Losing to win: the paradox of industrial development

Why a spoonful of failure is an important part of any winning industrial policy recipe

South Korea, 1962. The armistice to end the Korean war, a war which killed an estimated 3 million people and saw Soeul invaded and occupied four times in a year, is a mere nine years past. Adjusting for inflation, it’s about as poor as Tanzania, South Sudan or Zimbabwe is today. It’s from this starting point that the leaders of this small, impoverished country hatched a hair-brained scheme to develop an internationally competitive car manufacturing industry. They drew up extensive plans: tailoring tax incentives, plotting protective import tariffs and fashioning financing arrangements. One of the firms chosen to be the recipients of these measures was a little-known parts assembler named Shinjin that had done some rebuilding work for the US Army in the late ‘50s. 

Building good cars is notoriously hard, doubly so in a country with no prior expertise. But against the odds – and with a huge helping hand from government – this plucky little company saw its first commercial car roll off the production line in the autumn of 1963, only around a year after it began to receive serious support. As it started to grow, Shinjin – again with government assistance – entered into agreements with established global car manufacturers. Learning from these global leaders, the one-time parts supplier ramped up production in the 1970s and became the country’s leading car manufacturer.

But failure was lurking just around the corner. The company struggled to reach international competitiveness, and – without much in the way of export sales – the domestic market wasn’t big enough to keep it profitable. After taking on more debt than it could handle it started haemorrhaging money. In 1982, the government forced the closure of the one-time leading light of its industrial policy. Billions of South Korean Won had gone up in smoke; money that could have been used on hospitals, roads or schools.

What to make of this story? LSE economist Robert Wade once summed up the economic orthodoxy's anti-industrial policy views as “industrial policy is government picking winners; and everyone knows that governments can’t pick winners”.¹ Judging by this parable of government hubris, you might be tempted to agree. But there is more here than meets the eye. Yes, the government’s intervention did create a large, debt-ladened failure. But the same policies also produced Hyundai and Kia, two shining symbols of what a well-executed industrial strategy can achieve.

Today, Hyundai’s annual sales revenue is 50 times the GDP of the entire country when it was founded in 1967, five years after the Automotive Industry Promotion Law of 1962 was passed. Amid stunning industrial success, Shinjin was not the only failure; other brands such as Asia Motors and Daewoo also fell by the wayside. But these failures do not mean that industrial policy does not work, nor do they diminish the Korean strategy’s success. Rather, they were an inevitable – even valuable – part of the strategy as a whole.

South Korean leaders knew that to develop a quality car industry, government support would be vital. But they also suspected that some element of competition was necessary. So rather than picking a winner to support, they supported multiple firms and let them compete so that winners emerged at least semi-naturally. The government set the destination and let the firms compete to find the best and fastest way there. Crucially, the South Korean state was willing and able to cut off the failures and let them be cannibalised by their more successful peers.

Look too at China’s approach to electric vehicles. At some point in the mid 2000s, the Chinese government realised it was never going to catch up in the traditional motor car market. It pivoted significantly and piled support into electric vehicles. Between 2009 and 2022 the government gave out $29 billion in subsidies and tax breaks to nascent EV companies. These were backed up by aggressive measures to cultivate a domestic market for EVs, as well as huge state investment in battery technology.

These policies were widespread, providing state support to hundreds of domestic EV companies. The vast majority of these manufacturers still aren’t profitable today. Many never will be. But that doesn’t mean the strategy isn’t working. China’s BYD is tipped to soon overtake Tesla as the world’s leading EV seller, and the company reported profits of around $2.5bn for 2022. It just takes a few big successes for the failures to pale into insignificance.

In both South Korea and then China policymakers who supported the emergence of internationally competitive car manufacturers were comfortable in the presence of failure. They knew that providing support to a range of firms allowed enough competition to drive productivity growth, all the while funnelling support to where it could best be used. Without the presence of some failure, this is not a viable strategy.

But why do we need government intervention? Can’t we just leave this to the private sector and watch while Schumpeterian creative destruction works its magic? Two reasons. Firstly, industrial policy is especially risky. Work by several economists – including Mariana Mazzucato’s work on the “entrepreneurial state” – has shown that the private sector tends to shy away from the investments necessary for industrial development because of the level of risk involved and the often decades long pay-off horizon. Finance for car manufacturers in South Korea, for example, had to be directed by the state. According to one estimate, at one point in time the destination of 50-70% of domestic credit was dictated by the government.

The second reason is perhaps more important: the goals of private sector actors differ from those of wider society. Left to their own devices, existing firms make investments that perpetuate the existing industrial structure. The private sector is fantastic at incremental improvements within this structure, but only industrial policy can deliver step-change transformations. Without direction from government, new priority sectors can’t immerge. The private sector in South Korea in the 1960s had neither the vision nor the means to spur a successful car manufacturing industry.

There is of course a danger in this line of thinking. Recognising that failure is an inevitable part of even the best industrial policy risks legitimising poor policymaking that creates little or no value. Zombie firms can continue to be kept artificially alive by government handouts ad infinitum. And it can provide intellectual cover for crony capitalism to thrive.

Industrial policy doesn’t always work, and governments aren’t always wise. Sometimes the policy mix just isn’t right. The main challenge then is to make industrial policy responsive to citizens’ desire for their tax dollars to be used wisely in the pursuit of prosperity. How can we build an industrial policy framework that accepts some amount of failure as necessary, but also mitigates the risk of zombie firms and crony capitalism?

Strict decision criteria can help. Setting minimum conditions which need to be met if the policy is to continue can help governments and donors assess whether to pull the plug. A criteria that can successfully guide industrial policy is exports, which serves as an effective proxy for productivity growth. It’s also hard to fake. Governments from South Korea to Ethiopia to China to Japan have used ‘export discipline’ to great effect, withdrawing support from firms with stagnant exports. Even if a firm is losing money, they can still prove they are worth government’s continued support if exports are moving in the right direction.

The problem with this approach is that it requires a state that is willing and able to withdraw support if exports do not grow. If handouts ostensibly given in the name of industrial policy are actually necessary to sure up political support, then failure to meet export targets is unlikely to be met with government discipline. As always, politics matters. Yet even under weak political settlements industrial policy is possible – but that will be the subject of a future blog post. 

Returning to the theme of this post, failure is a necessary component of success. This may seem banal. But it is an important and necessary point to make when it comes to industrial policy. When in 1985 Nobel laureate Gary Becker wrote that “the best industrial policy is none at all”, he epitomised the view of most economists at the time. Many economists today still share this view, just look at the vehemence of the criticism the US government faced for their support of solar and EVs after some early flops.

These critiques miss the point. Failure is a necessary ingredient in the industrial development recipe. It is of course not the only one. Future posts will cover the other essential ingredients and what it means to be a good chef. But suffice it to say for now that when it comes to industrial policy, if you want to win, you must be willing to lose.

1. Robert Wade made these comments at the 2014 United Nations Conference on Trade and Development. He was summing up the views of the Washington Consensus on industrial policy. A previous version of this article erroneously named Robert Wade as an opponent of industrial policy. He is in fact a supporter. Apologies to Mr Wade!