectors to this output to be confidently measured. I discuss the political, economic, and intellectual contexts in which the birth of national accounting occurred, for these were not immaterial. And I explain why, with national accounts having since become over time, a crucial tool for the demonstration of economic and social worth, a critical understanding of the "productive boundary" that traverses such accounts is so important.

  The last two sections of the chapter address the more specific question of the place of financial services in general, and banking in particular, in the differentiations effected and the representations generated by national accounts. To distill this question to its pure essence, which side of the production boundary did banks sit on? The answer to this question, we shall see, has been multi-dimensional and manifestly problematic ever since the time of its initial posing in the national accounting context in the 1930s. 
  • Section 4 closely considers the nature of the question and the nature of the problems it elicited. These problems are reducible, in large part, to the fact that if banking services and revenues were treated in the national accounts in the same way as other companies' services and revenues, the accounts would show banks as being only negligibly positive contributors to national economic output or, more likely, as negative contributors─or unproductive "leeches" on the income stream, as one national income statistician would late, famously, put it. My discussion focuses on how and why this "problem" arose and on why─for it is not necessarily self-evident─this was generally (though certainly not always) regarded as a problem.
  • The final section of the chapter analyzes the different responses to this "problem" in selected countries, by those countries' national income statisticians. For, in the early days of national accounting, international harmonization of accounting standards was limited and many different methods were employed, ^particularly^ vis-a-vis the thorny question of accounting or banks. ( ... ... ) The response, as we shall see, was subtly but substantively different in each place, yet in none, at least initially, did the national accounts envision banks as economically productive (and nor would they in France or the UK for several decades).  The importance of this withholding of productiveness, in the prevailing context of protectionist national banking and finance systems, is the main subject of concern in this section. I then conclude the chapter with a brief segueway to Chapter 4, since the latter will shift our focus to the single major economic power of the day whose national accounts did show banking as a productive economic sector, and it will argue that the identity of this country should come as no surprise: it being the country which had the greatest immediate capacity to re-internationalize its banking sector, which had financial institutions agitating already for open international financial markets in the 1940s, which was the lightest use of capital controls internationally in the late 1940s and the 1950s, and whose banks would indeed become in the 1960s the first actually to re-colonize foreign markets. This, of course, was the United States.


3.1 The Retreat to the Nation  (p. 106)


3.2 Bretton Woods and the Properties of Mobile Capital  (p. 108)


3.3 The Birth of National Accounting and the Rebirth of Economic Productiveness   (p. 113)

p. 116:

  The output or product method arrives at an aggregate GDP figure by summing output on an industry-by-industry basis. It generates, in doing so, the "production account." ( ... ... ) National accountant have two different ways of getting around this [double-counting] problem. One is to sum, for each industry, only "final" goods and services output, which is to say those acquired by consumers or exported. The other, more common approach is to sum not outputs but "gross value-added" (GVA), which essentially equates to outputs ^minus inputs^ for each industry. Total GVA, summed across all industries, is effectively─although not exactly─total GDP.[n.33]  Meanwhile, the reason why this third method for calculating GDP is the most material for our purposes should, by now, be evident: for it is ^only^ with the product method that national economic output can be readily broken down into the estimated relative productive contributions of different industrial sectors, banking among them.

  This still leaves one final, crucial question to be addressed, however. ^Which^ industries─or, more broadly, which sets of social agents─are considered to be economically productive, and thus to merit an output or GVA calculation in the national accounts? ( ... ... )

  ( ... ... ) Nevertheless it is helpful, I think, to indicate at this juncture the generic shape of the answer which has essentially informed and animated Western national accounting from its origins through to the present day. ( ... ... )

Thus, ^in general^, activities are considered productive if they are deemed to produce a "useful output" and if a value can be assigned to this output through either imputation (a concept and practice we will have occasion to consider in some detail later) or payment actually made. The "production boundary" separates those activities which fulfill both of these criteria from those which do not.

  In my discussion below of the treatment of banking services in national accounts, I will be focusing closely on this question of where the production boundary is placed, and on how different placements necessarily entail different framings of particular economic activities. I do so, and indeed insist on the importance of doing so, because as Anwar Shaikh and Ahmet Tonak observe in their critique of Western national accounting, what counts as "productive" and what does not clearly "changes the very nature of the accounts, the picture that we see, and the conclusions that we draw."[n.34]  What is especially interesting to note, however, is that while there is a longstanding tradition of questioning ^exclusion^ from the product domain, and the effects of such exclusions on the social valuation of the activities in general─the classic example being household labor─much less critical attention has been paid to what gets ^included^.[n.35]  How and why do certain activities come to be considered "productive", and what are the effects of ^these^ (mis)representations?

  Shaikh and Tonak's critique is actually an important exception in this regard. They consider unproductive a range of activities which Western national accounts, today, typically treats as productive. One such activity, notably, is financial services provision. Such services, the authors claim, "derive their revenues from the recirculation of the money flows generated by the [economy's] primary sectors" and hence merely "preserve or circulate [social] wealth, or help maintain and administer the social structure in which it is embedded."[n.36]  This argument is directly inspired by the Maxian distinction, discussed in Chapter 1, between production and circulation, and by Marx's own placement of finance in the latter sphere. If we follow Marx, Shaikh and Tonak argue, finane is unproductive, and thus should be excluded from our estimates of economic output. They go on the show that this─excluding finance─was precisely the approach of the Material Product national accounting system used in the former Soviet Union, Eastern Bloc countries and, until 1993, China. They rightly point out, furthermore, that there remains considerable confusion around this system, particularly the widely-held view that its restriction of "productiveness" to physical goods was derived from Marx. While disabusing readers of this particular misconception (about the Material Product system, and about Marx), however, they effectively reproduce another: which is that we need to look to the former Communist states for real-world examples of national accounts that treat financial services as unproductive. This, as we shall shortly see, is simply, and critically, not the case.


3.4 The Banking Problem   (p. 118)

Reviewing in the mid-1980s the various approaches taken internationally, over the previous four decades, to the treatment of financial services in national accounts, Bryan Haig reflected that all had made attempts to one extent or another “to solve the banking problem.”[n.37]  This so-called “banking problem” is the primary focus of the rest of this chapter. The present section discusses, first, how and why this problem arises and what form it takes. As we shall see, it centers on the question of how the productiveness (or otherwise) of banking is to be measured. More particularly, it concerns the placement of a specific subset of financial services─banks' intermediation services─vis-a-vis the production boundary: which side of this boundary do such services properly sit on? The “problem”, in short, is that the generic national accounting treatment of interest flows necessitates a placement of intermediation services on the “wrong” side of the fence, thus (usually) also rendering the banking sector in toto unproductive in terms of its reported contribution to national economic output. The question of why such an outcome should be considered a “problem,” and by whom, is the second issue addressed in this section.
[n.37] B. Haig, “The treatment of interest and financial intermediaries in the national accounts of Australia,” ^Review of Income and Wealth^, 32, 1986, pp. 409-424, at p. 415.
  Two pieces of scene-setting are required before we can delve into the question of why and where the banking problem materializes. First, it is necessary to highlight the way in which national accounts distinguish between "income" and "transfer" items. As we saw above, for an activity to be considered "productive" in the national accounts it must generate a "useful output" to which it is possible to assign a value. Usually, this value is based on payment rendered for the output in question. But there are certain economic situations in which payments are made but production is not considered to have taken place─namely, where there is deemed to be no "output," or where any output is not regarded as "useful"─and it is in these situations that the income/transfer distinction arises. The payments in question, which include things like unemployment benefit and state pensions, are treated as transfer items rather than income. The conceptual reasoning for this treatment is that such payments represents "a redistribution of existing incomes" rather than "any addition to current economic activity"[n.28]  They are not, in other words, remuneration for productive work.

  The second form of scene-setting concerns the various different activities carried out by banks,and the importance of categorizing these in a meaningful fashion. ( ... ... )

  The first thing to note is that we are not concerned with insurance activities; our field of interest, rather, is banking, of all types. Broadly speaking, banks engage in three types of activities and make money in different ways from each. 
  • First, they provide services for which they are explicitly paid fees: capital raising; mergers and acquisition advice; fund management; current transfer; and so on. 
  • Second, they provide so-called intermediation services, where their role is to facilitate exchange between buyers and sellers of financial assets. The most obvious example of such market-making is the taking of plain-old cash deposits and the making of cash loans─where depositors are the sellers of cash (or, more accurately, of the right to use that cash until such time as it is reclaimed), and borrowers are its buyers. Much of the difference between "high street banks" and "investment banks" is located in the fact that they simply make markets in different types of assets: the former in cash, the latter in, inter alia, debt and equity securities and derivatives thereof. The key point here, however, is that with all such cases of financial intermediation, banks generate positive net revenues only through setting a margin between the prices at which assets, or the use of such assets, can be bought and sold: either a "buy/sell margin" (different cash prices, for instance, for the purchase or disposal of company shares) or an interest margin (different interest rates, most recognizably, on cash deposits or loans). 
  • Third, and last, banks strategically buy and sell assets themselves, using their own funds. Sometimes this is called trading (particularly over short time horizons), at other times investing (and at still others, of course, speculating) but in all cases, the aim is to profit from any ongoing income generated by holding such an asset, and/or from selling the asset at a higher effective price than that at which it was purchased.
With this tripartite classification of banking activities in hand, we can proceed directly to the aforementioned banking problem in national accounting. For, of these three categories of activity, two have historically posed no substantive problems for estimating GDP. Services where an explicit, direct fee is charged are deemed to sit comfortably on the productive side of the production boundary, and assigning a value to such services in the production account is usually a straightforward affair. It is clear what service has been provided, who has provided it, and, ordinarily, how much they have been paid for it.

  All monies generated through the buying and selling of assets with proprietary bank funds, meanwhile, whether such monies take the form of interest, dividends or capital gains, have historically been excluded from this account─just as they would be if it were an individual doing the buying or selling. No service or product has been tendered, and hence no value, the argument goes, has been added. Interestingly, the late 1990s and early to mid-2000s saw something of a push, in certain quarters of the international national accounting community, to reconsider this stance. Perhaps, it was suggested, this traditional approach was wrong, and earnings from so-called "own funds" and/or from holding gains should be included as part of banks' "productive output"?[n.39]  Doing so would, of course, tend to inflate the estimated contribution of the banking sector to national GDP. Not surprisingly, the financial crisis which began in 2007 abruptly killed off such notions: the idea that national accounting might have been underestimating banks' productiveness rapidly became sacrilegious.

  If banks' fee-based services and proprietary investment activities are treated as productive and unproductive respectively, what, then, of their final category of activity─intermediation services? These, it turns out, are where full responsibility for the "banking problem" lies. Part of the problem is essentially conceptual. What useful "service", if any, is actually being provided by the banks here? Or, to turn this question around: if holding of deposits or the making of loans really were a commercial "service" of some kind, would companies and individuals not be prepared to make explicit payments to banks in kind, aside from nominal regular maintenance charges on some accounts and occasional ancillary fees such as for early loan repayment? Where, as is ordinarily the case, banking is provided "free," how can it be a productive, valued service?

  The numerous historical attempts to solve the "banking problem" have therefore been partly about tackling this conceptual conundrum. Practitioners have offered various different answers as to the essential nature of the supposed intermediation "service." "Some explanations," observed Carol Carson and Jeanette Honsa in an article published in 1990, "refer to the services of liquidity provided by the financial institution. Others stress checking and bookkeeping services or safety." But the very fact that different protagonists were seen to identify different service forms underlines the essential problem. There was, at this stage at least, no agreement on what the core, useful intermediation "service" actually is. Carson and Honsa acknowledging that the "precise nature of the services" provided by banks, and which national accountants still struggled to define, "is not clear."[n.40]

  Yet such conceptual issues have never been the primary concern for the national accountants who have grappled with the "banking problem." Their main concern, rather, has been with how the output value of banks' intermediation activities might be quantified if such activities are treated as providing a service with a useful output. Once more, it bears reiterating: only activities with useful outputs to which a value can be assigned cross the production boundary. What is this value?

  We begin to grasp the crux of the banking problem when we recognize that one cannot simply use actual service payments as the basis for such value, since, as we have seen, explicit fees paid by discrete, identifiable customers are not ordinarily levied in respect of intermediation (the "service" is provided "free," such provision being funded through a margin as opposed to a transparent charge). This, therefore, leaves only one possible means by which banks' intermediation activities can be enabled to successfully negotiate the production boundary: an imputation of the value of the "output" of such services. As we shall see Chapters 4 and 5, this solution has gained increasingly widespread traction over the course of the history of national accounting and has taken many different individual forms in different times and places.

  While the exact methodologies of imputation are not relevant at this juncture, one comment on a shared characteristic of such methodologies is extremely pertinent. This is that, directly or indirectly, they all include the net interest revenues generated through intermediation within what is deemed to be the "productive" revenue of the banking sector. And in so doing, national income statisticians make a key exception from the generic conventions of the national accounting calculus. For, as Haig explains, using the transfer/income distinction introduced above, interest is as a rule excluded from the productive domain:
It has long been accepted by national income statisticians that receipts and payments of interest are transfer items which do not result in current productive activity. Accordingly, net receipts of interest are not counted in national income and expenditure, nor in the contribution of industries or sectors to total national output.[n.41]
[n.41] B. Haig, “The treatment of banks in the social accounts,”^Economic Record^, 49, 1973, pp. 624-628, at p. 624.
In other words, net interest earned by retailers, automobile manufacturers, pharmaceutical companies and so forth is excluded from "productive" proceeds; it is a transfer item, not income, representing economic redistribution rather than additive economic output. Where an imputation is made for intermediation, banks are therefore made an exception. Granted, this exception has a putative rationale (banks' net interest earnings being uniquely derived from the provision of a service). Yet it is an exception nonetheless.[n.42]  And it is an exception, as we shall see, which not all Western national accounting communities have always considered necessary or desirable.
[n.42] As Frits Bos, “The national accounts as a tool for analysis an policy; past, present and future,”Unpublished PhD thesis, Universteit Twente, 2003, p. 194, notes starkly of the conceptual implications of the exception: “lending money by a bank is production” whilst “lending money by non-financial producers or households is no production.” Cf. R. Speagle an L. Silverman, “The banking income dilemma,” ^Review of Economics and Statistics^, 35, 1953, pp. 128-139, at  p. 129: “the lending institution is ‘productive,’ but the wealthy relation financing his nephew's college education is not.”
  In the absence of an imputation, meantime, banks' intermediation activities necessarily sit on the unproductive side of the production boundary, and the only banking activity to register a positive output entry in the national production account is the first of our three categories: services for which banks are explicitly paid fees. Which brings us, finally, to the ultimate crystallization of the “banking problem.” If banks' GVA is calculated on the basis of only limited revenues (including no proprietary investment or intermediation-based earnings) but a full set of inputs (i.e. all costs, excluding interest payments on intermediated funds), the accounts' picture of the sector's contribution to national output will inevitably be a distinctly unflattering one.

  How unflattering is obviously one important question. In the 1940s and 1950s, when national accountants were first tussling with these quandaries, the picture sans imputation was extremely unflattering, which explains in part the broad and rapid consensus that there was a problem with how banks were to be represented. In that era, banks were still, in very large part, intermediation-focused business; fee-earning activities and trading with own funds were much less significant than today. As such, if net interest earnings were simply excluded from the calculation of GVA, there was very little left on the positive side of the ledger, and more often than not a negative GVA─that is to say, larger sector inputs than outputs─would be the result. This was precisely the outcome, for instance, in the official Australian national accounts of 1946 and 1947.[n.43]  Thus, the influential UK national income statistician Richard Stone confirmed for a wider audience in as early as 1947 that no imputation for intermediation services a banking sector "deficit rather than a surplus would appear."[n.44]  Writing a decade later, Paul Studenski put the matter more starkly, noting that in the many instances where banking sector operating costs were indeed greater than the sum of explicit charges to customers, such an accounting approach would entail the representation of banking as “a drain on national income rather than a contribution.”[n.45]  Perhaps most stark of all, however, were the words of the US statistician John Gorman another decade on. “In effect,” he wrote in 1969, “our simple commercial bank would be portrayed as a leech on the income stream.”[n.46]

  The Western banking sector has, of course, changed enormously since Stone, Studenski and even Gorman articulated the concerns which Haig later captured with the "banking problem" epithet. As fee-earning businesses and the investment of proprietary funds have become increasingly material, the relative contribution to sector revenues of intermediation-derived net interest income has gradually fallen.[n.47]  Such income remains, however, highly material. Ewald Engelen and co-authors have recently shown that across six major Western markets, the average contribution of interest income to bank profits fell by more than 20% between 1984 and 2007 alone─but that even subsequent to this decline, the average contribution was still above 50%.[n.48]  Not all such income is generated through intermediation, of course, but much of it is, and it is thus clearly the case that placing intermediation services outside the production boundary would still, today, result in a serious dent in estimated sector value-added. The economic basis of national accounting's banking "problem" has not disappeared, therefore; even in the context of a sector less reliant on intermediation, what two UK national income statisticians recently referred to as the "threat" of showing a negative or marginal banking GVA remains, theoretically, live.[n.49]  The implied GVA number would not, in most countries, be quite as unflattering today as it would have been 50 or 60 years ago; but it would be considerably less flattering than a number allowing for the incorporation of contemporary net interest earnings via imputation.

  All of this leaves open, however, the question that is perhaps most important of all─namely, why it was that many early national accountants saw an envisioning of the banking sector as a "drain" or "leech" on national income as a problem, and why most of their successors have continued to do so. (In his 2005 history of national accounting, Andre Vanoli, articulating the broad consensus amongst his contemporary peers, dismissed such an envisioning as “totally unrealistic.”[n.50]) What exactly is the nature of the "threat"?

  It would not be altogether unexpected if some readers were to suspect that a "negative" representation of banking productiveness was and is considered problematic in a political sense. After all, as we saw in the introduction to this book, the financial services sector takes considerable interest in how it is represented in the political and public eye, not least in terms of the nature and scale of its contribution to the economy; and the sector has never been slow to contest directly and forcefully representations that it considers unduly or unfairly negative. It is also demonstrably the case that banks and other financial services companies, in many countries, have considerable resources at their disposal to conduct just such representational battles. It is certainly not inconceivable, therefore, that national income statisticians might have deemed the envisioning of an unproductive banking sector as a political problem─as a representation that was and is unlikely to curry favor amongst powerful industry executives (and perhaps, by extension, among their supporters in government), and which was and is unworkable accordingly.

  The research conducted for this book suggests, however, that politics has never been the source of national accountants' enduring "problem" with the banking sector and how to account for its output. ( ... ... )

  In fact the discussions threw up findings that were, to this author at least, somewhat surprising. Until perhaps the late 1980s at the earliest, bankers appear to have shown little or no specific concern with the mechanics of national accounting and of the treatment of the banking sector it offered, even in countries where the resulting picture of the sector's output contribution was indeed unsympathetic. ( ... ... )

  ( ... ... ) Prior to 1990, and especially since, the calculus of national accounting has been mobilized at the very heart of the self-aggrandizing discourse of bank "productiveness."  Yet the point to be emphasized here is that this has more often than not been─and I used the word advisedly─an ignorant mobilization. As the Swiss national income statistician Philippe Stauffer writes: "Bankers use national accounting figures, but often do not really know what they are using."[n.52]

  ( ... ... )

  By the stage in the late 1990s and the 2000s that such meetings were occurring internationally on a more regular basis, the banking "problem"─the risk of representing banks as parasites─had, as we shall see later on, effectively been resolved in most Western countries. The most notable of those countries which had represented banks as such in their national accounts, no longer continued to do so. The theoretical "threat" to banks had therefore diminished. It is therefore somewhat ironic, though I do not wish to dwell on the irony, that only now─having only recently begun to take an active interest in the methodology of accounts generation─did bankers pause to question whether the picture of their economic contribution was a sufficiently positive one. Once they did so, moreover, they frequently found themselves to be dissatisfied. Stauffer notes that before the recent financial crisis, bankers were "convinced that the national accounts measures were biased, and underestimated the contribution of bank output to GDP" (though he goes on to say that "the crisis ended this feeling").[n.56]  Yet even then, Stauffer confirms, "there was no attempt to lobby national accountants" for a "better" picture.[n.57]

 ( ... ... ) What remains clear, however, to this author at least, is that national accountants have never seen the "problem" of potentially envisioning the banking sector as a "leech" as a political problem─they have never felt politically compromised, directly or indirectly.

  I will argue here, instead, that the problem has always been principally an intellectual one─but that is no less significant or interesting for that. ( ... ... ) We need to think closely, that is to say, about the relationship between accounting for wealth and theorizing about wealth.[n.59]  For, ultimately, the reason why the banking "problem" has typically been a problem is that according to the conceptual framework inhabited by most national income practitioners, the notion of banks being economically unproductive simply does not make sense. ( ... )

  A creature of 20th century, national accounting developed as a calculative economic technology at a time when, as we have seen, the political-economic approaches of the 19th century had been thoroughly usurped by neoclassicism as the dominant intellectual toolkit for understanding economic dynamics. This was a transition of enormous relevance for the story of representational developments we are seeking to understand here. The national accounting offices which sprung up around the world from 1940s onwards were populated with economists trained at university not in political economy but in the marginalist theories of neoclassicism. ( ... ... )

  ( ... ... ) p. 126


3.5 Banking Problem? What Banking Problem?   (p. 129)

( ... ... )

  What, however, about the practical reality of national accounts? Did the practitioners who actually produced national accounts, on the ground, always and everywhere agree with this broad consensus, or were there substantive dissensions, reflected perhaps in contrary treatments─treatments that did not impute a positive value for intermediation services, and hence which did not staticize away the conceptual problem in question? Even where those with the responsibility for crafting accounting methodologies did agree that a negative representation of the banking sector would be problematic, was it always and everywhere expedient for them to find a satisfactory and workable solution to the problem? These are the broad questions addressed in this final section. Where Chapter 4, which follows, focuses on one country where practicing statisticians immediately took the bull by horns, so to speak, and implemented a working solution to the banking problem (the US), our focus here is on three countries displaying a very different statistical history. Each treated intermediation services, and thus the banking sector more generally, in a different way; but in none did the initial accounts framework allow for a directly positive envisioning of bank productiveness.

  The three countries in question are France, Germany, and the UK. While I will make reference, both here and later in this book, to the treatment of banking services in the national accounts of other countries─and of course in the methodological standards drafted by international statistical authorities such as Eurostat and the UN's Statistical Office, to which the various treatments have by no means always and everywhere conformed─the spotlight throughout will be trained largely on these three and the US. Why? ( ... ... )

  How, therefore, did the first sets of national accounts of France, Germany, and the UK envision banks in general, and their services of intermediation more specifically? We can take Germany first, since it is the simplest to deal with. Due to the occlusion of official government statistics from the mid-1930s to the end of WWII on account of their subsumption into the closely-held regime of power-knowledge that constituted the Third Reich, Germany's role in the early history of national accounting has often been underplayed. The fact of the matter, however, is that Germany was already producing such accounts from 1926: "the first regular series," as Adam Tooze observes in his brilliant book on statistics and statecraft in early-20th century Germany, "to be produced by any major capitalist state."[n.77]

  ( ... early German case... )

  How were banks envisioned in these [French] accounts? The first important observation we can make is that they─or "financial institutions" more generally─were deemed sufficiently distinctive to warrant isolating as one of the five main categories of economic agent in the accounts, alongside households, other commercial enterprises, government, and the rest of the world. The second, more critical observation is that such institutions were assumed to have zero output─to be, that is to say, wholly unproductive economic agents. This zero rating, moreover, was extended across the full gamut of financial sector activities: to own-fund investment, and to intermediation services, and even to services for which financial institutions did explicitly levy fees and receive direct monetary payment. This represented, in other words, a wholesale and categorical denial of financial sector productiveness; not so much a question of the banking "problem" being recognized but being deemed not to required a solution, rather an explicit denial of there being a banking problem at all. Envisioning banking as unproductive was not problematic; it was an accurate representation of economic reality as French national income statisticians appeared to see it.

  We will attempt below to try to explain this particular treatment. But it is worth appreciating here that it was no fleeting ephemera, a temporary aberration that was quickly perceived as such and rapidly corrected. The French national accounts continued to show banks as entirely without productive economic output until 1975. Certainly, there have been other examples of "Western" countries totally omitting income generated through banking from the national product; Australia, for example, did so through to 1937/38[n.81]  Then, of course, there are the countries of the Eastern Bloc, which, including East Germany after WWII, also rendered banks unproductive in the accounting tableau in their use of the Material Product System.[n.82]  To the best knowledge of this author, however, the national accounts of France presented such an image of banks for a longer period than any other Western nation in the post-war era.

  The UK, meanwhile, pursued yet another alternative course. The first official GDP estimates, for 1938 and 1940, were published in 1941, but these used "only" the income and expenditure methods of sizing the national economy. The question of how to deal with the banking sector, and in particular its services of intermediation, thus did not surface until the product/output method─and hence the production account─was added to the mix in 1946.[n.83]

  The UK's national income statisticians clearly agreed with Stone that bank's intermediation services, in not being explicitly paid for, presented a dilemma. They also, equally clearly, disagreed with their French counterparts, in the sense that they did not think that showing banks as unproductive actors was unproblematic. It was, or would be, a problem, conceptually if not politically. But the key feature of this particular story is that these statisticians were not ready to take the step that, as we shall see in the next chapter, their US counterparts took, which was to "solve" the problem in such a way that the troublesome representation in question was avoided. Instead, they offered a treatment whereby only the few financial services for which UK banks explicitly charged their customers were deemed ultimately to make a net positive contribution to GDP, and hence under which banks' overall proportional contribution to GDP would indeed register as negative or only very marginally positive. Before considering how the statisticians in question rationalized this choice, we need to examine the exact nature of their treatment of intermediation.

  Rather than excluding the revenue flows associated with financial intermediation services from the production account entirely, the UK's statisticians adopted a seemingly even more perverse and tortuous approach.
  • They did impute an output value to such services, namely by deducting banks' interest payments on liabilities acquired through intermediation (mainly cash deposits) from the interest income they generated on assets acquired in the same capacity (mainly cash loans)
  • However, instead of treating this net interest revenue as the input─or, in the lexicon of national accounting, "intermediate consumption"─of one or more other sectors of the economy (as in the US methodology we will look at later, for example), they treated it as the input of ... the banking sector inself! It was, in other words, output and input, and hence always netted out to zero
  • The result, of course, was that the banking sector's reported value-added was precisely the same as if those interest flows had been treated as transfer items rather than income (as in Australia in 1946-47), and simply excluded from the production account accordingly.

  Was this an ideal solution? Patently, not. But it was a better solution, the statistical office believed, than any others on the table. Treating the "output" of banks' intermediation services as a charge not on the banks themselves but on other economic agents was something that was actively considered, but ultimately resisted. Why? Because any such distribution of these "imputed charges" to third parties would be "purely hypothetical" and, as such, "would be more misleading than the paradox of financial concerns appearing to make a steady annual loss."[n.84]  In other words, while the envisioning of banks as unproductive, or as "leeches" on the national income stream, was indeed considered problematic, "paradoxical" and "misleading," it was not seen as problematic enough to warrant an accounting solution which was considered technically unsound; it was problematic, but not that problematic.

  ( ... ... )
  ( ... ... )
  ( ... ... )

  Given these two national contexts, the fact that the early national accounts of each country denied the decisive stamp of productivenss to the banking sector is of the greatest historical import. This import extended, moreover, to the philosophy, politics, and practice of banking across national boundaries, as well as within them─or so I shall argue shortly. But to get to that argument we need to consider the key question of why the French national income statisticians decided picturing an unproductive banking sector was unproblematic, and why the UK's statisticians decided it was not problematic enough to justify a compromised methodological solution.

  ( ... ... )

댓글 쓰기