Finally, over any length of time, the other things held as equal are unlikely to remain so. The marginal propensity to save in most mature economies is known to decline over time, in part as money markets mature and households finance successively more expensive homes. The response of currency markets to macroeconomic fundamentals is not entirely consistent; hopes of further gains in a dynamic stock market can create an endless demand for the local currency. When the equities market corrects, the foreign exchange market can correct much more dramatically. For this reason we can justify regressing the relationship between fiscal deficits and trade (or current account) deficits over discrete periods.

This paper addresses three phases in the US economy since 1960: the period 1960-72, when crude oil was cheap and the country was in a prolonged wartime expansion; 1973-83, a period of stagflation associated with the two actions by OPEC to drive prices upward; and the era since 1984, when the US economy has enjoyed a period of cheap oil, productivity growth, but also chronic trade deficits. During each of these periods there was a distinct, surprising, and pronounced correlation between the current account balance and the fiscal balance.

The Mundell Flemming Model moved to the new phase; instead of defending pegs which were now in abeyance, the OECD members sought to fight inflation using monetary policy. In France and the UK the agony of disinflation stretched out to ’86-87, as those economies stagnated for years. Traditionally low rates of unemployment in the (current) EU lurched to levels not seen since the War. The United States experienced its worst recession since ’38; unemployment reached 11%.

From the very beginning of the period, the US economy suffered major fiscal deficits. In ’72 the US began to suffer continual merchandise trade deficits, then overall current account deficits in ’77; after ’82 this became a permanent condition of the US economy. The dollar lost about 60% of its exchange value against the yen during the period, and performed badly against the German mark and the Canadian dollar.

The end of the economic crisis in the USA saw something very new and startling: the return of major net capital flows to the US economy. Since before the Spanish American War, the US had run a trade surplus; capital had streamed out of the US (net), not into it. The other quiet, but vital development, was that the real quantity of imports into the US was growing quickly. During the previous period, the import bill was rising because the trade-weighted value of the US dollar was sagging and the price of oil (our principle import) had increased so much. During the period ’84 to the present, the real price of oil declined from almost seven times its pre-crisis level, to about 2-2.5 times. The Japanese yen was to remain an anomaly; weighted for trade, the most prominent trend was a strengthening of the US dollar.

The currency trends show something else: while the US dollar did not perform so well against other celebrity currencies, the OECD economies as a group did very well against those of other nations (chart in appendix below). The “Spiffy Era” was actually a dud for Africa, Latin America, and much of the Arab world. The decline in oil prices, the Iran-Iraq War, the Gulf War, and botched liberalization efforts combined to ruin opportunities for expatriate laborers. This is significant because the secular decline in investment opportunities in those countries depressed demand for their currency, increasing the exchange rate risk of investment there.

So we can see that, by themselves, the fiscal trends leave most of the economic trends unexplained. And indeed, even if the results from period regressions had strong p-values and economic significance, we’d have to rationalize breaking up the periods thus. If a predictive model is any good it predicts all the time. If we have to switch models everytime there’s a historical discontinuity, the model is barely even descriptive.

Here, then, is the moment to return to a syncretical model. Syncretical approaches are more likely to cover discontinuous historical events, which trends change direction and cycles turn into waves. Most economic or social events have cyclical attributes; but these cycles are themselves subject to many other cycles influencing period and amplitude. The American economy of the ’70’s, as is well established, suffered discontinuous technology shocks in the forms of an oil crisis and a sudden intensification of foreign industrial competition. The oil shock is almost a cliché; the intensification of foreign competition in with America’s primary sector is the ignored, non-hydraulic feature of this analysis.

Observe the chart above. In mid-1973, there was a series of upward lurches in the price of oil. The initial effect of this was a jagged leap in prices that subsided later in ’74. The immediate result was a reduction in output even as prices soared. Even over the course of a quarter, there were falls in imports, declines in revenues from assets in the US (causing the CA balance to go even further into surplus), a flight of US assets abroad, and an increase in net savings. Perhaps coincidentally, the US government had begun a massive withdrawal of US troops from Southeast Asia in ’72 which was beginning to make itself felt; in ’73 year the budget was nearly balanced.

The money supply had been expanding briskly thanks to the November ’72 election; Arthur Burns, the Fed Chair, was exceptionally close to the Nixon White House. In the wake of the price shocks preceding the Yom Kippur War, there was a surge in savings reflected in the rightward movement of the IS curve. The surge in prices meant a fall in consumer demand and business investment, probably intensified by uncertainty; the Yom Kippur War, at the time, was screeching towards a superpower confrontation. But the rise in prices also meant a decline in real interest rates.

All of the classic events of a contraction were there except the fall in prices. The Watergate Scandal and the Yom Kippur War reached their conclusion while economic output was falling; they continued to fall until mid-75. The “automatic fiscal stabilizer” of the social welfare system did its job, in the sense that the fiscal deficit was enormous (for peacetime) and most of the states, especially California, suffered budget crises. Technically, inflation did subside somewhat during the recession, but when growth returned the nation suffered a seismic ratched effect of prices which I attempted to show in the chart. Each time prices actually jumped - inflicted by the price of oil - the next quarter suffered a reduction in growth. Savings would bounce up and investment would fall. Then the effects of the short-term interest rate were felt, and there was a monetary stimulus. But the surge in growth was under conditions of inelastic supply. This, at the very beginning of a recovery! The NAIRU had, quite simply, moved to a far higher level of unemployment. This is why the social welfare system was on overload. It was unsuited, especially ideologically, to address the needs of a brand-new underclass of structurally unemployed.

In fact, the fiscal balance of most countries was moving into the red. Canada’s federal government ran a deficit larger than ours, as a share of GDP; the general governments of Italy and Japan have done so continuously since before the ‘70’s and suggest the upper limits of deficit stimulus policy. In absolute terms, i.e., in real US dollars, the rest of the OECD was racing Washington into the wild red yonder. In absolute terms, the other industrial nations from whom the US might finance its burgeoning deficits were doing the same thing. This meant that the torrential capital influx predicted by the Mundell-Fleming Model was checked by competition in bond markets from countries such as New Zealand, Sweden, the UK and Australia. All of these countries ran general government deficits far larger than ours (as a share of GDP); in a perverse sort of way, the United States was actually the fiscal conservative of the international community during the ‘70’s, usually in the moments when world debt was greatest.

This state of affairs reversed in the late 1975. Again, as shown on the ISLM-ASAD diagram above, the constriction of oil drove prices upward, interest rates upward, and the fiscal balance downward. Other nations scrampled to reign in their deficits, partly because the cost of borrowing was fast becoming a scandal in the countries of Northern Europe. The domestic market for US manufactures weakened, and the capital influx from other nations seemed to finance a boom in consumption goods, especially for imported goods. After 1976 the nation began running a trade deficit in both goods and services; the invasion of cheap imports of manufactured Japanese goods was already the stuff of jokes and demagogery. The surge in income from foreign assets in the US also pushed the CA back into the red.

The Carter Administration was highly successful in its assault on the deficit, but this was much a product of the economic recovery as Carter’s “discipline, discipline, discipline.” Alas, the second oil shock of the ‘70’s reversed everything, moving us back to the left side of the ISLM-ASAD chart above. The budget deficit soared, domestic consumption plummeted, savings increased, investment fell, and the CA balance was positive again. The KA again was negative as capital fled the oil-dependent US economy; this was accompanied by steeply increasing interest rates, which by April ’80 would begin a breathtaking ascent with the appointment of Paul Volcker. The Balance of Payments behaved as one might expect, sagging in response to the federal government’s dissaving. Investment and savings, balanced off by their foreign income components, maintained a predictable balance until 1982 when the US deficit not only spun out of control, but that of the entire rest of the world was brought under control. The combination of world-beating interest rates, gigantic fiscal deficits, and imploding industrial output provided both a CA-balance stimulus (downward, deep into negative territory) and a KA-balance stimulus (high into positive territory); yet the balance of payments remained surprisingly well-balanced relative to the two flows. Relative to countries like Italy or (occasionally) Japan, the fiscal deficits of the American federal government were not terribly bad, but in absolute terms, the US accounted for nearly half of the total deficit spending by OECD nations (1985). The states, for the first time since the Depression, contributed significantly to the deficit — reaching imbalances almost as high as those projected for FY2003, and larger relative to the US economy.

Relative to most major currencies, the US dollar actually rose. We had a depressed economy, a terrifying trade deficit, and our bonds were flooding world markets, yet the US dollar was climbing to Bretton Woods-era highs. Oddly enough, most of the large industrial economies had some of these traits; Japan’s and Canada’s deficits were larger relative to GDP, Germany was suffering big deficits too, and Britain was barely getting its fiscal act together by ’82. Our industrial enterprises were developing a reputation for shoddy assembly, bad design, and worse conception. But the demand for dollars was rising; compared to other nations, our economy was doing better, and individual investors could expect a higher rate of return on investments in the USA than anywhere else…especially since, and as long as, the US dollar was rising in value.

In the late 1980’s economic recovery reversed the deficits; interest rates eased, capital and current account balances began to converge, exports surged, and the dollar fell. The most plausible reason for the loss of interest in dollar holdings, and one widely cited, was the stockmarket and real-estate bubble in Japan. This was tying up Japanese capital, which had historically flown into other capital markets such as the EU, been converted to US$, and used to buy American assets. Paradoxically, now that the Japanese were exporting less capital, they were doing more of it directly into US markets.

During the ‘80’s the US ran what were numerically the largest fiscal deficits ever seen. This record was broken by the US in 1992. It seems hard to believe, but the combination of massive global levels of unemployment with the ability of central banks to manage interest rates, the adoption of the EMU in ’92, and the effects of the Japanese financial system — all within a cordial entente of international bankers — simply eliminated any direct correlation between American fiscal balances and American trade balances. During much of this time, it is true, OECD trading partners ran even larger deficits than we did, and collectively this sometimes amounted to an absolute number far larger than those of the US; and yet, even if we could get a good time series for trade-weighted fiscal deficits for all OECD countries, we would still be vulnerable to the following caveats:

1. Savings and investment among all the OECD states would have to be known; the oscillation of these, as well as national fiscal balances, would have to part of the system of homogenous equations;
2. The bond markets are a part of the world money supply which would represent the constant in the homogenous system of linear equations describing such a relationship. The fact that they share the market with money in the form of credit issued by financial intermediaries renders this relationship even more complex
3. Finally, as we saw in the early ‘80’s, the KA account can expand “endlessly” to absorb deficits, if investors believe the national currency (in this case, ours) is headed higher. If this is the prevailing expectation, bond yields expressed in foreign currencies can go down — probably will — without reversing the demand for the bonds. As economists, we like to think this is constrained by NIPA, but evidently it isn’t.

The combination of rapid growth in technology and international trade of the last 30 years has masked the fact that the market is awash with unemployment. In the Third World, the modernization of agriculture, cash crops, and counterinsurgency actions (e.g., Colombia) has created urban labor armies. In the OECD, the oil shocks temporarily awakened the leadership to the dangers of fossil fuel dependency, but neither the EU nor the US have found a way to reduce unemployment and oil dependence at the same time; Japan has managed to conceal the problem with paternalistic business enterprises. Under conditions of severe unemployment, the NIPA relationships do not hold, and capital markets are actually grateful for safe havens.

UPDATE: In the comments section below, Conrad explains some of the limiting assumptions (but he was stymied in part because my comments section doesn't support notation) I've freely paraphrased his comments, and I hope this captures the correct sense of them:

Normally one would expect Labour supply to equal W/p_c where W is the level of nominal wages, p the price level and c the prices of domestic goods and services; the problem is that since firms are concerned with the general price level (p)—which includes the real prices of both domestic and foreign traded goods—changes in the exchange rate can have an effect on the aggregate supply curve.

One usually assumes that the open ISLM economy is a small one and is therefore a price taker in international goods markets. Such an economy therefore uses the law of one price p = Ep* where domestic price levels (p) and foreign price levels (p*) are affected only by the nominal exchange rate (E). In reality it is an extreme assumption because of factors like:

• countries specialising in different products and therefore in monopolistic competitive markets or oligopolies;
• similar goods having imperfect substitutes and the fact that there is a large non-tradable sector which doesn’t experience this kind of price pressure.

Of course in macro models, these are ignored to allow convenience of forecasting approximations.

2. Under the Basel Accords of 1988, signatory banks agreed to international standards of capitalization to allow globalization of financial services (different capital requirements across international borders allows countries to gain a temporary advantage [in financial services] with lower capital requirements…which could lead to semi-solvent banks winning the field and imploding in the next financial crisis). In the 90’s the Basel Accords were modified to ensure banks had to have the same grade, or quality of capital. Tier One is the higher-rated capital and includes owner equity in the bank.