Industrial policy? Get the basics right first

A response to Nick Timothy

Industrial policy: what and why

Nick Timothy — former adviser to Theresa May, current Telegraph columnist, and (likely) future Conservative MP for West Suffolk — has been banging the drum for ‘industrial policy’ (or ‘industrial strategy’) for many years now. In a recent article, which has garnered substantial attention, Timothy calls for ‘a serious strategy to reindustrialise, increase private and public investment, improve manufacturing output’. Timothy is hardly alone — while we are well past the peak of ‘industrial policy’ advocacy, the matter continues to divide the Right, both here and elsewhere. Where should we stand?

What is ‘industrial policy’ — or ‘industrial strategy’, the phraseology that Timothy prefers? In general, it is not entirely clear what distinction, if any, there is between ‘industrial strategy’ and ‘reindustrialisation’ — ‘industrial strategy’, despite the use of the word ‘strategy’, seems to refer to an end and not a means. For most, ‘industrial strategy’ seems to be the vague idea that manufacturing is good, and that we must, by some mostly undefined mechanism, rebalance the British economy away from financial services.

The association of ‘industrial strategy’ and the ‘industrial’, and thus with manufacturing (physical, masculine, tangible, et cetera) — is no doubt a big part of what has attracted so many on the Right (especially post-liberals) to the idea since 2016; these attractions, moreover, are usually as much (if not more) socio-cultural and political as they are economic. In particular, they are often motivated by regional policy concerns — with the rhetoric of the ‘Red Wall’, the ‘Rust Belt’, and the ‘Left Behinds’. These groups are often deemed to have suffered at the hands of an out-of-control ‘neoliberalism’ which, excessively concerned with economic efficiency, has taken the life out of these communities; in other words, the problem with a ‘neoliberal’ world without ‘industrial policy’ is not that it is bad for headline growth figures, but that it neglects what is needed for a good society. This is a strange inversion of the historical norm, where liberals accused neo-mercantilist advocates of ‘industrial policy’ of being out of touch with the public for favouring the interests of mega-corporations over small businesses and ordinary savers, and for favouring investment and exports at the expense of consumption and imports — obstructing citizens from saving, from running their own businesses as they please, and from buying the goods (domestic and foreign) that they actually want, all in order to bolster the seemingly nebulous concept of ‘national production’.

Let’s give our own definition, which hopefully should be acceptable to both proponents and opponents: ‘industrial policy’ — which we shall use interchangeably with ‘industrial strategy’, since no real distinction between the two terms ever seems to be made — is where the government puts a thumb on the scale in order to favour certain types of production over others, rather than allowing these decisions to be made for it by the free operation of the market mechanism. A national economy with a coherent ‘industrial policy’ must, therefore, have a clear vision of the country’s economic future; it must decide, top-down, what sort of industries should grow, and usually (as a natural corollary) what sort of industries should decline. This intervention could be highly intensive — as it was in, say, historical South Korea, where real interest rates for favoured borrowers in favoured industries could be as low as negative 10% — or less; and it could be highly precise in its selection — to the point of favouring individual firms rather than entire industries or sectors — or less; the point is not so much the intensity of support or selectivity, but the use of a mechanism that self-consciously departs from the ‘non-distortionary’ or ‘neutral’ preference of both laissez-faire and ‘liberal Keynesian’ fiscal and monetary policy. ‘Industrial policy’ is, in short, distinguished by being intentionally distortionary by sector, industry, or firm. According to our definition — which (correctly) makes no reference to manufacturing per se — it would, at least in theory, be just as possible for an ‘industrial policy’ to favour the services sector; which industries are to be favoured remains an entirely open question for the government concerned.

There are a number of reasons given for the view that some degree of ‘distortion’ in economic policy may be preferable, and that the market mechanism alone may not always be adequate for the maximal advancement of the national interest. The reasoning given for favouring a certain industry over others will almost always blend both economic and non-economic arguments, and arguments from the view of sectional interests and from the national interest, which makes them difficult to unpick.

Firstly, ‘industrial policy’ can be supported for reasons of national security: this is particularly true for the arms industry and agriculture, as well as industries linked to critical infrastructure (e.g., nuclear power), but this can also extend to other ‘civilian’ industries that are seen as critical for achieving relative autarky in an emergency, such as ‘basic’ industries, like steel or semiconductors. France’s secretive agency Sisse have become notorious for deploying this argument for state protection wherever they can, blocking or modifying the foreign acquisitions of such consumer brands as Evian, Danone, and Yoplait. (One would think that the French people would be glad to no longer be associated with such trash as Danone — apparently not...)

Secondly, ‘industrial policy’ can be supported for essentially economic reasons: in this view, at least from point of view of the nation, the market mechanism is not maximally efficient even at achieving high levels of economic growth. In particular, ‘industrial policy’ is argued to solve various coordination problems in developing (and sometimes even developed) economies. In a country with a shortage of capital, it might be necessary for the government to organise it and channel it towards productive uses. And perhaps more importantly, young businesses entering industries with heavy fixed costs, requiring substantial ‘learning’ to compete, will need to be supported and protected for some years before they can compete internationally; so called ‘infant industries’. Of course, the risk here is that these ‘infants’ never grow up.

Thirdly, ‘industrial policy’ can be supported for essentially ‘social’ (or indeed, political) reasons: such as to alleviate regional unemployment (as previously discussed at length in ‘The curse of being first’); or more generally because of the political muscle of certain industries — sometimes for cultural reasons particular to a certain country or region. Think of jute in Dundee, steel in Wales, or textiles in Lancashire.

---

The problem of selection

The first question we must answer, then, is the following: what sort of industry should be favoured, and why? As some commentators (most notably Mariana Mazzucato) are fond of rather glibly pointing out, there are few, if any economies that entirely forego ‘industrial policy’ under the above definition, especially if the idea of policy as ‘non-distortionary’ is very strictly applied — even something as basic as a national effort to train more engineers would (probably correctly) be interpreted as evidence of a general government view that the engineering industries should be favoured, however marginally, over the others; most forms of investment allowances will, on balance, favour those industries that rely more upon investment in plant and machinery; even a tight monetary policy will disproportionately hit those industries that tend, by nature, to be highly leveraged. But the ‘industrial policy’ of most countries already appears to go much beyond this. Unfortunately, for the most part, extant ‘industrial policy’ far better illustrates the risks of political capture and malinvestment than it does the benefits of such ‘industrial policy’ intervention.

There seems no doubt that today, agriculture — not manufacturing — is the most subsidised and protected industry on earth, both as a consequence of the political influence of the farmers’ lobby and as a consequence of more general national security concerns about an overdependence on food imports. We will ignore the arguments for and against agriculture subsidies for now, as this is hardly what most ‘industrial policy’ advocates are referring to when they call for intervention: this is simply worth briefly noting without extensive comment.

Many countries — even those not otherwise much known for ‘industrial policy’, like the United States — will favour industries connected to national defence needs: the so-called ‘military-industrial complex’. It remains fashionable to gesture vaguely towards the potential for economically valuable ‘spin-offs’ from this defence-orientated spending: for instance, Silicon Valley from American military spending in the Cold War. By this logic, the defence spending itself is simply the cloth which hides the true intention (or at least benefits) of the ‘industrial policy’ intervention. There seems little truth to such a view, which is, after all, mostly reliant on anecdote: in the excellent A Great Leap Forward, Alexander J. Field demonstrates that Total Factor Productivity (TFP) growth — essentially the ‘residual’ of growth that cannot be accounted for by increased inputs of labour and capital, and is thus used as a proxy for growth that arises from new ‘ideas’ (both technological and otherwise) — was clearly higher in the depressed interwar years than during the Second World War. As we would probably expect, pound for pound, decentralised civilian R&D spending in the direct pursuit of profit proved superior to military R&D in achieving economic growth.

Quite apart from anything else, there is no guarantee whatsoever that the country of invention will be the country that will economically gain the most, given how easily ‘ideas’ (and thus technology) can travel across borders in the modern world. A country can also be encouraged to continue with obsolete or inferior technologies that are seen as ‘native’, as occurred with Britain’s nuclear energy industry (itself a direct byproduct of our nuclear weapons programme). For the most part, then, military spending really is ‘wasted’ spending — other than in what it achieves for national defence — even if some of this ‘waste’ can eventually be clawed back (and even this is rare).

More recently, many countries have began aggressively promoting the so-called ‘green’ industries. Many Labour Party politicians seem eager to follow the United States into this trap (Green New Deal!), and, worse still, will be attempting to enter an already saturated and ultra-competitive market from a weak position — for all the talk of the ‘industries of the future’, such state-led investment is very unlikely to be fundamentally profitable for the state, especially once the cost of disincentives to fossil fuels is fully accounted for.

Another recent driver of ‘industrial policy’ spending has been a desire, especially from the United States, to economically ‘decouple’ from China. The Pandemic and the Ukraine War seemed to expose the fragility of global supply chains in an emergency or a time of conflict. But that Britain should not overly concern itself about the coming conflict in the Pacific, and that we should pursue sensible economic ties with that country, has already been argued by another contributor in the pages of this very journal. This seems to lead us to the policy conclusion that, for Britain (if not the United States), ‘decoupling’ from China should not be a priority — certainly not if it requires tens or even hundreds of billions of pounds of government spending (as it almost certainly would). Much of this would likely come out of taxes, given that the era of ultra-easy money seems to well and truly be over.

A large chunk of this spending in the United States has focused upon semiconductors: Biden has allocated over $50bn for semiconductor subsidies in the CHIPS Act alone. It seems unclear, to say the least, whether this spending will pay any real dividends: countries such as Taiwan and South Korea have entire university departments dedicated to semiconductors; in the much higher wage United States, attracting elite human capital to the industry may prove challenging. Many, upon hearing about the centrality of semiconductors to modern economic life, will want the UK to follow the US into promoting the industry. But as the excellent Rhian Chad Whitton notes, how can a country that is seemingly incapable of doing steel even think about doing semiconductors (or batteries)? Moreover, we would be attempting to enter a market where the dominant players — and especially Taiwan’s TSMC — have an enormous ‘moat’ for us to cross. You’re trying to enter perhaps the most difficult industry to enter in the entire world. Given this, it is probably a good thing that Britain’s announcement of funding for semiconductors was so paltry relative to the scale of the problem. Better not to do something at all than to do it wrong.

All this is to say that typically, ‘industrial policy’ will favour industries that have political muscle and non-economic justifications for state support. Even at best, investment will often be excessive and wasteful in ‘trendy’ and/or ‘prestige’ industries — nuclear energy, computers, and aerospace in the 1950s; semiconductors, artificial intelligence, and green energy in the 2020s. It would be best, if we are to have an ‘industrial policy’, to avoid allocating so much money to such industries, but the political incentives to do so may prove irresistible. Such are the risks of relying upon centralised (and often highly politically influenced) ‘knowledge’ rather than allowing the decentralised wisdom of the market to take centre stage.

Not all of these aforementioned ‘industrial policy’-incentivised investments will be wasted, then, but there is good reason for us to be suspicious. Rather ominously, a very wealthy American investor and ‘industrial policy’ advocate once suggested to me that perhaps the best method by which Britain could obtain the necessary political consensus for long-term ‘industrial policy’ support (which is necessary for success) would be through ‘the NHS’, which he saw as the political equivalent to the US military over here. (Imagine the nightmare of ‘NHSPharma’ trying to make its own version of semiglutide at taxpayer expense.) With ‘industrial policy’ — assuming that our goal really is a more prosperous Britain, rather than simple pork-barrelling — the risk of political capture is not mere idle talk. It is a concern that must be factored into discussions from the very beginning.

So, to return to our original question: what sectors should the British state seek to expand, and which sectors should it seek to shrink? It is unsurprising that it is rare for advocates of industrial policy — especially right-wing advocates of industrial policy — to be specific on this point. After all, in the abstract, industrial policy can seem popular, a good thing for everyone; in the specific, many will become worried as soon as they realise that the subsidies are not actually for them. No-one wants to be told that their industry not to be favoured, or worse still, that it is to be disfavoured and should decline — it is much easier for the impersonal and amoral free market to make these choices for us. It is, moreover, much more challenging for a ‘developed’ country to determine its industrial future than an undeveloped one. An undeveloped country will know the industries that previous industrialisers went through: first textiles, then steel, then automobiles, and so on and so forth. But where is the economy going in a developed country? And in a country with an adequate supply of capital, why has the market failed to make the correct assessment of the future for us?

Britain’s history of industrial policy is not a happy one when it comes to selection. Some of this has been discussed in my previous article: we supported steel, shipbuilding, and coal, all for regional policy reasons; all eventually collapsed. We promoted the wrong type of nuclear reactor. We semi-forcibly merged a good car manufacturer (Leyland Motors) with a terrible one (British Motor Corporation), and ended up sinking both. Our aerospace industry — never competitive with the American giants from the mid-1950s onwards, despite repeated mergers — joined Airbus late, and then (mostly) left before it cemented its supremacy over Boeing.

Like many other struggling industries, our computing industry would eventually be merged into a single firm: ICL. ICL was also uncompetitive, reliant almost entirely on government contracts. Thatcher wanted to let it collapse, but this was deemed impractical, as ICL was by now symbiotic with the government, with specific infrastructural knowledge of IT systems that could not be found elsewhere. Instead, the Japanese were persuaded to acquire it: ICL is now known as Fujitsu UK.

---

The problem of mechanism

So much for the problem of which industries to favour and why. But equally troubling is the lack of attention paid to the precise mechanism by which we will promote the selected firms, industries, or sectors.

Many traditional ‘industrial policy’ mechanisms are currently unavailable to Britain. This remains true even after leaving the European Union, which bound us to exceptionally rigid anti-‘state aid’ rules. Our continued membership of the World Trade Organisation, despite its much weaker ‘state aid’ rules, means that it is still difficult for us to hike tariffs to encourage domestic production. (Although it is popular in much of the scholarship to contrast ‘export-orientated industrialisation’ (EOI) and ‘import-substitution industrialisation’ (ISI), all cases of the former (e.g., East Asia) were combined with elements of the latter — they were, at heart, still ISI strategies in many ways.) Even for those who downplay the importance of tariffs per se, and emphasise the importance of export promotion, or ‘outward orientation’ as a stick to discipline recipients of state support, it seems unlikely that many of the incentives used to promote exports in previous decades would be tolerated internationally today.

There seems no question that, by modern standards, countries like Japan were very frequently engaging in ‘dumping’ — i.e., selling abroad at a lower price (or even at a loss) than in the domestic market; partly to earn foreign currency, partly to bankrupt competitors through predatory pricing, partly to maintain scale economies. In theory, ‘dumping’ has always been unacceptable in international trade, but action was often not taken against more mechanically complex export promotion schemes; nowadays, ‘dumping’ is rather more aggressively guarded against. In terms of currency, the rigidities of Bretton Woods meant that there was no real way to force a surplus country (like Japan or West Germany) to revalue upwards: a surplus country that was content to continue to pile up mountains of foreign exchange was doing nothing wrong within the system, as it was the responsibility of deficit countries (like Britain) to adjust. Outwardly, this all seemed to change after 1971, but even after the demise of Bretton Woods and the steady switch to floating rather than fixed currencies, many countries had a ‘dirty’ rather than a ‘clean’ float — i.e., they continued to intervene extensively in the foreign exchange market. This was usually to try and prevent their currency from appreciating so that exports would remain competitive, sacrificing the enjoyment that their citizens would derive from the consumption of foreign goods for the interests of outwardly-orientated big businesses. With the exception of China — who continue to undervalue their currency, being in the position to ignore American pressure — it seems unlikely that this sharp-elbowed behaviour would be tolerated to such an extent today.

In hindsight, then, many countries that operated ‘industrial policy’ from early post-war period into the 1980s benefitted from the global hegemon, the United States, turning a blind eye to East Asian practices (and pushing its European allies to do the same) — preferring to ensure geopolitical alignment, and the economic development (and thus military strengthening) of critical allies in the former ‘Third World’, even if this meant foregoing securing maximum economic advantage — rather than engaging in tit-for-tat tariff hikes and devaluations, despite long-standing stereotypes of the protectionism of Congress. This seems to have been particularly true under the administrations of Eisenhower, Kennedy, and Johnson; Nixon, as most will be aware, began to bring this system to an end by clearly reasserting the primacy of domestic American economic interests.

What about methods of intervention that do not involve the manipulation of international trade? Direct subsidies on the fiscal side are an option, but this comes with risks: this is probably the form of intervention that is most vulnerable to political capture, as it is a form of intervention that is highly transparent to all, even stupid and emotional politicians and the stupid and emotional general public. It is for good reason that much of the literature on ‘industrial policy’ will emphasise the importance of achieving a (relative) insulation of policymaking from petty politics. And where would the money even come from? It should be clear that at present, no one wants to pay more taxes, and the government is not in a position to dramatically step up borrowing.

Perhaps a better option would be on the monetary side — so-called ‘credit policy’, which, as is poorly understood by many ‘industrial policy’ advocates, was probably the primary method of ‘industrial policy’ intervention in countries as diverse as France, Japan, and South Korea. ‘Credit policy’ is a form of ‘industrial policy’ that operates by using the central bank to manipulate interest rates: in short, favoured industries would pay less, disfavoured industries would pay more — rather than such matters being determined solely by the market. The precise method used to achieve this varied from country to country, and as such is not easy to generalise.

There are probably two main benefits to intervention on the monetary over the fiscal side. The first is that the intervention becomes opaque. Savers are being ‘taxed’ by artificially low interest rates for big businesses (thus creating artificially low interest rates for savings), but headline tax rates remain the same. Political capture becomes more difficult due to the complexity of the intervention mechanisms. The second is that a form of (modified) market competition is in some sense ‘embedded’ into this form of intervention: decisions are made in a much more decentralised manner by overwhelmingly private actors pursuing their economic self-interest. This is usually less true of most interventions on the fiscal side.

There are many difficulties of operating such a scheme in the present day. Firstly, the Bank of England would have to be placed under the control — either de facto or de jure — of the Treasury (or another economic department). It would require a total change in perspective on the part of those responsible for monetary policy, who currently focus on controlling inflation. We are heavily indebted, and probably cannot afford to spook the markets in this way. But above all, such schemes probably required that there was only a very limited mobility of capital. They also generally seemed to require heavy state control of the banking system, whether nationalised or not. Otherwise, in response to such manipulations of interest on savings, capital would flee abroad in search of returns.

---

Getting the basics right first

None of the difficulties with ‘industrial policy’ presented above are, in my view, insuperable. But I hope that I have demonstrated to readers that, despite my acceptance of many of the theoretical arguments for intervention (especially nationalistic infant industry arguments), ‘industrial policy’ is risky, and that we must be clear-sighted about this. Most advocates, who are often simply seeking political advantage by shovelling pork to their favoured regions (or, worse still, are responding to psychosexual neuroses about the effeminacy of ‘email jobs’) — rather than actually wanting to further our country’s economy — are usually not.

It is for good reason that advocates of ‘industrial policy’ often seem to never advance their arguments further than repeatedly ‘owning the neoliberals’ with endless regurgitations of case studies of East Asian industrialisation in decades past. They do not dwell much on ‘industrial policy’ failures, not least those of Britain. They spend significantly less time telling us, in detail, what we in Britain should want now, and how we should get it — a much more difficult task, and one that is likely to bring you more enemies than friends. It is not so much that I reject the notion that we could, or even should, specialise in something beyond our supposedly ‘revealed’ comparative advantage in financial and professional services; or that we could, if we really tried, modify our ‘comparative advantage’ by intelligent intervention. It is just that I remain unsure about how to achieve this, and so too do most advocates of ‘industrial policy’.

The truth is that Britain’s main economic problems are much simpler — if not actually easier (especially politically) — to solve than the problem of providing centralised ‘industrial policy’ direction to the economy. Grand talk of ‘industrial policy’ is a distraction in such circumstances. We suffer from exorbitantly expensive energy, especially for industrial users — partly due to misguided green energy policies, partly due to geopolitical naivety. We have foolish rules on land use, which have led to sky-high rents and the mislocation of industry due to shortages of lab and office space. Our cities are dominated by the unproductive, often foreigners, who are warehoused at artificially low rates in some of the most potentially productive real estate in the country. Mass migration — ‘Deliveroo Britain’ — has favoured low-wage and low-productivity businesses and business opportunities while punishing, directly or indirectly, those that want to invest and create the sort of jobs that a developed country should actually want. Indeed, the resolution of some of these problems — especially our energy problems — is actually required for the success of ‘industrial policy’ in the first place.

None of these problems require expert centralised foresight about the future of our economy to resolve. Before we spend billions upon billions on highly complex and risky ‘industrial policy’ schemes — which Britain clearly lacks the will to see through — let’s get the basics first.

So, ‘industrial policy’ advocates: come back to us five years after we win. If things still aren’t working out for the British economy — then we’ll talk.