The desire to “make supply chains robust” has been a major talking point for protectionists (and other industrial policy supporters).  This rhetoric has accelerated since the COVID-19 pandemic allegedly showed how fragile globalized supply chains are.  A few years ago, I wrote a post questioning the validity of such claims from a market-failure perspective.  Here, I’ll question the theoretical and empirical foundations of the claim.

The argument that protectionism can make supply chains robust is specious.  Prima facie, it makes sense: when supply chains are spread out, they’ll be subject to more political, social, and economic factors in a larger area.  For example, if a firm’s supply chains go through Argentina, China, Germany, and Canada, then political and social upheavals in those areas could affect the supply chain.  If the chain was wholly domestic, then political and social problems in those countries would not necessarily affect the firm.*  

However, some consideration shows the fragility of such an argument.  It is common sense, they say, to not put all your eggs in one basket.  Rather, diversification is the way to minimize the risk of catastrophic loss.  If all your eggs are in one basket and that basket should break, then you lose all your eggs.  If your eggs are spread out over many baskets, your risk of loss is much lower if a single basket breaks.

The same holds true for supply chains.  If firms rely on a single supplier, then they are highly vulnerable to production shocks (for a technical discussion of this point, see either this paper by Acemoglu et al or this paper by me).  A single shock has a cascading effect throughout the economy, potentially affecting firms far removed from the original shock.  Indeed, the impact of the shock becomes larger when there is less diversification, akin to an avalanche, than when there is more diversification.  

In theory, protectionism would make supply chains more fragile than under free trade.  And, empirically, we see this effect play out.  A recent paper out of Japan looking at Asian firms during the COVID-19 pandemic found that firms with greater ties to the global economy had more robust supply chains and better performance than those with weaker ties.  When supply shocks started to hit, globalized firms had more partners to choose from and could subsequently offset the shocks.  Firms with fewer ties to the global market could not so easily offset the shocks and thus performed worse.  

In theory, we would expect protectionism to make supply chains more fragile.  Empirically, this is indeed what we see.  If politicians really want to protect supply chains, then getting out of the way and letting firms build their own network of partners will do more good than protectionism.  By increasing the costs to domestic firms of forming such robust supply networks in the global economy, protectionism weakens the very thing it means to strengthen.  Protectionism does not do good; only harm.

*In the case of a globalized world like ours, this last statement is not strictly speaking true.  Many items are traded globally, so anything that affects the global price will affect the firm, regardless of their connection to international trade.  But, in order to steelman the protectionist argument, we will ignore this reality.