The single market and the eurozone; what a single currency gains and what it gives up; Mundell’s optimal currency area criteria (labour mobility, wage flexibility, fiscal transfers, synchronised cycles); the gains–losses diagram and the threshold for joining; the UK’s five tests and the 2010–12 crisis. Through Bulgaria’s first year in the euro.
Bulgaria adopted the euro on 1 January 2026, nineteen years after joining the European Union and twenty-six years after fixing its lev to the euro at 1.95583. To qualify it had to meet the Maastricht convergence criteria: inflation no more than 1.5 percentage points above the average of the three lowest-inflation member states, a deficit below 3% of GDP, debt below 60%, two years of exchange-rate stability and long-term interest rates close to the best performers. In the twelve months to April 2025 its inflation was 2.7%, against a reference value of 2.8%.
By September 2026 Bulgarian inflation had reached 5.6%, the second highest in the euro area, against an area average of 3.8%. A country with its own currency might raise interest rates or let the exchange rate rise. Bulgaria now does neither: the ECB sets one rate for 21 countries, judged on the average. The question this lesson asks is when that trade is worth making.
Source: Eurostat, HICP flash estimates, September and October 2026; European Commission and ECB convergence reports, June 2025; Council of the EU decision on Bulgaria’s euro adoption, July 2025.
The European Union is 27 countries in a single market: free movement of goods, services, capital and people, a common external tariff and shared regulation. The eurozone is the 21 of them that share the euro, with monetary policy set by the European Central Bank. Denmark has an opt-out; the others are committed to join once they meet the convergence criteria.
An optimal currency area (Mundell, 1961) is a region where the gains from one currency outweigh the loss of separate monetary policies. The gains: no transaction costs, no exchange-rate risk, transparent prices, more trade and investment. The loss: when a shock hits one member and not the others, an asymmetric shock, it cannot cut its interest rate or let its currency depreciate. The OCA criteria say what else can absorb such a shock: labour mobility between regions, wage and price flexibility, fiscal transfers from a central budget, and economies integrated enough that shocks are mostly shared rather than asymmetric.
On 1 January 2026 Bulgaria became the 21st country to use the euro. By September its inflation was 5.6%, against a euro-area average of 3.8%.
The European Central Bank sets one interest rate for all 21 members. Bulgaria had pegged its currency to the euro at a fixed rate since 1999, and it qualified in 2025 with inflation of 2.7% against a limit of 2.8%.
What did Bulgaria give up by joining?
A single currency removes two adjustment tools: the interest rate and the exchange rate. Whether that costs much depends on how often a country’s economy diverges from the rest and what else can adjust instead. That is what an optimal currency area is about.
The horizontal axis measures how integrated a country is with the currency area: the share of its trade, investment and finance done with members. The vertical axis shows gains and losses as a share of GDP. GG slopes upward: the more business a country does inside the area, the more it saves in transaction costs, exchange-rate risk and price comparison, and the more extra trade and investment a single currency brings it.
Check: The gains from joining a currency area are larger for a country that:
The GG line slopes up: the more integrated a country already is, the more it saves.
LL slopes downward. The loss from joining is the cost of giving up your own interest rate and exchange rate when a shock hits you and not the others. A country barely connected to the area will often need a different policy from the one the ECB sets. A country whose trade, prices and business cycle already move with the area rarely would: its loss is small.
Check: The losses from joining fall as integration rises because:
The loss is the cost of a policy that does not fit. The closer your cycle tracks the area’s, the smaller that cost.
Where GG and LL cross is the threshold θ1 = 50. Above it joining pays; below it, it costs. A at 40 gains 2.4% of GDP but loses 3.4%: a net loss of 1%. B at 70 gains 4.2% and loses 2.2%: a net gain of 2%. The same currency can be right for one country and wrong for its neighbour. The question is not whether the euro is good but whether this economy is integrated enough.
Teacher tip: Students want a yes or no on the euro. The diagram forces the better question: integrated enough, and able to adjust without the exchange rate? Ireland and the Netherlands answer differently from Greece, and the UK in 2003 differently again.
Check: Country A does 40% of its trade with the area; its gains from joining are 2.4% of GDP and its losses 3.4%. Country B at 70% has gains of 4.2% and losses of 2.2%. Which should join?
The threshold θ is where GG crosses LL. Join above it, stay out below it.
The OCA criteria move LL. If workers leave a region in recession for one in boom, if wages fall where demand falls, and if a central budget sends money to the member in trouble, each asymmetric shock costs less. LL shifts down to LL2 and the threshold falls to θ2 = 35. A, unchanged at 40, now gains 2.4% against losses of 1.9%: joining pays. The criteria decide where the line sits.
Check: The area agrees a large central budget that transfers money to members in recession, and workers move freely to where the jobs are. On the diagram:
The OCA criteria lower the cost of losing the exchange rate; they do not raise the gains from trade.
The UK judged itself below the threshold. Its 1992 opt-out and the Treasury’s five tests in 2003 concluded that its housing market, trade pattern and cycle were not aligned enough. Bulgaria judged itself above it: with a fixed peg since 1999 it had already given up monetary independence, so LL was low for it. Its 5.6% inflation in 2026 is the cost arriving; the lower borrowing costs and trade are the gain.
Check: Between 1999 and 2008 the ECB’s interest rate, set for the area as a whole, was too low for booming Ireland and Spain and too high for slow-growing Germany. This is an example of:
One rate cannot fit a boom and a slump at once. The booming members needed higher rates and had none.
The student shifted GG up. Labour mobility does not raise the transaction-cost and trade gains from a single currency; it lowers the cost of losing the exchange rate, because workers can move instead of wages or the currency falling. LL shifts down, not GG up. Both moves lower the threshold, but the mechanism is different and examiners mark the mechanism.
Evaluate the view that the eurozone is not an optimal currency area.
An optimal currency area is a region in which the gains from a single currency, lower transaction costs, no exchange-rate risk, transparent prices and more trade and investment, outweigh the loss of an independent interest rate and exchange rate. On the gains–losses diagram the gains rise and the losses fall with a country’s integration with the area, and a country should join only above the threshold where the two cross. Mundell’s criteria say what else can absorb an asymmetric shock once the exchange rate cannot: labour mobility, flexible wages and prices, fiscal transfers from a central budget, and cycles synchronised enough that shocks are mostly shared.
On those criteria the eurozone falls short. The EU budget is about 1% of its GDP, against a US federal budget of more than 20% that automatically transfers income to a state in recession. About 4% of working-age EU citizens live in another member state, so labour moves far less than between American states, and wages fall slowly. The result showed in 1999–2008, when one interest rate set for the average was too low for Ireland and Spain and fed credit and housing booms, and in 2010–12, when Greece and Spain adjusted through unemployment above 25% because they could not devalue. Italy’s unit labour costs rose against Germany’s and, without a lira to fall, the gap appeared as lost exports and stagnant growth. In 2026 Bulgaria, newly joined, has inflation of 5.6% against an area average of 3.8% and no rate of its own to raise.
**However,** the eurozone meets the integration criterion well: most members’ trade is with each other, and the single market, the euro itself and common regulation have pushed their cycles closer, which is the argument that the criteria are endogenous and improve after joining. The area has also built substitutes for what it lacked: a banking union, the ECB’s 2012 commitment to buy the bonds of members under attack, which ended the crisis within months, and the €750bn recovery fund of 2020, the first large common borrowing. And the cost falls unevenly: for Bulgaria, pegged since 1999, or for the Netherlands, closely tied to Germany, the loss line was already low; the judgement is per country, not for the area as a whole.
On balance the eurozone is not an optimal currency area on Mundell’s original criteria, and the 2010–12 crisis was the cost of pretending otherwise. But it is more of one than in 1999, because trade and cycles have converged and because it has started to build the fiscal and banking institutions that make a one-size-fits-all rate bearable. The view is right as a description of the design and increasingly incomplete as a verdict on its future; the honest answer is that it depends on which member, and on whether the fiscal union stays an emergency measure or becomes permanent.
Euro-area figures: Eurostat, HICP flash estimates and releases for September and October 2026 (euro area 3.8% in September; Bulgaria 5.6%); European Commission and ECB convergence assessments, June 2025 (Bulgaria’s 12-month inflation 2.7% against a 2.8% reference value); European Commission, Annual report on intra-EU labour mobility 2024 (about 10 million working-age EU citizens living in another member state). The gains and losses on the diagram are invented. Diagrams drawn with Rough.js and perfect-freehand, both MIT licensed.