Unregulated private cryptocurrencies are increasingly used in the payments system and regulators and central banks around the world are trying to understand the benefits and risks posed by this form of private money creation (see the Vox debate on “The future of digital money” here). One prominent risk they pose is their lack of price stability, since the value of these monies ultimately depends on the stability of the (unregulated) balance sheet of the issuer. Therefore, even when they are ‘nominally pegged’ to a known government fiat currency like the dollar (as in the case for stablecoins), the value of such currencies can be very volatile, as their issuers face risks of runs and bankruptcies. Therefore, decentralised, unregulated and privately produced monies are intrinsically unstable (see also Aizenman 2019). This instability likely affects economic decisions that require long-term planning or transactions that take place over long horizons.
Yet little is known about the real economic consequences of broadly adopting such unregulated privately issued money. History, however, provides a natural experiment for estimating the effects of ‘stabilising’ the value of private monies such as stablecoins. Using the context of the National Banking Act of 1864 in the US, we show that improving the stability of privately created money promoted the traded sector of the economy, which was particularly sensitive to transaction frictions (Xu and Yang 2022).
Creation of national banks and stable money
For most of the first half of the 19th century, thousands of different types of privately printed bank notes were the primary source of money in the US. Notes were not fully backed, and so they bore the credit risk of the underlying assets (often bank loans). This risk was reflected in the way that notes often traded at a discount relative to each other, especially across town borders. Discounts were positively correlated with distance and lack of information about the issuer, and risks in these notes were well recognised by market participants (Gorton 1996, 1999). These large discounts raised transaction costs and impeded trade. Policymakers recognised the costliness of this system where many privately traded monies with uncertain values co-circulated, and it led them to pass the National Banking Act of 1864.
The National Banking Act created a network of national banks that were subject to federal regulation. One of the biggest changes was that national bank notes were required to be fully backed by treasury bonds, which eliminated their credit risk and caused them to trade at par. However, the notes were supplied by local national banks, and cities with their own national bank had access to a larger supply than places without.
In our paper, we identify cities that obtained access to this fully backed currency by exploiting the fact that national banks also faced different regulatory costs to enter in a city. These entry costs were tied to a town’s population and featured discontinuous cut-offs. We use the cut-off for towns around the 6,000 population mark. According to the regulation, national banks established in towns below the threshold had to raise $50,000 of equity capital while those above the threshold had to raise twice as much.
Using the discontinuous jump in the regulatory capital requirement as an exogenous source of variation in the likelihood of national bank entry across cities, we first show that towns directly below the threshold were 30% more likely to establish a national bank. This sizable difference in bank entry translated into differences in access to the stable, fully backed currency. This setting then allows us to study how access to such fully backed currency affected real outcomes, especially those in the traded sector of the economy.
Our findings: National bank entry shifts agriculture and employment, and persistently promotes manufacture
In the agricultural sector, places that gained a national bank significantly altered the composition of crops they produced towards those that were traded. This shift occurred while total output and capital expenditure in the farming sector however did not increase. The lack of impact on agricultural production is consistent with the regulatory environment where national banks mostly did not extend credit to the agricultural sector (Snowden 1987). These results instead likely arise from the reduction in transaction costs from gaining access to a stable form of money that increased the relative profitability in the traded sector.
We find similar evidence of growth in the traded sector when we analyse employment. We study professions that are transactions-oriented (while not being involved with production) such as commission merchants and shippers. There is a significant increase in relative employment in these professions compared to placebo professions such as doctors and teachers. We find similar differential effects between traded and non-traded goods when looking at prices. Consistent with the cost of transactions going down specifically for traded goods, we find that their prices decreased relative to non-traded goods.
Therefore, our results on agriculture and employment in tradable versus non-tradable sectors strongly suggest that as cities got access to the fully backed, stable money, the traded sector expanded.
We then focus on a sector that is by nature particularly trade-intensive – namely, manufacturing.
Figure 1 Persistent positive effect on manufacturing outcomes
Figure 1 illustrates long-term effects of national bank entry on manufacturing growth. There is a significant and persistent impact on production. Relative to the sample mean at the time, accessing the stable currency led to an increase in manufacturing production by 37%.
The census of manufacturing allows us to decompose the change in output into the different inputs. This decomposition shows that gaining access to the stable currency stimulated output by allowing manufacturing firms to buy more and better inputs, rather than investing more in physical capital. The growth coming from the change to inputs is consistent with access to stable money reducing transaction frictions that allow local manufacturers to import more and better inputs (Goldberg et al. 2010). Finally, consistent with the reduction in transaction costs improving market access and raising the incentive to innovate, we find that there is more patenting activity in places with national banks.
Our paper provides empirical evidence that stabilising the value of privately created money was economically beneficial, especially for sectors that were likely more exposed to payments frictions. As financial technology progresses and new digital currencies present alternative payment methods, these lessons from the National Banking Era can provide policymakers additional guidance on the costs of transactions frictions arising from reducing the liquidity of monetary instruments.
Adrian, A and T Mancini-Griffoli (2019), “The rise of digital currency”, VoxEU.org, 9 September.
Aizenman, J (2019), “On the built-in instability of cryptocurrencies”, VoxEU.org, 12 February.
Goldberg, P K, A K Khandelwal, N Pavcnik, and P Topalova, (2010), “Imported intermediate inputs and domestic product growth: Evidence from India”, Quarterly Journal of Economics 125(4): 1727–1767.
Gorton, G (1996), “Reputation Formation in Early Bank Debt Markets”, Journal of Political Economy 104(2): 346–397.
Gorton, G (1999), “Pricing Free Bank Notes”, Journal of Monetary Economics 44: 33–64.
Bordo, D (1998), “Currency Crises (and Banking Crises) in Historical Perspective”, Stockholm School of Economics Research Report 10.
Snowden, A (1987), “Mortgage rates and American capital market development in the late nineteenth century”, Journal of Economic History 47(3): 671–691.
Xu, C and H Yang (2022), “Real effects of stabilizing private money creation”, Working paper.