The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
I I SHADOW OPEN MARKET COMMITTEE Policy Statement and Position Papers September 21-22, 1986 PPS-86-6 CENTER FOR RESEARCH IN GOVERNMENT POLICY & BUSINESS Graduate School of Management University of Rochester SHADOW OPEN MARKET COMMITTEE Policy Statement and Position Papers September 21-22, 1986 EPS-86-6 1. 2. 3. Shadow Open Market Committee Members, September 1986 SOMC Policy Statement, September 21, 1986 Position papers prepared for the September 1986 meeting: Economic Outlook, Jerry L. Jordan, First Interstate Bancorp Circumventing the Intent Fidelity Bank of Gramm-Rudman-Hoilings, Mickey D. Levy, Time Series Analysis of "Velocity" Concepts, Robert Rasche, Michigan State University and Arizona State University SHADOW OPEN MARKET COMMITTEE The Committee met from 2:00 p.m. to 7:30 p.m. on Sunday, September 21, 1986. Members of SOMC: PROFESSOR KARL BRUNNER, Director of the Center for Research in Government Policy and Business, Graduate School of Management, University of Rochester, Rochester, New York. PROFESSOR ALLAN H. MELTZER, Graduate School of Industrial Administration, Carnegie-Mellon University, Pittsburgh, Pennsylvania. MR. H. ERICH HEINEMANN, Chief Economist, Ladenburg, Thalmann & Company, Inc., New York, New York. DR. JERRY L. JORDAN, Senior Vice President and Economist, First Interstate Bancorp, Los Angeles, California. DR. MICKEY D. LEVY, Pennsylvania. Chief Economist, Fidelity Bank, Philadelphia, PROFESSOR ROBERT H. RASCHE, Department of Economics, Michigan State University, East Lansing, Michigan; and Department of Economics, Arizona State University, Tempe, Arizona. DR. ANNA J. SCHWARTZ, National Bureau of Economic Research, New York, New York. POLICY STATEMENT Shadow Open Market Committee September 21, 1986 Pessimism about the current economy and its near-term prospects is overstated. years. The This economy has grown at a 2 1/2% rate for the last two is slightly lower than the average for the last century, but the difference is small. Concerns about near-term prospects have encouraged a return to the policies of fine-tuning that failed in the past and will fail again. The Treasury Department, Wall Street economists and various seers urge the Federal Reserve, the Bundesbank and the Bank of Japan to lower interest rates. Efforts and to force interest rates lower, to depreciate the dollar to stimulate the economy to head off protectionist legislation are based on the mistaken belief that we have learned how to stimulate and prevent inflation later. to now Efforts to get Germany, Japan and others replace stable, noninflationary policies with additional stimulus are similarly short-sighted and wrong-headed. The administration is repeating the mistakes of the Carter Admin- istration. We have been playing "locomotive," and we are now urging others to play locomotive in turn. A Long-Term Problem The U.S. has a long-term problem. We spend more than we produce and finance the difference between spending and production by borrowing from the rest of the world. tive investments, there If more of the spending were for produc- would be no problem. The returns from the productive investments would retire the debt and pay the interest as it 1 came due. We would have more capital to work with, higher productivity and a higher standard of living as a result. Living standards would rise. If, instead, tion, we will we borrow to maintain or increase current consump- live better today but worse tomorrow. To repay the increased debt, we must lower our standard of living in the future. Our problem is NOT the twin deficits in the government budget and in the trade balance. -- the fact that Our problem is the way in which we use resources we are increasing consumption at a high rate but increasing investment in productive capital at a more modest rate. If, as a nation, we continue to borrow overseas between $50- and $150- billion annually for the next few years -- and do not invest more consume less --by the end of the decade our standard of and living will fall in absolute terms and relative to other countries. During the 35 years 1951-1986, the share of GNP used for personal consumption has remained between 61.6% and 65.6%. most recent quarters, of its range. For 1985 and the two the consumption share is close to the upper end During the same period since 1951, the share of gross non-residential investment has remained between 9.0% and 12.1%. Despite other or accelerated depreciation, the investment tax credit and encouragements to investment -- measures that will be eliminated curtailed in the new tax code -- the share of stayed near the middle of its postwar range. quarters, investment GNP invested has For the most recent two the share of gross non-residential investment was 11%. is much lower, Net only 2.2% of GNP in the second quarter (see the attached chart). If we do not invest more and consume less now, we will have less from which to consume and invest in the future. 2 Each addition to our foreign debt carries an obligation to pay interest. The delay closing the gap between production and spending, est we will production have to pay and the larger the amount must exceed future spending. longer we the more inter- by The reason is which future that, event- ually, we have to close not just the deficit in the net exports but the current and account deficit. the net The latter includes the net export deficit interest payments which will be due to the rest of the world. Choices The current account deficit problem will not remain unaltered. we do nothing to increase production, lower our imports hold standard of living, increase our exports and This will reduce our until households and firms around the world become willing the rate. the dollar will fall. available stock of dollars at a relatively To If stable to exchange solve the problem by depreciating the dollar will require a reduction of about 4% in the U.S. standard of living by the end of this decade. This will amount to more than $400 per person. loss will not be uniform. lose at all. The Some will lose much more, and some will not Exchange rate intervention cannot prevent this loss. administration's current effort to depreciate the dollar one way to reduce living standards. restrict Of course, the It is not the only way. is We could imports by protectionist measures, or by forming or joining cartels, as we have done with textiles, steel and microchips. Protectionism would run the risk of retaliation, would shrink and everyone would be made worse off. where would goods at competitive prices. lose the opportunity to purchase higher so that trade Consumers every- quality foreign Producers would be forced to use higher 3 cost domestic increasing lower inputs, thereby reducing their ability to costs and reducing efficiency. compete Retaliation would by further our standard of living by reducing our exports and the benefits of trade for everyone. The administration has recently added a cartel microchips to the existing cartels in textiles. Cartels raise costs to consumers and producers. fits steel, they are will products that use microchips will be at a effect. and The beneAlthough intended to maintain output and employment in almost certainly have the opposite in food of trade are lost and standards of living are lowered. cartels The automobiles, arrangement the U.S., Producers competitive of disadvantage. administration also uses export subsidies to sell goods, particu- larly farm output, consumers and abroad. business, These subsidies will be paid for by taxing thereby lowering standards of living and raising costs at home. The growth of dollar-denominated debt held by foreigners increases the temptation to inflate away part of our obligations. Current Fed- eral Reserve and administration policy of raising money growth to lower interest rates temporarily -- and depreciate the dollar permanently -- carries high risks of inflation. Inflation the burden investment. reduce lower will raise the effective corporate tax rate. of this tax will fall on capital, expand, However, In addition, inflation will wage payments and real costs of production. unemployment currency will get of less Inflation, particularly if it is unanticipated, also will the exchange rate. real so we Much will fall and the trade depreciation, 4 deficit temporarily Output will may decline. achieved by inflation, will produce little lasting reflect the benefit. inflation, As wages and costs of production rise to the trade balance will worsen once again and unemployment will rise. Each of these measures -- currency depreciation, protection, export subsidies months. solve Each and inflationary policy -- has been our in recent of them will work to lower the standard of living. our current international imbalance while creasing, tried standard maintaining, of living, we must invest more to To or inincrease productivity. This year's tax reform has many desirable features. It should be the first of a two-part reform to increase efficiency and productivity. The second step should be substitution of a broad-based consumption tax for the corporate income tax. This step would encourage investment and cause will a temporary postponement of consumption. lead to higher productivity and a rising Increased standard investment of living. Productive investment also would help to service the nation's debt owed to domestic and foreign lenders. To we must increase our living standards in a world with low cost have efficiently low-cost capital, without and increase productivity. subsidies, to labor, produce Reductions in taxes on capital would reduce the cost of capital, and, by improving the capital stock would increase labor productivity. Elimination of the corporate income tax, and substitution of a broad-based consumption tax, would be most useful step government could take to reduce our trade without reducing living standards. 5 the imbalance Monetary Policy Current Federal Reserve policy is irresponsible. high price to reduce inflation, administration, has After paying a the Federal Reserve, urged on by the returned to the short-sighted policies that pro- duced the inflation of the 1970s. For between more than the monetary base an 8% and a 9% annual rate. annual rate. with two years, the has increased at Output has increased at a 2 1/2% Base velocity has been little changed. current growth rate of base money, average 5% to 6% over the next several years. This means that inflation is likely to As the effect of dollar depreciation spreads through the economy, prices are likely to rise at more than a 5% to 6% rate for a time. Supply-side incantations are not a substitute for rational policy. Encouraging Germany and Japan to expand demand does not solve the longterm trade problem. To base avoid should Nor does it avoid inflation. the coming inflation, be the growth rate of the reduced to a rate consistent with price monetary stability. Research prepared for this committee suggests that that rate is in the neighborhood of 3% to 4%. This goal should be achieved by the end of the decade. Policy Coordination For the past year, coordination. should act there has been increased discussion of policy The meaning usually given to this term is that countries to dampen fluctuations in spending. In the most common formulation, nations are urged to adjust monetary and fiscal actions so 6 that world trolled. demand and its distribution among countries would be conProponents argue enhanced in this manner. There is that exchange rate no known way for central banks and governments exchange stability this way. caused by that could be This is coordination by concerted action. crease Speculation stability Most exchange rate variation shifts in policy actions and about to in- other unpredictable monetary and other policies has been a main is events. reason interest rates, exchange rates and other asset prices have been extremely variable. Greater policies. price certainty be achieved by setting compatible If all countries were to adopt credible policies to achieve stability, actions could one major source of variability -- unstable monetary -- would be removed. This is coordination by common objec- tives . We urge the Federal Reserve, the Bundesbank and the Bank of Japan to adopt price stability as their common objective. the furthest from this objective, mainly on the United States. of the the burden of adjustment will fall But this burden would be lighter if all countries were to accept the same goal and strategies to achieve it. Since the U.S. is 7 announce credible INVESTMENT IS DOHN AND CONSUMPTION IS UP P E R C E N T 12.5/ Net Investnent Left Scale, Line 10. 0Z 66,25'/ 65.00V; 7.5'/ 63.75'/ 0 F G N P 5.0-/. 2.5'/ [62.50*/ Consumption Right Scale, Dot 61.25'/ 1952 1956 1960 1964 1968 1972 1976 1980 1984 Notes: Net investnent equals business fixed investnent plus residential investnent plus inventory investnent. less econonic depreciation plus net foreign investnent. Consumption equals personal consumption expenditures. Sources: Wharton Econonetries; Heinenann Econonic Research P E R C E N I 0 F G N P ECONOMIC OUTLOOK Jerry L. JORDAN First Interstate Bancorp Bancorp August 5, 1986 QUARTERLY HI 1986 ECONOMIC UPDATE Inflation Policy Assumptions • Inflation should move higher in 1987 because of ending of oil price decline, impact of dollar's drop, and delayed affects of rapid money growth. • Monetary policy will continue expansive, with Ml growth of approximately 12% in both 1985 and 1986 and 8% in 1987. • Federal deficit will remain large: $212 bit. in FY 85, $216 bit in FY 86, and $175 bil. in FY 87. • Consumer prices up 3.5% in 85 (fourth qtr. to fourth qtr.), 1.5% in 86 and 4.8% in 87. • Impact of Gramm-Rudman still in question. Interest Rates • Tax reform implies lower real interest rates than would otherwise exist because of restraining impact on investment. Much of reduction may have already occurred. • Tax reform represents major structural change for U.S. economy. Nonresidential building, capital spending, housing, and consumer borrowing and expenditures on durables will be restrained by tax changes. Lower tax rates, however, will help many firms, especially in services and retailing, and raise disposable income and spending of many consumers. • Interest rates arc expected to remain relatively flat during third quarter. Fed uncertainty about economy and low inflation arc likely to offset concern about foreign demand for U.S. securities and deficit. Economic Growth • There is no shortage of demand. Consumer spending has grown at an annual rate of 4% during the past 6 qtrs., 6% last quarter. • Forces depressing growth should start to be mitigated: (1) Auto inventories - now under control (2) Net exports-gradual improvement (3) Oil priccs--$i M5/b in 86, $12-16/b in 87 (4) Tax reform -. removal of uncertainty. Initial negative impact on investment, followed by higher consumer spending when tax rates are reduced. • Real GNP growth, equal to 2.4% in first half of 86, should average 3.5% in second half, 3.9% in 1987. • Interest rates should then move gradually higher in fourth qtr. and in 87 because of: (1) (2) (3) (4) (5) Faster economic growth Continued concern over deficit Weakness of the dollar Firming of oil prices Higher inflation. • Short-term interest rates, flat in third qtr. of 86, should rise about 25 basis points in fourth qtr. and 100 basis points in 87. Similar pattern for longterm rates, depending on instrument Some steepening of yield curve. For further information contact Lynn Reascr at (213) 614-3486 FCC 1205 6/86 ft 1 MONETARY BASE AND M1 NET OF OTHER CHECKABLES QUARTERLY PERCENT CHANGE OVER YEAR AGO 10.0 j MONETARY BASE 8.0 -h 6.0 4.0 4 M1 NET 2.0 0.0 -2.0 4 -6.0 4 .0 -\ 1 1980:1 1 1 1 1 1981:1 1 1—I 1 1982:1 h H 1 1983:1 FIRST INTERSTATE ECONOMICS 1 1 1 1 1984:1 1 1 1 1 1985:1 1 SEPT. 4, 1986 1 1 1 1986:1 2 MONETARY BASE AND M1 QUARTERLY PERCENT CHANGE OVER YEAR AGO 14.0 T 2.0 H 1 1980:1 1 1 1 1 1981:1 1 1 1 1—I 1982:1 1 1 1 1983:1 FIRST INTERSTATE ECONOMICS 1 1 1 1 1984:1 K—I 1 1 1985:1 1 SEPT. 4, 1986 1 1 1 1986:1 3 M1 AND M1 NET OF OTHER CHECKABLES QUARTERLY PERCENT CHANGE OVER YEAR AGO 1980:1 1981:1 1982:1 1983:1 FIRST INTERSTATE ECONOMICS 1984:1 1985:1 SEPT. 4, 1986 1986:1 4 ST. LOUIS AND BOARD ADJUSTED MONETARY BASE QUARTERLY PERCENT CHANGE OVER YEAR AGO u.u - ADJUSTED MONETARY BASE 9.0 - r*"V. 8.0 - K^/\_^ 7.0 ST. LOUIS FED MONETARY BASE 6.0 5.0 4.0 3.0 - 2.0 -1— i — i — i — i — i — i — i — i — i — i — i — i — i i — i — i — i — i — H 1980:1 1983:1 1984:1 1981:1 1982:1 FIRST INTERSTATE ECONOMICS 1 1 1 1 1985:1 SEPT. 11, 1986 1 1 1 1986:1 5 GROSS DOMESTIC PURCHASES & LAGGED MONETARY BASE QUARTERLY PERCENT CHANGE OVER YEAR AGO 16.0 -r GROSS DOMESTIC PURCHASES ADJUSTED MONETARY BASE LAGGED 2 QTRS H 1980:1 1—I 1—I—I 1 i 1981:1 1982:1 \ 1 1983:1 1 FIRST INTERSTATE ECONOMICS 1 1 h 1984:1 1985:1 .,—I—| 1986:1 SEPT. 11, 1986 3-MONTH TREASURY BILL & INFLATION RATE* 20.00 T -5.00 1111111 f 1111 f f f 111 i 111111111111111111111111J111111)I»)1111111111111111111111111111111111 1979:1 1980:1 1981:1 • 1982:1 1983:1 1984:1 1985:1 1986:1 * Percent change in CPI from 3 months ago, annual rate FIRST INTERSTATE ECONOMICS SEPT. 4, 1986 2 3-MONTH TREASURY BILLS - INFLATION RATE* REAL INTEREST RATES -8.00 1111 f 11111111111! 111111111111111111111111) I 1982:1 1983:1 1979:1 1981:1 1980:1 1111111)111)]1111111111111111111111 1984:1 1985:1 1986:1 * Percent change in CPI from 3 months ago, annual rate FIRST INTERSTATE ECONOMICS SEPT. 5,1986 1 YEAR TREASURY BILLS & INFLATION RATP oo 1111111 i 11111111111111111111111111111111111111111111111111111111111111111111111111111111 1979:1 1980:1 1981:1 1986:1 1982:1 1983:1 1985:1 1984:1 Percent change in CPI over year ago FIRST INTERSTATE ECONOMICS SEPT. 4, 1986 ONE YEAR TREASURY BILLS - INFLATION RATE* REAL INTEREST RATES 111111M11111111111 li 11111111111111111111111111111111111111111111111111111111111111111111 1982:1 1983:1 1979:1 1980:1 1981:1 1984:1 1985:1 1986:1 * Percent change in CPI over year ago FIRST INTERSTATE ECONOMICS SEPT. 5, 1986 5 5 YEAR TREASURY BOND & INFLATION RATE* 5 YEAR BOND 3.00 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 1 I I I I H I I I I I H I I I I I I I I H I I I I I 1 I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I 1979:1 1980:1 1981:1 • 1982:1 1983:1 1984:1 1985:1 1986:1 Percent change in CPI over five years ago FIRST INTERSTATE ECONOMICS SEPT. 5, 1986 6 FIVE YEAR TREASURY BOND - INFLATION RATE* REAL INTEREST RATES * Percent change in CPI over five years ago FIRST INTERSTATE ECONOMICS SEPT. 5, 1986 10 YEAR TREASURY BOND & EXPECTED INFLATION 16.00 -r 10 YEAR BOND 14.00 4 8.00 4 6.00 4 4.00 11111111111111111111111111111111111111111111111111111II11111111 f 111111111111111111111111 1986:1 1980:1 1981:1 • 1982:1 1985:1 1983:1 1984:1 1979:1 * Drexel Burnham Poll FIRST INTERSTATE ECONOMICS SEPT. 4, 1986 8 TEN YEAR TREASURY BOND - EXPECTED INFLATION RATE* REAL INTEREST RATES 8.0 -r 0.0 H-H-H+H+H 1979:1 1980:1 1981:1 1982:1 1983:1 1984:1 1985:1 Drexel Burnham Poll FIRST INTERSTATE ECONOMICS SEPT. 11, 1986 1986:1 9 SPREAD BETWEEN LONG AND SHORT-TERM REAL INTEREST RATES 1979:1 1980:1 1981:1 • 1982:1 1983:1 1984:1 1985:1 * 10 YEAR REAL INTEREST RATE - 3 MONTH REAL INTEREST RATE FIRST INTERSTATE ECONOMICS SEPT. 11, 1986 1986:1 QUARTERLY ECONOMIC UPDATE 1986-1987 IV) First IntersMte Economics August 18, 1986 FEDERAL DEFICIT (Billions, fiscal years) -50 •+ CTN -100 -+ -150 •+ -200 •+ -250 TRADE-WEIGHTED VALUE OF THE DOLLAR (Index, March 1973=100) 160 -r 150 -f 120 100 1982 1983 1984 1985 1986c 1987f DOMESTIC DEMAND vs. PRODUCTION (Cumulative Change From 4th Qtr. 1983 83:4 84:1 2 3 4 85:1 2 lndex=1.00) 3 4 86:1 2 REAL GNP (Percent change over prior quarter, annua! rate) 1982 1983 1984 1985 1986e 1987f TOTAL MOTOR VEHICLE SALES (Millions of units) 16 nr 14 •+ 12 o 10 •+ 8 •+ 2 •+ 1982 1983 1984 1985 1986e 1987f HOUSING STARTS (Millions of units) 2.0 - r 1.5 -+ oo 1.0 •+ 0.5 0.0 1982 1983 1984 1985 1986e 1987f INFLATION (Percent change in year-end CPI over prior year) 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986e 1987f 3-MONTH TREASURY BILLS & INFLATION (Percent) 14 T \ 1982 1983 * Percent change In CPI from prior qtr., annual rate 1984 1985 1986e 1 1 1 1 1 1987f 30-YR. GOVT. BOND AND MORTGAGE RATES (Percent) 18 - r e-l 1 1 1982 1 1 1 1 1983 1 1 1 1 1984 1 ; 1 1 1985 1 1 1 1 1 1986e 1 1 1 1 1987f 1 NONFARM EMPLOYMENT: FIRST INTERSTATE TERRITORY AND THE U.S. (Percent change from prior year) 1982 1983 1984 1985 1986e 1987f PERSONAL INCOME: FIRST INTERSTATE TERRITORY AND THE U.S. (Percent change from prior year) . 11 T 4H 1982 1 1983 1 1984 1 1985 | 1 1986e 1987f CIRCUMVENTING THE INTENT OF GRAMM-RUDMAN-HOLLINGS Mickey D. LEVY Fidelity Bank The and positive trends in the federal budget outlook forecast by OMB the CBO in February 1986 were exaggerated. On the plus side, deficits should decline from current levels, and the disturbing, sharp rise in the federal debt-to-GNP ratio is projected to recede, marking a clear improvement. looned to On the negative side, a record high --it may reach $230 billion -- and the sub- stantially lower deficit targets legislated by the Control Act of 1985 (Gramm-Rudman-Hollings, achieved. report Deficit estimates Emergency Deficit or GRH) are unlikely to be prepared for the Initial Sequestration are biased downward by accounting requirements mandated by Balanced Budget Act. across-the-board $144 billion in sequestration order for FY1987, but actual GRH target. FY1988 budgetary the Certain tactics will probably be used to avoid an will be at least $180 billion, target the FY1986 deficit has bal- and perhaps $200 billion, far above the This will make the $108 nearly impossible to attain. maneuvering, it deficits billion In light is unclear whether the of deficit recent Administration or Congress take the deficit targets of the Balanced Budget Act seriously. Nor is it clear that GRH, with its overly ambitious scheduled deficit cuts and its porous sequestration process, provides a viable guideline that contributes to sound fiscal policy. The record-breaking deficit in FY1986 reflects weak economic growth that suppressed tax revenues, and higher-than-expected expenditures for due largely to rising agricultural subsidies, deposit insurance, and the rapid pace of defense 37 a jump in outlays spending. The deficit remains well above 5 percent of GNP, GNP ratio has risen substantially, and the federal debt-to- to nearly 42 percent from 38.3 percent in FY1985, reflecting the rising debt and sharp slowdown in GNP growth. Achieving cutting over GRH's FY1987 $144 billion deficit target $80 billion from deficits -- slicing the requires deficit to approximately 3.2 percent of GNP. The sharply lower deficit targets legislated by GRH remain even though the original process was declared unconstitutional. the fallback General mechanism now in place, the role of the law, Under Comptroller is replaced by the Temporary Joint Committee on Deficit Reduc- tion, composed of the entire memberships of the Senate and House Budget Committees. submitted In accordance with GRH, on August 15, the CBO and their budget base levels for FY1987 to the Joint 0MB Committee. TABLE 1 Budget Base Levels for FY1987 (in Billions) Budget Aggregates 0MB Estimates CBO Estimates Average Revenues 826.4 827.8 827.1 Outlays 982.6 998.5 990.5 156.2 170.5 Deficit 163.4 Memo: GRH Deficit Target 144.0 0MB forecast the FY1987 deficit to be $156.2 billion and the CBO forecast above $170.6 billion; the $163.4 billion average was the $144 billion GRH target (see Table 1). Since this estimate is more than $10 billion above the GRH target, 38 $19.4 billion average the report for FY1987 includes calculations for the amounts to be sequestered. removing numerous applying special rules that limit cuts to other programs, sequestration programs spending report programs exempt from calculates that qualified sequestration To avoid this sequestration process, October non-defense 1 to enact deficit cutting legislation. and the initial spending be cut 7.6 percent and defense spending programs be percent. After cut 5.6 the Congress has until A revised and final sequestration order, reflecting any change in laws or regulations since August 15, will be submitted on October 6 and effective October 15. The OMB-CBO average deficit forecast is significantly below FY1986 levels and allowed, made only modestly suggesting above the $154 billion maximum deficit that substantial progress on the deficit has been and that the 1987 GRH target is easily within reach. Moreover, sharply lower deficit forecasts for FY1988 through FY1991 by the Administration (Mid-Session Economic and Budget Review Outlook: of the FY1987 Budget) An Update) and the (The CBO imply that future GRH deficit targets are also possible to achieve. However, a closer look at the budget base estimates in the Initial Sequestration Report reveal a different story: for the deficit estimates FY1987 based on the accounting principles imposed by GRH are tematically avoid the biased downward; sys- several methods likely will be used GRH across-the-board sequestering for deficits will be much higher than estimated. FY1987; and to actual This deceptive maneuver- ing by economic policymakers adds to public confusion about the deficit and reduces credibility in the budget process. The GRH deficit targets in later years likely will be impossible to achieve and, Balanced Budget Act may lead to poor fiscal policy. 39 in fact, the GRH Budget Base Levels for 1987 The average OMB-CBO deficit estimate for FY1987 under Gramm-Rudman accounting rules is $163.4 billion, but a more realistic deficit estimate would be at least $180 billion and as high as $200 billion. OMB-CBO estimates August 15. The reflect laws and regulations that were budget base estimates for the Balanced in The effect Budget Act (called the "Gradison base", named after Congressman Gradison, a member of the House Budget Committee) include the deficit-cutting provisions of the Budget Reconciliation Act of 1985 and the initial GRH sequestering in March 1986. levels, without However, they are based on FY1986 adjustment for inflation, because no have yet been enacted for FY1987. projections for appropriation appropriations This significantly reduces spending programs that require appropriations, downward bias to the deficit estimates. and imposes a Exact measurement of this bias is difficult to calculate, but it is instructive that the CBO baseline projections, also published in August, be $184 billion, estimate the FY1987 deficit to or $13.4 billion higher than the budget base deficit it estimated, based on the Balanced Budget act accounting requirements. OMB line seems particularly guilty of artificially lowering the base- deficit estimates in the Initial Sequestration Report. mates $15.9 billion lower spending in its budget base than and the lower estimates in several spending programs are OMB's budget base does not tionable. a 1987 though President's Mid-Session the highly CBO, ques- include additional money to fund pay increase for civil servants or the It esti- military personnel, even Review proposes a 4 percent raise for military personnel effective October 1, 1986, and a 2 percent raise for civilian employees effective January 1, exclusion 1987. OMB justifies this on the grounds that the FY1987 appropriations bill 40 has not yet passed. This omission biases down OMB's spending and deficit base by $2.9 billion, compared to the CBO budget base. OMB also estimates defense outlays in FY1987 to be $4.7 billion lower than the CBO because it assumes spending major slower spending rates. rate source is not new; is higher not, defense OMB's slower assumed spending rate was a Congressional Since CBO's spending rate is based on the histori- relationship between defense budget authority and OMB's the of contention in the recent debate on the Budget Resolution. cal This argument about outlays, while and because defense outlays in FY1986 have than anticipated, OMB's estimate seems suspect. been much The CBO also estimates $5.1 billion higher outlays for the Commodity Credit Corporation (CCC) farm price supports, largely reflecting advanced deficiency payments for the 1987 crop year. approach is the more conservative, light CBOfs assumed Again, the CBO and probably the more accurate in of recent patterns of agricultural outlays and continuing weak- ness in agricultural prices. Other deficits sources of higher-than-anticipated higher than forecast. For example, spending may drive a surge in failures of depository institutions could lead to higher outlays for deposit insurance, as in FY1986. In general, the legislative slippage in the fight to reduce budget deficits, reflected mainly by supplemental appropriations, is perpetually underestimated. Besides projections optimistic. real GNP through these biases in the base calculations, are based on economic growth assumptions that seem OMB gains assumes real GNP to grow 3.7 percent in of at least 4 percent in each quarter 1987:111) the lower deficit and the CBO assumes slightly 41 slower overly 1987 from 3.2 (with 1986:111 percent growth rate. Both assume approximately a 3 percent rise in deflator and modest interest rate increases. that Importantly, GRH requires the final sequestration report (due October 6) be based on August in the GNP these 15 economic assumptions, even though the downward GNP revision second quarter and the clear weakness this quarter nearly assures real GNP to be lower than the level used to calculate the budget base. This implies a shortfall in tax revenues and a higher deficit than the budget base. too optimistic, in Moreover, the growth assumptions through 1987 may be and continued economic weakness would generate a fur- ther shortfall in revenues. By eliminating this downward bias in OMBfs budget base levels and assuming somewhat slower economic growth, the FY1987 deficit will be at least $180 billion -- far above the maximum deficit amount that trig- gers sequestration. Suspension or Avoidance of Sequestration The Balanced sequestration Budget there If, that the quarters or if either OMB or the CBO forecast a decline in Real GNP rose at a 0.6 percent annual rate in 1986:11, but is only a slim chance it will be below 1 percent in suspends process if real GNP growth for two consecutive is below 1 percent, real GNP. Act includes a provision fact, growth is sufficiently slow to halt the this quarter. GRH process, deficits would rise and move even further away from the GRH targets. With the 0MB-CB0 average estimate deficit of $163.4 billion $9.4 billion above the maximum allowable deficit, several avenues only are available to reduce the deficit base estimates by October 1 and avoid a sequestration order. October 1, If the pending tax reform bill is passed prior to then an estimated $12 billion revenue gain in FY1987 may be 42 used to reduce the budget base deficit below the magical $154 billion level. Unfortunately, this would give only temporary deficit relief. Since the tax reform bill is designed to be revenue neutral, presumably there is a combined $12 billion revenue loss in FY1988-1990, adding to deficits in those years. Furthermore, these are static revenue impacts that assume shift. no change in economic behavior in response to the policy If the tax bill has a short-run adverse economic impact, the initial year revenue impact may be negative, not positive. More generally, lowering the deficit by increasing taxes is not solution to the budget dilemma. Nevertheless, a desirable this estimated revenue gain in, FY1987 may be used to avoid sequestration for FY1987 if the tax bill passes. A second method of temporarily escaping GRH sequestration would be the sale of certain government financial assets, such as government housing loans. The proceeds of these asset sales, which are counted as negative spending, would legally reduce the deficit estimates. The government's loan holdings are sufficiently large that such sales could reduce the selling estimated deficit in FY1987 below $154 billion. While the government's financial assets and dropping its participa- tion in certain credit markets has merits on other grounds, doing so to avoid sequestration in FY1987 is deceptive. The transfer of assets would provide a one-time reduction in spending which would be offset by higher future deficits due to the loss in interest income from no longer owning the assets. Some combination of these methods will likely be used to avoid the sequestration meeting order for FY1987. However, these methods will GRH deficit targets more difficult -- if not impossible FY1988 and beyond. 43 make -- in Another Congress possible action should not be ruled out. Suppose merely declines to vote on the joint resolution based on Final Sequestration Report, or votes to delay consideration? this that scenario will be unfolding one month before the the Remember, national elec- tions, and nobody in Congress would benefit from across-the-board cuts. Without any precedent, the outcome is uncertain. The Outlook for the Balanced Budget Act Forcing Congress and the Administration to become constantly aware of the enormous deficit problem and to consider lower deficit are two positive several It aspects of the Balanced Budget Act. large weaknesses. calls strictly because GRH has The required cuts are dramatic and for a balanced budget in FY1991. ad hoc requirement whose an But targets Balancing the appropriateness rigid. budget is is uncertain, it is not based on or supported by theoretical considerations. Reducing the deficit to $144 billion in FY1987 requires cutting the standardized-employment deficit from 4.3 percent of GNP to 2.5 percent. Excluding deficit net of billion. are are approximately $90 billion in FY1986 to a surplus of uncertain. They depend on the exact composition $1.3 policy of the the extent to which spending reductions in some federal programs offset by private sector or state and local provision of the pro- gram, this this involves shifting from a primary The economic impacts of such a large shift in fiscal highly cuts, interest costs, and the Federal Reserve!s monetary policy response. However, shift in deficit policy will occur simultaneously with major reform, tax whose short-run impact on the economy is likely negative. The weak economic expansion may be threatened by this turmoil in budget and tax policy. 44 The task of achieving the ambitious GRH deficit targets in later years intensifies. Excluding net interest costs, the deficit target of $108 billion in FY1988 requires running a primary surplus of $36 billion. If in fact, the actual deficit in FY1987 is closer to $180 billion, the FY1988 GRH target will be virtually unachievable. Numerous exemptions from the across-the-board sequestration pro- cess and special provisions that limit cuts to other programs introduce additional defense serious spending security, flaws. is In FY1987, over two-thirds of excluded from net interest, sequestration, the earned income credit, all non- including social certain low-income programs such as AFDC, child nutrition, medicaid, food stamps, SSI, and WIC, veterans compensation and pensions, state unemployment benefits, and outlays from prior year appropriations. Medicare cuts. and Over Other programs, guaranteed student loans, are subject to such as only limited one-third of defense spending is also exempt from seques- tration due to prior contractual obligations. Clearly, these rules grossly violate GRH's original intent that the burden of deficit cut- ting be distributed evenly. GRH's exemptions and uneven application may actually inhibit efforts to substantially lower the deficit because they protect spend- ing social security and transfer payments, a primary in rising spending and deficits. spending GRH's programs exemptions As a practical matter, should be considered candidates severely for source of all government budget hinders the simple arithmetic of cuts. deficit reduction. So far, cutting the Balanced Budget Act has elicited short-term actions that may not be consistent 45 with deficit- long-run program reform. These "quick fixes" are not necessarily good public policy if they not generate long-run savings or if they fail to do address, or preclude addressing, some of the structural flaws of spending programs. These flaws probably doom the Balanced Budget Act. Yet rather than consider changes that would improve the effectiveness of GRH, the Administration and Congress are effectively taking steps that undermine the Act. These actions only contribute to poor public policy by breed- ing public misunderstanding and cynicism. eventually will take GRH seriously, or make improvements to they are off to a rather inauspicious start. 46 Perhaps budget policymakers it, but TIME SERIES ANALYSIS OF "VELOCITY" CONCEPTS Robert RASCHE Michigan State University and Arizona State University A significant related innovation of the last decade in the literature to the demand for money function has been the investigation of the time series properties of the ratio of a measure of the money stock to a measure of income (or the inverse of this concept. This (1974), literature, Nelson concerned including and Plosser (1982), studies ratio) -- "velocity" by Gould and Nelson Haraf (1986) and Poole (1986) is with the question of whether the "velocity" measure is more appropriately characterized as a difference stationary (DS) or a time stationary (TS) univariate time series process. This question involves whether an autoregressive component a unit root. addressing root in an ARMA model of "velocity" has A corresponding statistical literature has the question of the appropriate test statistic for the unit hypothesis in simple univariate ARMA models (Dickey 1979, developed 1982; Evans and Savin, 1981, 1984). on velocity has and Fuller, The time series literature applied some of these univariate tests to century of annual GNP velocity measures for the U.S. and to almost a quarterly GNP and Final Sales velocity measures for the 60s and 70s. The general conclusion is that the data do not reject the simplest DS process --a random walk with drift. Several research rather and criticisms been levied against the derived conclusions. widespread "velocity" have model feeling First, that the issue this time series there seems to be of whether a univariate is "TS" or "DS" is irrelevant -- the whole issue 47 a is clearly "BS" since such tests are plagued by specification error. argument for specification error starts with the proposition "velocity" is a meaningful concept, "well known" that it must be founded in aggregate demand for money function (Wallich, 1984). The a stable However, it is that the income elasticity of aggregate money demand significantly less than unity, if is and the interest elasticity is signifi- cantly less than zero, both in the short run and in Hence, univariate tests on "velocity" impose inappropriate constraints on the parameters of the money demand function, addressed in an unconstrained multivariate the long run. and the issue must framework. A be second criticism of this approach concerns the interpretation of the events of the past several years. process for 1959-79 is argues, without univariate Haraf (1986) concludes that a univariate several quarterly measures of "velocity" over the rejected in favor of a univariate DS explicit tests, DS process. "shift" is the source of the "velocity slowdown" of problem with rejection this this He the then the unexplained 1980s. arbitrary selection of a shift point is The that the of the TS univariate process in favor of the DS process was developed under the restriction that no parameter instability in period that a "shift in the drift" of process occurred in about 1980, and TS either model during the sample period. occurred It is not evident that the rejection of the TS process in favor of a DS process is robust with re- Gould and Nelson (1974) are quite careful to note that their conclusions on the random walk nature of velocity do not rule out an interest elastic velocity function if interest rates are also characterized by a random walk process (H4 and H5, p. 417). They did not address the implicit income elasticity constraint in the construction of the velocity measure. 48 spect to parameter shifts within the sample period. ring allegations of shifts in the aggregate Given the recur- money demand function during the 1970s, this problem needs to be considered. Fortunately, the issue of "shift in the drift" can be investigated using more formal techniques that the visual examination of velocity series undertaken by Haraf. shift in The likelihood ratio test statistic for a location parameter at an unknown point a within a sample developed by Hawkins (1977) and Worsley (1979) applies directly to the "drift shift" problem. Three measures of "velocity" are investigated in Table 1.1. involved the Ml measure of the money stock with a shift nationwide concept NOW accounts in January-April, is 1981. All adjustment for The first velocity the traditional GNP measure of velocity. The second is based on final sales to domestic purchasers (GNP - inventory investment + net exports), deficits have since inventory changes and the large current account been mentioned prominently in recent discussions velocity. The third measure is based on personal income. typically does not show up in "velocity" discussions, but is used the relevant (or only available proxy) in monthly demand specifications (Farr, 1985; Judd, 1984). of This measure aggregate as money It has the advantage of being measured independently on a monthly basis, so that the effect of time aggregation can be investigated on three levels. The 1952. sample This choice was somewhat arbitrary, but is motivated by considerations. Accord period. for periods for the tests in Table 1.1 begin in January, First it marks, roughly, Second, the beginning of the Post- it is typically the starting point of samples money demand specifications estimated in the late 1960s and three 49 early 1970s -- the "golden age" of short run money demand estimates. several experiments with testing for "drift shift" in Third, samples that extended back to 1947 suggested parameter shifts around 1951. Two major Table 1.1, conclusions emerge from the test results "drift measures in regardless of the velocity measure chosen and regardless of the level of time aggregation. of reported shift" in The first is that there is no evidence a random walk model of the various during the 1952 through 1981 sample period. that when the sample is extended through 1985 the The velocity second is tests unanimously second of these two conclusions is particularly interesting point to a change in the drift parameter in 1981. The because of the existing controversy over the timing of velocity and/or money demand shifts. early as occurred bearing Haraf (1986) appears prepared to date a shift as late 1979 or early 1980. Others have that shifts as a result of financial deregulation that permitted interest checkable deposits such as NOW and SNOW accounts (Higgins Faust, 1983; Roth, 1984; Paulus, 1986). or argued 1980 admits and Dating shifts as early as 1979 the possibility that the shift could result from a monetary policy regime change in 1979 or the credit controls experience in 1980. Dating the shifts as late as the end of conclusions dubious. the nationwide measured money 1981 makes such Dating shifts in late 1981 would seem to suggest extension stock of NOW accounts as an influence, but here has already been adjusted to the filter the impact of portfolio shifts associated with this regulatory out change. Dating shifts as early as late 1981 casts doubt on the proposition that the shift is related to the introduction of checkable deposits with market determined own rates, since SNOWs did not become available until January, 1983. Thus, the time series properties of the various 50 Ml velocities do not seem to support many, alizations about if any, of the popular ration- what has caused the early 1980s to differ from the preceding 30 years. The possibility remains that the conclusions about velocity shifts are illusions investigated created by misspecification. later This question will be while the adequacy of the (0,1,0) ARIMA model of velocity will be discussed here. Table 1.2 presents estimates of the univariate (0,1,0) velocity model for the 1952-85 sample period at all three levels of time aggregation, after ning of 1982. allowing for a shift in the constant term at the beginThis dating of the shift is off by one-quarter from the point identified by the Hawkins1 likelihood ratio tests shift monthly and quarterly data above, but it was chosen comparisons across monthly, gation. 81 to in the facilitate quarterly and annual levels of time aggre- The estimated autocorrelati ons of the residuals for the 1952- sample equations from Table I.l and the estimated equations from Table 1.2 are given in Table 1.3. With one marginal exception (quarterly GNP velocity), the hypothesis that the residuals of these velocity models are white noise be rejected. A second feature of Tables I.l and 1.2 is cannot that the introduction of the shift in the constant term at the beginning of 1982 has greatly reduced the difference in the residual standard error the 1952-85 sample compared with the regardless corresponding 1952-81 sample, of the velocity concept or level of time aggregation. money and income measures used in these regressions are constructed conventional arithmetic and if averages. If geometric averages The as had been used the residuals at the monthly level are white noise, 51 of then the only consequence of residual variation. intervals the aggregation would be a reduction in the In fact, aggregating from quarterly to annual time produces a greater reduction in the residual that the 1.731 factor predicted by a white noise process. The conclusion from this analysis is the data do not reject the hypothesis that the random walk model of Ml velocity is well characterized by a "shift in the drift" at the beginning of 1982. explanation of this phenomenon is the financial One possible deregulation hypothesis (FDH). As noted above, a number of analysts have argued over the past several years that the introduction of interest bearing checkable deposits both in the form of NOWs with effective interest rate ceilings above zero or in the form of SNOWs with market determined rates fundamentally (e.g., Ml changed the nature of the current measure of Ml. Paulus, measure have Some 1986) have gone so far as to argue that if the current is stripped of the other checkable deposit money is measured as what was called MIA in 1980-1, component and then the velocity investigate the FDH we have examined the time series behavior relationship of the 60s and 70s reasserts itself. To of a number of other monetary aggregates, including the currency component of Ml, the adjusted monetary base published by the Federal Reserve Bank of Under St. Louis, and Ml net of other checkable deposits (MlA) . FDH, none of these aggregates should exhibit any "shift in the drift" in the early 1980s. The results for the currency component of Ml are shown in Tables 1.4 and 1.5. of the Regardless of how velocity level of time aggregation, there appear to be breaks in the "drift" of currency velocity. at the is measured, and regardless two distinct The first of these occurs end of 1961 or beginning of 1962 and the second occurs 52 around the third quarter of 1981. With different velocity concepts the other of these shifts may appear stronger, one or and in some cases the shift in late 1981 is either not significant or only marginally significant when measured by the Hawkins likelihood ratio However, in location parameter tests. of case, at statistic. this test points a possible change the same point indicated for the in the Ml velocity The hypothesis that the same "shift in the drift" occurred currency 1.5 every test in velocity as occurred in Ml velocity is investigated in Table where dummy variables are introduced for a shift at the beginning 1962 (D62) and at the beginning of 1982 (D82). In every case the estimated coefficients on these dummy variables are highly significant. The magnitude of the estimated coefficients on D82 also suggests the 1981 "event" should not be considered solely an Deposit problem. 1981 measured shift Checkable The shift in the growth rate of currency velocity in in Table 1.5 is of the order of 40-60 percent in the growth rate of Ml velocity measured in Table strong rency Other of 1.3. result in support of a shift in the drift parameter of velocity against that ARIMA model can be interpreted as strong the This a curevidence the FDH that the difference behavior of Ml velocity should be attributed solely to the advant of interest bearing checkable deposits. Certainly interest rates on currency have not been deregulated, nor is there anything in the financial deregulation literature to suggest that It should be noted that the currency component of Ml manipulated for NOW shifts in 1981. 53 has not been an implication of this deregulation is a continuing increase 3 in the demand for current relative to various income measures. The hypothesis occurred a "shift in the drift" simultaneously with that of Ml currency Table 1.6. is that ingly similar to those in Table 1.2 and 1.5. the there it is not or quarter-to- In every case, regardless of a highly significant shift in the drift the beginning of 1982. income applied, of base velocity The estimated coefficient of the shift dummy variable is generally 65-75 percent of the estimated for the corresponding Ml velocity concept in Table 1.2. were the necessarily The results are strik- level of time aggregation or the concept is around in growth rates of the base, hence the results presented here do not immediately follow from those in Table 1.5. of is investigated that currency should dominate the month-to-month quarter velocity While the dominant component of the adjusted monetary base currency in terms of the level of the base, true of base coefficient Several tests undertaken that are not reported in detailed tables here. Adjusted Monetary Base velocities were investigated for a First shift around the beginning of 1962 such as found for currency velocity above. In no case was the estimated coefficient of a shift dummy variable at The underground economy hypothesis could be invoked to argue that much of the shift in the drift of Ml velocity should be attributed to the change in behavior of currency velocity, and the latter is a consequence of the rapid growth of the cocaine and other hard drug industries, which are necessarily currency oriented transactions and are not properly measured in GNP or other income concepts. There are several problems with this hypothesis. First, it cannot be refuted, since accurate data on the demand for currency for such transactions will never be available. Second, the shift in the currency velocity is smaller than the shift in Ml velocity and since currency is only about 20 percent of total Ml, there remains a large shift in the velocity of the checkable deposit component of Ml. 54 this point significant for the Adjusted Monetary Base. Second, autocorrelation functions of the residuals of the regressions in Table residual 1.6 were examined carefully. The monthly and the reported quarterly autocorrelation functions gave no indication of any signifi- cant serial correlation. some evidence of a weak first order moving average process, but given the relatively In some of the annual regressions there small samples, these estimated coefficients significantly different from zero. is are not Third, after introducing the shift dummy beginning in 1982, the estimated standard error of these (0,1,0) ARIMA as models of Adjusted Monetary Base velocity are virtually constant the sample period is extended from the end of 1974 to the end of 1981 to the end of 1985. The results ments erties logical question that follows from all of these is "Why?". time The following section will review numerous series argu- in the money demand literature that suggest the above time propshould not be observed. 55 TABLE 1.1 Tests of the Stability of DS Models of Velocity m*[lnVt - lnVt ] - a + e Full Sample Drift Estimate (Annual Rate) Velocity Measure Data Interval 1. Personal Income Monthly (m=1200) 52,1-85-12 2.800 (.36) 7.21 81,9 5.66** 2. Personal Income Monthly (m-1200) 52,1-81,12 3.455 (.36) 6.90 81,9 2.50 3. Personal Income Quarter (m-400) 52,1-85,4 2.77 (.34) 3.93 81,3 6,40** 4. Personal Income Quarter (m=400) 52,1-81,4 3.44 (.29) 3.18 54,3 2.50 5. GNP Quarter (m-400) 52,1-85,4 2.58 (.41) 4.73 81,3 4.76** 6. GNP Quarter (m-400) 52,1-81,4 3.20 (.39) 4.25 54,3 2.15 7. Final Sales Quarter (m=400) 52,1-85,4 2.76 (.31) 3.66 81,3 4.91** 8. Final Sales Quarter (m-400) 52,1-81,4 3.26 (.30) 3.29 67,1 2.01 9. Personal Income Annual (m-100) 52-85 2.87 (.39) 2.28 81 5.32** 10. Personal Income Annual (m=100) 52-81 3.43 (.27) 1.46 54 2.70 11. GNP Annual (m=100) 52-85 2.66 (.42) 2.46 81 4.28** 12. GNP Annual (m-100) 52-81 3.19 (.33) 1.79 54 2.78 13. Final Sales Annual (m-100) 52-85 2.84 (.33) 1.95 81 4.52** 14. Final Sales Annual (m-100) 52-81 3.28 (.26) 1.39 73 2.26 Sample Period 56 LR Statistic s TABLE 1.2 (0,1,0) ARIMA Models of Velocity 1952-85 with Constant Shift a + @ D, fc + e m*[lnVt - lnVt J kt Velocity Measure Data Interval a B Personal Income Monthly (m=1200) 3.45 (.37) -5.56 (1.07) 7.00 1.99 Personal Income Quarter (m-400) 3.44 (.32) -5.71 (.93) 3.48 1.48 3. GNP Quarter (m-400) 3.20 (.40) -5.28 (1.18) 4.42 1.62 4. Final Sales Quarter (m=400) 3.26 (.31) -4.28 (.91) 3.40 1.67 5. Personal Income Annual (m=100) 3.43 (.31) -4.78 (.90) 1.69 2.27 6. GNP Annual (m-100) 3.19 (.36) -4.55 (1-06) 2.00 2.48 Annual (m=100) 3.28 (.28) -3.72 (.82) 1.55 2.39 7. Final Sales 57 d-w TABLE 1.3 Estimated Autocorrelations for (0,1,0) ARIMA Models of Velocity Table ( l i n e ) 1 2 Lag 3 4 5 6 Chi Squared(6) 1. 1.1(2) -.07 .07 .01 -.03 -.04 .05 5.2 2. 1.2(1) .00 .08 .04 -.02 -.03 .03 4.6 3. 1.1(4) .17 .06 .04 -.05 .03 .03 4.4 4. 1.2(2) .25 -.06 -.04 -.13 -.06 -.09 12.7 5. 1.1(6) .11 .13 .16 -.06 -.06 .08 8.5 6. 1.2(3) .18 -.11 -.09 -.12 -.12 .03 11.2 7. 1.1(8) .08 .09 .10 -.14 .07 .14 8.0 8. 1.2(4) .16 -.05 -.10 -.16 -.02 .06 8.7 9. 1.1(10) .06 -.09 -.01 .02 -.10 -.02 .7 10. 1 . 2 ( 5 ) -.17 .04 .06 -.01 -.09 .04 1.4 11. 1.1(12) -.17 .14 .02 .08 -.18 .00 2.5 12. 1 . 2 ( 6 ) -.28 .04 .16 -.03 -.14 .02 4.1 13. 1.1(14) -.06 .10 .04 .05 -.28 -.02 2.9 14. 1 . 2 ( 7 ) -.20 .03 .14 -.01 -.22 .02 3.5 58 TABLE 1.4 Tests of the Stability of DS Models of Currency Velocity m*[lnVt - ln(v tl )] - a + € Full Sample Drift Estimate (Annual Rate) Velocity Measure Data Interval 1. Personal Income Monthly (m-1200) 52,1-85,12 1.937 (.31) 6.27 81,9 4.42** 2. Personal Income Monthly (m-1200) 52,1-81,12 2.364 (.33) 6.31 61,6 3.43** 3. Personal Income Monthly (m-1200) 62,1-81,12 1.573 (.39) 6.09 81,9 2.55 4. Personal Income Quarter (m=400) 52,1-85,4 1.904 (.27) 3.12 81,3 5.44** 5. Personal Income Quarter (m=400) 52,1-81,4 2.367 (.26) 2.87 61,4 4.19** 6. Personal Income Quarter (m=400) 62,1-85,4 1.10 (.30) 2.96 81,3 4.57** 7. GNP Quarter (m=400) 52,1-85,4 1.71 (.37) 4.29 62,1 3.55** 8. GNP Quarter (m=400) 52,1-81,4 2.12 (.39) 4.25 62,1 2.86 9. Quarter (m=400) 62,1-85,4 .93 (.40) 3.94 81,3 2.95** 10. Final Sales Quarter (m=400) 52,1-85,4 1.89 (.28) 3.23 62,1 4.57** 11. Final Sales Quarter (m-400) 52,1-81,4 2.19 (.28) 3.11 62,1 4.05** 12. Final Sales Quarter (m-400) 62,1-85,4 1.14 (.31) 3.07 81,3 2.29 GNP Sample Period 59 s k LR Statistic TABLE 1.4 (continued) Tests of the Stability of DS Models of Currency Velocity m*[lnV ln(vtl)] - a+ e Full Sample Drift Estimate (Annual Rate) Velocity Measure Data Interval Sample Period 13. Personal Income Annual (m-100) 52,1-85,1 1.96 (.35) 2.03 62,1 4.16** 14. Personal Income Annual (m-100) 62,1-81,1 1.21 (.34) 1.67 81,1 3.97** 15. GNP Annual (m=100) 52,1-85,1 1.75 (.40) 2.35 62,1 3.34** 16. GNP Annual (m-100) 62,1-85,1 1.06 (.40) 1.97 81,1 2.84 17. Final Sales Annual (m-100) 52,1-85,1 1.94 (.32) 1.86 62,1 4.83** 18. Final Sales Annual (m=100) 62,1-85,1 1.21 (.31) 1.51 81,1 2.33 60 k LR Statistic TABLE 1.5 (0,1,0) ARIMA Models of Currency Velocity 1952-85 with Constant Shifts m*[lnV4 a + BD62 + 0T)82 + e lnV t-i> Velocity Measure Data Interval 6 a e d-w Personal Income Monthly (m=1200) 3.94 (.55) -2.37 (.68) -2.84 (.28) 6.09 2.09 2. Personal Income Quarter (m«400) 3.82 (.43) -2.18 (.52) -3.21 (.74) 2.70 1.60 3. GNP Quarter (m-400) 3.56 (.64) -2.16 (.79) -2.78 (1.11) 4.05 1.65 4. Final Sales Quarter (m=400) 3.68 (.47) -2.24 (.58) -1.77 (.81) 2.99 1.82 5. Personal Income Annual (m-100) 3.77 (.45) -2.09 (.55) -2.84 (.78) 1.42 1.71 6. GNP Annual (m=100) 3.41 (.62) -1.90 (.76) -2.67 (1.08) 1.96 2.22 7. Final Sales Annual (m=100) 3.68 (.44) -2.17 (.54) -1.76 (.77) 1.40 2.21 61 TABLE 1.6 (0,1,0) ARIMA Models of Adjusted Monetary Base Velocity 1952-85 with Constant Shifts m*[lnV - l n V t l ] - a + BD82t + e Velocity Measure Data Interval a 6 s d-w 1. Personal Income Monthly (m-1200) 2.74 (.36) -3.98 (1.04) 6.78 2.17 2. Personal Income Quarter (m-400) 2.74 (.28) -4.20 (.82) 3.09 1.55 3. GNP Quarter (m-400) 2.49 (.39) -3.78 (1.13) 4.26 1.60 4. Final Sales Quarter (m-400) 2.55 (.31) -2.78 (.90) 3.37 1.86 5. Personal Income Annual (m-100) 2.76 (.28) -3.70 (.82) 1.55 1.95 6. GNP Annual (m=100) 2.52 (.36) -3.44 (1.03) 1.95 2.45 7. Final Sales Annual (m-100) 2.61 (.28) -2.65 (.81) 1.52 2.19 62