A History of the Federal Reserve, Volume 2 (129 page)

The Federal Reserve returned the remaining unissued large-denomination notes to the Treasury for destruction. As outstanding notes returned to the Federal Reserve, they were destroyed also.
²⁵⁷

Concerned by the rising number of bank robberies, Congress passed the Bank Protection Act directing the banking agencies to agree on regulations that recognized differences in the types of banks that were most subject to holdups. Governor Robertson was the Board’s representative on a committee to coordinate regulation by the different agencies. The original proposal by the Board’s staff was highly specific and detailed. Although Maisel and Mitchell objected, the Board issued the regulations for comment. “The comments received were quite devastating. . . . [T]he Federal Reserve had made itself appear unbending, adamant, and pig headed . . . whereas the other agencies had been able to arrive at a proposal that indicated a more flexible solution was possible” (Maisel diary, January 10, 1969, 2). After further efforts, the Board’s staff accepted a proposal similar to the other agencies, and the Board voted regulation P to implement the proposal.

CONCLUSION: WHY DID INFLATION CONTINUE?

The Great Inflation that began in 1965 was not inevitable. It could, and should, have been prevented by the Federal Reserve in the interest of preserving money values and avoiding costly disinflation. Like the Federal Reserve’s other great mistake, the Great Depression, it began and continued because of mistaken beliefs about the obligations and responsibilities of a central bank. The Federal Reserve was not alone in these mistakes. They were widely shared. And, for the Great Inflation, the costs of ending inflation loomed large. Burns (1987) gives several reasons for permitting inflation to continue.

There were two kinds of cost. One was the social cost, approximately measured at the time by the unemployment rate or lost output that society would pay during the transition to price stability. Second was the political cost to the administration in power from recession or slow growth, particularly if the transition to lower inflation occurred in an election year. The second was often more relevant to the officials who made the decisions.

Set against these costs were the gains accruing to the government from
the inflation tax; inflation reduced the real value of the government’s monetary liabilities, and unanticipated inflation reduced the real value of the government’s debt. The nominal value of the government’s promises to pay remained the same, but the real value that holders of debt and money received was less than the nominal amount. The difference accrued to the government; hence it was a type of tax.

257. In the last half of the 1960s Congress passed several bills providing for civil liberties, equal protection of the laws, equal employment opportunity and non-discrimination in lending. The Board developed rules and procedures for implementing such legislation even if, as an independent agency, it was not covered by the legislation.

The government’s gains from inflation took other forms as well. Income tax rates did not adjust to inflation at the time, so taxpayers moved into higher tax brackets as their nominal incomes rose with inflation. Depreciation of fixed capital also was not adjusted to inflation. The replacement cost of capital rose with inflation, but depreciation was tied to original cost, so the replacement cost exceeded the accumulated depreciation of existing capital. Inflation also transferred wealth. Ceiling rates of interest prevented owners of time and demand deposits from receiving interest payments that compensated for inflation. Banks that issued these liabilities received the transfer. Member banks held non-interest-bearing required reserves. Homeowners who issued fixed-rate mortgages received transfers from lenders. Fischer (1981) discussed these and other costs and gains of inflation. See also Cukierman and Wachtel (1979), Cukierman (1984), and Feldstein (1982).

Theoretical economic analysis emphasized the tax on cash balances and ignored most of the other costs, perhaps because most could be prevented by institutional changes such as indexing depreciation or tax brackets to inflation or removing regulation Q ceilings. The loss from inflation was typically described as the cost of more frequent adjustment of cash balances. Frether (1981) noted that this cost was estimated at about 0.7 percent of GDP, not negligible but a small part of the cost that the public bore in practice.
²⁵⁸
A large part of the true cost was the efficiency lost by obscuring the allocative signals sent by relative price changes and the redirection of management efforts away from the search for productivity and efficiency gains.

We have no evidence that the tax on cash balances or similar gains influenced either policymakers or the public. It is never mentioned in Federal Reserve discussions. Federal Reserve and administration records suggest
that the 1946 Employment Act, and the ideas that gave rise to that act, were much more important. The imprecise language of the act, calling for maximum employment and purchasing power, became a commitment to full employment, in turn translated to mean avoiding unemployment above 4 percent. In practice, the act became an argument for an activist policy to reduce the unemployment rate. And although the 4 percent rate was not a constant, the records show that it was treated that way for many years.

258. An influential paper, Tobin (1980) compared this cost of inflation (called a Harberger triangle) to the loss from unemployment (called an Okun gap). Tobin justified opposition to anti-inflation policies by pointing out that the Okun gap was larger than the Harberger triangle. This is an invalid comparison because the loss from unemployment was a transitional cost with little or no permanent effect. (The qualification recognizes possible hysteresis arising, say, from loss of skills during unemployment.) The loss from inflation that he used considered only the steady state costs of a fully indexed and fully anticipated inflation. For a more complete criticism, see Brunner and Meltzer (1993, 187–88).

Versions of the Phillips curve used from the early to the late 1960s reinforced the belief that policy could permanently reduce the unemployment rate by accepting more inflation. The choice was seen as a social judgment; at the cost of a small additional tax on cash balances, society could employ more people, especially people with low marginal productivity and little education and skill.
²⁵⁹

Though advanced by many academic economists and policymakers in the Kennedy and Johnson administrations, there is not much evidence that President Johnson or Chairman Martin accepted this reasoning. Johnson was an old-time populist who disliked “high” interest rates and used his famous persuasive skills to prevent or delay interest rate increases. Martin was a committed anti-inflationist who made many speeches about the dangers of inflation. Yet consumer prices rose at a 6 percent average rate in the last twelve months of his term. This was the highest inflation rate since the Korean War nearly twenty years before. And contrary to the original Phillips curve, the unemployment rate had started to rise also.

There is ample reason to believe that Martin disliked this outcome and truly wanted to prevent it. The longer he and his colleagues at the Federal Reserve allowed inflation to persist, however, the more firmly held the anticipations that made it costly to end. The Federal Reserve and the administration were aware of the inflation. Why did they permit it to continue once it had started?

Martin’s beliefs, the absence of a relevant theory, errors, and institutional arrangements explain why inflation started. The first two eventually changed, but inflation continued; thus, the reasons it continued are separate from the reasons it started. Two main institutional arrangements contributed to inflation under the circumstances of the 1960s.
²⁶⁰

259. At the time, empirical estimates suggested that the Phillips curve was relatively flat; small increases in inflation generated relatively large reductions in the unemployment rate. As inflation rose, estimates of the tradeoff became steeper. The likely reason is that at the low inflation rates of the early 1960s, the public was not very concerned about inflation. David Lindsey pointed out that by 1971 or 1972, the Board’s staff estimated that the sum of lagged inflation on wage increases reached unity. Inflation could not permanently increase employment by reducing real wage growth.

260. The rest of this section draws on Meltzer (2005).

First, even keel policy caused the Federal Reserve to delay taking appropriate policy action, sometimes for months. I noted in chapter 3 that even keel became very frequent in the late 1960s. During even keel periods, usually lasting for two to four weeks, the Federal Reserve often permitted large increases in reserve growth that it did not subsequently remove. It is, of course, true that the System could have prevented the inflationary impact. It failed to do so because the cost of reducing reserves (or reserve growth) always seemed large. It could have eliminated even keel by auctioning securities, as it eventually did.

Several years later Chairman Arthur Burns accepted the importance of even keel policies for the beginning and continuation of inflation.

While the Federal Reserve would always accommodate the Treasury up to a point, the charge could be made—and was being made—that the System had accommodated the Treasury to an excessive degree. Although [Burns] was not a monetarist, he found a basic and inescapable truth in the monetarist position that inflation could not have persisted over a long period of time without a highly accommodative monetary policy. (FOMC Minutes, March9, 1974, 111–12)

Second, Martin’s acceptance of policy coordination with the administration prevented the Federal Reserve from taking timely actions and contributed to more expansive policies than were consistent with price stability. The System delayed acting in 1965 despite Martin’s early warnings about inflation; it eased policy in 1968 to coordinate with fiscal restriction, and it again delayed acting in 1970. Despite well-known arguments from the permanent income hypothesis, Arthur Okun and the Board’s staff expressed concern in 1968 about fiscal overkill from the surtax. Martin had promised President Johnson that passage of the temporary tax surcharge would lower interest rates. The Board moved to ease policy by encouraging reductions in the discount rate against the wishes of most of the reserve bank presidents. Output continued to rise and unemployment to fall. By December 1968, the annual rate of CPI increase was 4.6 percent, 1.8 percentage points higher than a year earlier. The unemployment rate was 3.4 percent, the lowest since 1951–53. Monetary base growth for the year reached 7.15 percent.

Martin acknowledged the error in easing policy. Reversing the error proved costly. As Okun eventually recognized, we could not “get back to where we were in 1965, the good old days . . . That’s exactly what we thought would happen. That’s exactly what didn’t happen” (Hargrove and Morley, 1984, 308). Expected inflation shifted the Phillips curve on which he relied.

The Nixon administration had a different analytic framework. It accepted the vertical long-run Phillips curve and paid attention to money growth as an indicator of future inflation. Initially, it chose a gradualist policy and, in its internal memos, was willing to tolerate an unemployment rate as high as 4.5 percent. By the end of the 1969–70 recession, the unemployment rate reached 6 percent with annual CPI inflation of 5.4 percent. The administration shifted its concern from reducing inflation to increasing employment.

Arthur Burns, the new chairman of the Board of Governors, decided that inflation could not be reduced at a politically acceptable unemployment rate. He told President Nixon: “Wage and price decisions are now being made on the assumption that governmental policy will move promptly to check a sluggish economy” (letter, Burns to the president, Burns papers, Box B-B90, June 22, 1971, 2). He also blamed cost-push factors, the power of labor unions, and welfare programs, along with expectations that inflation would persist. By mid-1970 he argued that inflation had become entirely cost-push, independent of previous excess demand. Soon thereafter, he claimed that the rules of economics had changed. Standard macroeconomic policies were virtually impotent against inflation. He favored controls or guideposts to break expectations. As the 1972 election approached, President Nixon accepted that advice. The administration chose political benefit over economic fundamentals.

Inflation continued because of the unwillingness of policymakers to persist in a political and socially costly policy of disinflation. During the 1960s, and afterward, there was little political support for an anti-inflation policy in Congress and none in the administration, if it required unemployment much above 4 percent. Polling data show that inflation was not named by many people as “the most important problem facing the country.” Beginning in January 1970, the number of respondents that named inflation never exceeded 14 percent; during the rest of 1970–71, it was usually well below 14 percent. Often it came fourth or fifth on the list of most important problems.
²⁶¹
Without political support, the Federal Reserve was back in a position similar to 1946–50. It had greater independence on paper; it had not committed to maintain interest rates at or below a fixed ceiling, as it had in 1942–50. The unemployment rate functioned in much the same way, however. It limited the extent to which the System would persist in a policy to end inflation or reduce it permanently. Soon after unemployment rose, the administration and the Federal Reserve shifted
their operations and goal from lowering inflation to avoiding or ending recession and restoring full employment.

261. I am greatly indebted to Karlyn Bowman of the American Ent
erprise Institute for retrieving the Gallup data.

In several papers, Athanasios Orphanides attributed continued inflation to a mistaken estimate of 4 percent as the unemployment rate consistent with non-accelerating inflation. This encouraged policies that proved excessively expansive. I am persuaded that he is correct in pointing out this error (see especially Orphanides, 2002). But inflationary policies persisted in the Carter administration after recognizing that the 4 percent rate was too low. Also, there is considerable evidence that neither the administration nor the FOMC was willing to accept a large temporary increase in unemployment to achieve a permanent reduction in inflation.

Andrew Brimmer, a Board member from 1966 to 1974, explained that employment was the principal goal. “Fighting inflation, checking inflation was the second priority” (Brimmer, 2002, 22). The FOMC never took an explicit vote to order these priorities, but the decisions taken at critical times support Brimmer’s interpretation.

Reversals had lasting effects. Inflation fell quickly in 1966–67 without a recession but with major disruption of the housing market and strident opposition from Congress and the politically powerful thrift and home building industries. The public learned from this attempt to reduce inflation that anti-inflation actions did not last once unemployment (or other costs) started to rise. The policy focus then shifted, reinforcing the public’s growing belief that inflation would continue and even increase. These beliefs made it harder for the Federal Reserve to persuade the public that it would persist the next time it tried.

The next time was 1969–70. A new administration was in power. The principal economic policymakers did not subscribe to the idea of a permanent tradeoff between unemployment and inflation. They accepted the logic of Milton Friedman’s (1968a) analysis showing that any reduction in unemployment achieved by increasing inflation was temporary. It persisted only as long as the inflation was unanticipated. But the public and Congress were unwilling to accept the temporary increase in unemployment that would substantially lower or end inflation. Officials learned subsequently that by refusing to pay the costs of transition to lower inflation, they increased the costs they would face subsequently by reinforcing beliefs that the public held.
²⁶²
They called this mixture of inflation and un
employment “stagflation” and found it puzzling and mysterious because they ignored the anticipations that the policy actions fostered.

262. I suspect that at least some of them would have paid these costs if they would not go on too long. By the time they generally recognized that their policy was working very slowly, the presidential election was less than two years away. President Nixon was not inclined to sacrifice his second term to end inflation and probably not convinced that his advisers and
the Federal Reserve could deliver. He believed that he lost the 1960 election because of rising unemployment and was determined to not repeat the experience.