Roger C. Bird

Competition in Financial Services: Who Is Ahead?

What is the Current Situation?

• Insurance industry: 1/2 win, 4 losses; team breakdown—no unity on the bench
• Banking industry: 3 1/2 wins, 1 loss; potential turnaround champs
• Mutual funds: everybody’s favorite
• Credit unions: dark horse players
• S&Ls: out of the game
• Finance companies: barbarians at the gate (players without rules)

What Is the Nature of the Change?

In brief, in the last 5 years we have seen the most remarkable apparent recovery of the stock market performance and public approval of the commercial banking industry compared with the dark days of 1989–1990. In the meantime, the insurance industry is assailed on all sides for poor management, poor teamwork, poor public relations, and mediocre results. Furthermore, the referees (the courts) keep calling fouls against it. The mutual-funds phenomenon is unstoppable, winning every contest in the public eye. Credit unions are in contest with the banking industry and are a dark horse to achieve an unassailable niche in personal finance markets. The savings and loan industry is finally quelled—some members will become banks—the rest will wither away or be merged. And finally the finance companies (including private insurers of asset-based securities and mortgage companies)—an unpredictable set of players without many rules—have been entering the financial services game in strength, and they are positioned to continue to avoid many regulations, yet still gain market advantage through technology and sheer muscle power. These few players are not the only subcomponents of the financial services industry, but they represent the major action involving recent changes and controversies.

Market Share. Market share of assets is one of the few numerical criteria of achievement that can be used for comparison purposes among financial services participants. If available, return on stockholder’s equity would be another important measure, but it would leave out all of the mutual style organizations that are mainly found outside of commercial banking. Table 10-1 shows the latest available figures on shares of assets by industry segment, reflecting the relative decline in both banking and insurance since 1980 compared with mutual funds and pension funds. Thus far, life insurance companies at least have managed to avoid losing much share (strike one loss for banks and 1/2 win for insurance companies).

The insurance industry is under stress in several other arenas, however, including technology and innovation, regulation and litigation, public relations, and globalization. Banks are subject to some of the same stresses but appear to have coped better than the insurance industry in several respects. Moreover, in spite of potentially rancorous divisions among banks based on size and locale, as a group they have acted far more cohesively than have insurance companies when subject to these stresses.

Technology and Innovation. The interactive computer revolution and the use of computerized databases for marketing, distribution and analysis, and new product innovations threatens to leave the insurance industry in the dust. Banks, credit unions, mutual funds, the brokerage industry, and especially the major finance companies (for example, GMAC, GEC, etc.) have all automated their relationships with their clients at a much faster rate than have insurance companies through automated teller machines (ATMs), automated telemarketing systems, various proprietary on-line banking and brokerage services, and on-line marketing and delivery through the Internet. The result has been a booming set of new business opportunities in areas traditionally reserved for insurance companies and old-line retail brokerages. The use of systematized depositor databases coupled with credit application information, all coordinated with new globalized credit reporting systems has enabled explosive growth in credit use (especially credit cards) and interactive client account management. This innovation enables banks, finance companies, mutual funds, and credit unions to all increase their shares of household usage of financial services. Their management of household assets has of course grown relative to insurance companies as well. Banks also have rebounded from their loss of place in the business sector, with a rebirth of commercial and industrial lending. They are learning how to apply mass computerization techniques to small business loans as well, following their success in the household area.

Financial innovation in the form of securitization of credit card debt, mortgage debt, auto debt, and small business loans has only been possible because of the ability to use computers to standardize the problem of packaging and servicing myriad individual loans and debts. This has opened up new avenues for use in marketing and distribution of information derived from interactive on-line databases with sharing of massive amounts of information through credit agencies and other clearing houses. Insurance companies, in the meantime, are having trouble agreeing to share their company data among state regulators, (the NAIC) let alone among themselves through an industry clearing house. The great potential ability for insurance companies to use their collective knowledge about assets and liabilities, risk preferences, and financial needs of their clients is no less than the banks’ current ability. But they appear to be late in the game. On this issue the score is probably banks, 1; insurance companies, 0.

Regulation and Litigation. The banks are clearly advancing in this arena as well. Following the Barnett Bank decision and its various interpretations in federal courts, national unlimited geographic sales of any and all kinds of insurance anywhere in the U.S. are effectively permitted as long as the operation is centered in places of less than 5,000 inhabitants. This is a loophole that a fleet of Brinks trucks could drive through. Some state legislatures (for example, Rhode Island and Pennsylvania) are attempting rear-guard action at the behest of the insurance agents’ lobby, but the insurance companies themselves have conceded the field and many are rapidly designing marketing and distribution strategies for alliances (affiliations) with various banks and bank holding companies (BHCs). Even though the Barnett decision referred to national banks, state-chartered banks are following in train using the principle of reciprocity in the dual-banking system.

In the Congress, it is only a matter of time before the Glass-Steagall Act of 1933 is effectively repealed as regards banking powers. The act

• separated commercial banking from investment banking except for federal, state, and local government securities
• established by amendment the concept of deposit insurance
• generally prohibited banks from engaging in business activities unrelated to banking (such as insurance)

Most observers expect the repeal in this Congress in order to avoid controversy in the proximity to the next presidential election in the year 2000. Several versions of legislation effectively repealing the act are in play currently in the Congress. But one common thread among all parties is that banks or BHCs will be permitted to operate in every financial services field, including insurance and securities underwriting (with very few restrictions on the type or scope of entry). Relative to players already in the field, debate still lingers about the powers of nonfinancial entities (called commercial firms) to enter the fray and buy, be bought, or affiliate with, banks and BHCs. The finance companies are the major players here, since the main ones are owned and/or controlled by major industrial-type firms such as GM, GE, Ford Motor, and such. Microsoft is also in this act due to its alliance with Citibank, and its ambitions to facilitate financial transactions on the Internet. On this score, Intuit is a potential player as well.

The major issue with respect to conglomeration by banks is related to the implicit or explicit use of insured deposits (a publicly protected resource exclusive to banks and credit unions) to enter and/or operate in the new financial services arena. To the extent that the insured deposit funds (held in the Bank Insurance Fund or BIF) are regarded as a public good within the traditional function of banks as money-creating instruments of macroeconomic public policy, natural monopoly rules may apply. For this reason another major distinction among all of the major legislative proposals has to do with the degree of firewalls built between the various functions of banks and other financial institutions. Functional regulation is harder to achieve when functions correspond less and less to institutions or entities. Overall, this concern is unlikely to stop the legislation and some compromise will be reached soon.

On balance, the relative legislative gain for banks versus others in the financial services industry is clear. However, all financial services firms will gain something regarding their ability to more closely match their comparative advantage to the new and exciting markets of the future.

Insurance companies have one big disadvantage in the current legislative and regulatory arena—they have no single regulating body at the national level with the clout of the agencies who regulate the banks (whether federal or state chartered) or the securities firms. Such agencies frequently serve the role of advocates as much as regulators. Insurance companies’ cherished regulators are all at the state level and are too independent to win contests at the national level. This disunity among regulators, even though patched together through the good offices of the National Association of Insurance Commissioners (NAIC), is evident in their missing voice at the referees’ table when the Federal Reserve Board (FRB), the Office of Comptroller of the Currency (OCC), and the Federal Deposit Insurance Corporation (FDIC) are all looking out for their respective banking constituents, and the Securities and Exchange Commission (SEC) is looking out for the securities industry. Couple this with the disunity among and between insurance companies and insurance agents, and it is clear that lobbying efforts will be diffused. Consequently attempts to secure the position of the insurance industry are problematic, and attempts to obtain any new legislative ground are even more problematic.

Other examples of disunity are the lingering disputes between insurance commissioners over settlements in the Prudential Insurance Company cases dealing with past alleged sales force churning of client life insurance contracts. Another episode of disunity is the continuing saga of CIGNA's disputed creation of a run-off company to handle residual environmental and health-related (asbestos) claims under litigation. These cases not only illustrate the problem of unified action, but they also lower federal legislators’ opinion about the regulatory environment of the insurance industry itself, whether life or property-casualty.

Finally, merger and acquisition activity among banks and BHCs has been positively encouraged by the FRB, OCC, and FDIC, with the full acquiescence of the antitrust division of the Justice Department. This is all part of a strategy to strengthen the banking sector and to avoid as many claims on the deposit insurance trust funds or the general budget as possible. The horrible federal budget experience of 1989–1992 when the savings and loan debacle caused massive spillover effects upon the commercial banks (most of which had to be picked up by the general public to the tune of more than $160 billion) has seared itself into the memories of Congress and the regulators. Even though Congress and the regulators now may be going too far in the other direction, strengthening the banking sector has been a clear policy goal of the Fed and the others since 1990. It is showing in merger and acquisition activity, as well as in stock prices of the banks themselves. Here, again, the relative position of the banks has been and will be strengthened vis à vis the insurance industry. Combining these regulation and litigation factors on our subjective scale of wins and losses would again show banks, 1; insurance industry, 0.

Public Relations. Coincidently, banks appear to be less prone to receiving bad press in the last 5 years compared with the insurance industry. The ongoing Prudential case as well as similar settled cases with regard to New York Life and Metropolitan Life have been digested in endless detail in the popular press and on TV exposé shows. Insurance agents continue to get low marks in public polls about their performance. The move to managed health care systems under the aegis of several insurance companies has placed a whole new set of public relations issues in the sights of the press—HMOs are regularly seen as malign insurers (not benign as was more or less the case with most of the Blue Cross/Blue Shield system less than 10 years ago). The old lingering accusations about whole life insurance being a bad financial deal for the average household have not yet been put to rest (Jane Bryant Quinn of Newsweek magazine continues to beat this drum, for example). This impression persists in spite of the fact that the insurance industry has long since moved to emphasize other products and services (such as variable life, and universal life insurance). The public relations advertising of the industry leaders frequently seen on TV continues to stress security of savings and assets, rather than the fact that U.S. insurance companies and agents probably have the best training and the best personal and household financial planning skills in the world. The training of agents is far stronger than that found among banking staffs or mutual fund and securities brokers. Yet it is the latter that are gaining market share among households as a group, probably because (a) they are ahead of the insurance companies when it comes to utilizing the latest computerized marketing and distribution systems and (b) these latter groups also are regularly advertising the skills of their staffs rather than the size or safety of their assets. What prevents insurance companies from advertising their agents’ skills at financial planning? We would guess fear of litigation and disunity between agents and companies. In the meantime, the industry suffers both relatively and absolutely.

The banks are not without their own public relations problems. There is vocal apprehension in the press about the impact of certain massive mergers and acquisitions on local economies. Consumers are noticing restructurings (downsizing of the workforce, in particular), and new and/or rapidly growing fees for services (ATM usage, teller usage, telephone access usage, overdrafts). The financial press has a growing sense that credit cards are being flogged to the public beyond prudence, at the same time that credit card interest rates remain high. Community development lending continues to be a sore point with minority groups as the mergers affect many local community banks, upset established relationships, and regularly appear to set the clock back as regards meeting of Community Reinvestment Act (CRA) goals. In fact, it is Allstate and Nationwide that receive the worst press from a minority relations point of view. The banks seem to keep most of their bad CRA news contained in the back pages of the Federal Reserve Bulletins. Moreover, all of these foul points against the banks have not slowed them down in the legislative or regulatory arenas. And bankers still appear to score higher in public opinion polls than insurance agents do. As a group, they are paying a dear price in terms of public perception for the shenanigans of a few.

Overall, the public relations score on our measure is banks, +1/2; insurance industry, 0.

Globalization. This overused term describes the process whereby firms and industries are (1) taking advantage of the blossoming opportunities for selling, and/or buying, and/or producing abroad in a globalized system of decisions; and (2) obtaining thereby economies of scale or other efficiencies that give them comparative advantages at home as well as abroad. The banks have always had clearer advantages here than the insurance industry, and they are using their new-found strength to prove it. The opportunity for international bank lending and the provision of myriad other international financial services is exploding for U.S. banks. This situation is occurring as the dollar strengthens in international financial markets and as most Japanese and some European banks pull back from their prior incursions, leaving the field to the rejuvenated U.S. commercial-banking sector. Cross-border trade in insurance has steadily grown along with the world economy, but these opportunities are nothing like the growth opportunities presented to bankers. Indeed, in the property-casualty area the global reinsurance market appears flat. While some merger and acquisition activity is regularly occurring among life and P&C companies, none of it appears to be driven by any new market opportunities based on geography, technology, or new products and services—unlike the banks. Again, the score would be something like banks, 1; insurance industry, 0.

Other Financial Service Sectors. We have already mentioned the looming demise of savings and loan banks (S&Ls) and mutual savings banks. The final death throes will occur when the Savings Association Insurance Fund (SAIF) is phased out in 1999, and the merger of the BIF and SAIF is finally accomplished as the last chapter of the savings and loan industry debacle. The charters of the banks and the S&Ls will then be merged and any remaining S&Ls will effectively be banks chartered under federal or state rules. If, as believed, the impediments of Glass-Steagall are removed in the next 2 years, then the remaining banks can turn their complete attention to expanding their businesses into the newer arenas of insurance, mutual funds, and annuities. Therefore the remaining S&Ls will effectively become competitors of the independent insurance agencies, if they are not already.

As noted above at several points, the mutual funds industry has the fastest growth rate of assets of all the major financial sectors. The combination of technology, public awareness, robust stock markets, and favorable legislation on tax deferral (especially with regard to 401(k) plans, 403(b) plans, and IRAs) has led to the booming growth of mutual fund assets. Every sector of the financial services industry wants to be aligned with the mutual funds industry. It is clearly the winning vehicle for long-term household savings in the United States.

Credit unions are an interesting case of an old-fashioned nonprofit cooperative arrangement between close-knit communities-of-interest (for example, employees of a company or union members) that have a common bond. Their tax-free status has made them extremely popular and the number of participants has grown steadily to over 70 million members, even as the number of credit unions has declined through consolidation and merger. The fact that interest paid on insured deposits (called share accounts) is higher than at commercial banks or S&Ls, and that net interest charged is at highly competitive rates (with share dividends declared annually), means that many consumers and household savers are members. Credit unions have steadily increased their share of consumer credit. Moreover, they offer very low-cost, convenient access to the stock, bond, and mutual fund markets. They are also technologically adept at keeping up with (and even surpassing) the skills and capabilities of the commercial banks in most respects. As further financial liberalization takes hold, they will be a recognizable force in offerings of new products and services. Insurance (life products, especially) is a currently underdeveloped area of potential credit-union interest.

The biggest risk to credit-union growth comes in the form of lawsuits over the issue of liberal interpretations of the common-bond requirement of the 1934 Federal Credit Union Act. A federal appeals court has slowed down the growth of credit unions significantly in the last half of 1996 pending an appeal to the Supreme Court. The issue will probably end up in the Congress where some compromise on permissible growth will be reached. On balance, even with some unfavorable ruling, credit-union assets will continue to grow in opposition to banks, which are seen as more and more distant from their small customers.

What Will Be the Effects upon the Industry?

Competition in financial services offerings has never been greater. Household and business clients are all finding new ways of borrowing and saving through the new modes and mechanisms of financial intermediation. The main sources of this development are technology, the push to deregulate, and the federal government’s efforts to encourage tax-deferred savings. To the extent that diversity and competition are healthy, the public is being well-served in the current environment.

Appropriate safeguards are the issue. Rules make competition healthy. The addition of new ways of doing business, and the breakdown of previously established barriers can lead to free-for-all tactics that go well beyond the current ability of regulators to foresee. Thus, three key decisions yet to come are as follows:

1. Can deposit-taking institutions (banks and credit unions) with insured deposits be permitted to operate in any manner they choose with those deposits, or are there restrictions on use or capital requirements for use?

2. Can any firm own or control a bank, or only certain types of firms?

3. Except for government-insured deposits, and assuming financial disclosure to regulators is sufficient and available to investors, can the market be trusted to regularly cull out the weak and inefficient and potentially insolvent, among the new conglomerated financial institutions, or will the old cycles of market euphoria and depression reappear?

These three questions—and there are many ancillary consumer-protection questions—have a remarkable similarity to the questions that have been raised over and over again as financial legislation has been considered throughout American history. Certainly, they were all considered at the time of the passage of the Glass-Steagall Act of 1933—which is now subject to repeal.

We expect that a consensus will emerge that makes incremental adjustments to the current regulatory regime. In other words, evolution not revolution will be the norm. In this scenario, the likely outcomes are already set and in most cases (discussed below) have already been allowed or implemented.

The banks’ new Barnett decision powers will be confirmed. Securities and brokerage firms will be permitted to merge with bank holding companies. Bank holding companies will be permitted unrestricted engagement in investment banking through arms-length nonbank subsidiaries. All customer databases of banks and securities firms will be shared across all subsidiaries of the BHCs for any legitimate uses. Insurance companies will be permitted to merge with securities firms and to become BHCs. Finance companies will be permitted to become or merge with BHCs.

Insured deposit-taking powers will be tightly controlled with more, not fewer, firewalls between deposit-taking and other activities of banks and BHCs except for customer database sharing. The Bank Insurance Fund for insured deposits will continue to be regulated by the FDIC with new capital and risk-assessment requirements. However, the powers of the Federal Reserve board and the Office of Comptroller of the Currency will be realigned giving more power to the Fed over BHCs (mainly for reasons of continued macro-economic control over the creation of money and the conditions of credit availability) and less power to the OCC over the national banking subsidiaries of the BHCs with regard to charters and powers. The BHC form of organization will progressively win out both at the state and federal levels. While state regulation will continue as an anachronism of bygone days, mainly with regard to licensing and granting of state franchise, it will progressively recede in importance. Consolidations at the state level between banks, securities, and insurance regulators will increase, paralleling the other merger-acquisition-consolidation moves in the financial services industry.

How Does This Change Affect Clients?

How will the typical client of these new financial conglomerates be treated compared to today? The major differences will occur due to technology rather than the regulatory regime—in other words, substance, not form, will rule in determining what services are provided and in what packages they come. Cross-marketing and cross-selling of services—except for gross and illegal tying arrangements (for example, making a loan contingent upon purchase of insurance)—will be more commonplace than today, but the mechanisms of cross-selling and -marketing will be ever more sophisticated. Customer databases covering a household’s actual or imputed (estimated) assets, liabilities, income, buying habits, and family-member biographies will all be integrated. Regular telephone or on-line interviews of household members to establish needs, goals, and risk-tolerance profiles as well as contingency plans will also be co-integrated with the basic data. Household financial planning (actually a form of lifetime portfolio management for each household) can then begin to be treated en masse with the latest software techniques of portfolio optimizers. The financial planning capabilities of financial services providers will be delivered via the Internet to those who prefer this mode. But, the main point is that financial planning can and will become a mass market through the immense information-processing capabilities of the comprehensive financial services firms.

What Should Be Done?

In this new world, there will be a premium on those financial planners who can best communicate the lifetime portfolio management process to clients, and yet still treat the clients as individuals with individual needs and goals. Success in the arena of financial services will go to those with such skills and training.

Herein, perhaps, lies the salvation of the life insurance industry and its agents. No other industry group has devoted as much time and energy to understanding client financial needs and assessing goals. With their superior institutionalized training and professional development programs, they are in a strong if not unique position to capitalize on their human skill base in this brave new world of financial services described above.

Where Can I Find Out More?

• David S. Kidwell, Richard L. Peterson, and David W. Blackwell. Financial Institutions, Markets, and Money, 1995. Dryden Press.
• Robert E. Litan. The Revolution in U.S. Finance: Past, Present, and Future. The American College, 1991.
• Clifford E. Kirsch, editor. The Financial Services Revolution, Irwin, 1997.
• Board of Governors of the Federal Reserve System. Federal Reserve Bulletin, Washington, DC, monthly, and Flow of Funds, quarterly.

Price indices of various types have found their way into the core of the financial planning toolkit because they are necessary to assess inflation. Applications of such indices are critical to forward-looking estimates of taxes, social security, health care, business valuations, insurance needs, and estate and annuity planning, all of which have an inflation component. This section will focus on measurement and selection of price indices rather than applications (except for illustrative purposes). It will be clear, however, that correct application is dependent upon knowing about such measurement and selection.

What Is the Current Situation?

The Basics. All price indices are calculated using weighted averages of price relatives. A price relative is the ratio of an item’s (or group of items) price at one time to the same item (or group of items) price at another period of time. For some purposes the price relative may be composed of ratios across geographic areas rather than across time periods. Such indices are primarily used for state-by-state or international comparisons of prices and costs of living or costs of doing business. They will not be discussed in any detail here, but their principles of formation have many similarities to the period-to-period price indices.

Table 10-2 shows the price indices calculated for a variety of sets of items over a selection of periods. The sets shown for illustration are (1) consumer goods and services (CPI-U), (2) consumer goods at wholesale prices (PPI-C—actually defined as producer finished goods for consumers, (3) export goods (PDX—a chain-type price index), (4) all goods and services produced in the domestic economy (PDGDP—a chain-type price index), and (5) price of personal consumption expenditures (PDCE—a chain-type price index). This sample clearly illustrates the variety of indices available as well as the different stories about inflation that they tell. Inflation here is simply the annual average percentage rate of change of any one index between any two periods.

The price relatives of the items which compose each index are generally calculated in the same manner, and several items (for example, trucks/autos) will appear as subcomponents in all five of the indices. In fact, the great majority of item price relatives for all of these indices are created by the statisticians at the Bureau of Labor Statistics (BLS). The difference in treatment will primarily be due to the treatment of the weights in each weighted average; and here lies the second choice facing the statisticians and practitioners, the first choice being which items should be in the basket of items whose price change (inflation) is to be measured.

Choice of Items. At first glance, this decision seems straightforward. If consumers are involved, one should choose those items (goods and services) that consumers buy (for example, cars and haircuts). But, these items in fact change more or less over time due to (1) technology, (2) tastes, and (3) new items (goods or services) being introduced and old items being withdrawn. Technology here encompasses all the changes in function and availability, frequently called quality change, that accompany new discoveries (inventions) and new ways of doing things (innovation), whether in packaging, delivery, marketing, quality control, distribution (for example, supermarkets versus corner groceries), or any of the other many ways in which change occurs in the marketplace aside from invention. Tastes change due to a variety of reasons which are often misunderstood. For example, choices of entertainment have dramatically changed through time and not simply because of technology (for example, video) or advertising. This means that the entertainment items chosen will also have changed dramatically. Also availability of new medical techniques and treatments has led to lifestyle choices, due in part perhaps to advertising, that are arguably no healthier than prior home remedies and treatments. Finally whether items change due to technology or tastes, price relatives will have a less clear meaning as a measure of inflation than if the items in question were unchanged. Isolating the effects of such quality and taste changes then becomes a major task of the statisticians who compose the indices. This task is aside from the problem of what to do when a wholly new item is introduced or an old one disappears.

Choice of Weights. Even if items were to remain constant in quality once introduced, one has the problem of choosing the appropriate weight for each item in the basket of items. Again, if consumers are involved, the appropriate market basket of items presumably should reflect consumer choices in their budgets—or at least that portion of their budgets which they are willing and able to spend on market-priced goods and services. This latter caveat means that we exclude such normal budgetary items as taxes, charitable contributions, and all forms of savings. However, such a basket clearly does not correspond to any total cost of living. Here, the main issue is changes in the appropriate budget weights from one period (called the reference period) to any other time period. What period should be that reference? After all, as relative item prices change at different rates, consumers will choose more of some items and less of others; this is the substitution effect (for example, choosing more chicken and less beef as chicken prices fall compared to beef prices). Also as overall budgets change, the proportions of some items will increase or decrease depending upon tastes and preferences (this is the income effect—for example, choosing steak versus hamburger simply because income rises). Both of these effects will cause the budget mix to change in a later period compared with the reference period. Should the weights then be chosen referring to the past, the present, or something in-between? The answers to questions such as these have occupied a vast economic and statistical literature for many years.

The overall answer is that the choice of appropriate weights depends upon the question that is to be answered. For many purposes, the most common question is forward-looking: given the choices of budget items of some past reference period, how much more must one pay for the same basket of items today? This means using initial period weights and generates what is called a Laspeyres index. Of the indices shown in Table 10-2, the CPI and the PPI answer that question at least one year from their reference period ahead. The CPI and the PPI try to keep the market basket fixed for all years of the time series shown, whereas the other indices change the weights year by year using a chain-index formulation. These latter indices normalize the index to a base reference period, however.

A similar question, which is retrospective, is How much less would someone have paid for the same basket of today, one year earlier? Again, taking the reference period as the base, look one period earlier to obtain this answer from the table for the CPI and the PPI. This means using terminal period weights and generates what is called a Paasche index.

A third type of question would be, How much more or less must one spend this year versus an earlier or later year in order to keep oneself as well off as in the base period (that is, what would it take to keep one’s standard of living constant)? It is the answer to this last question that is the goal of a true or ideal cost-of-living index (sometimes called a COLI). One choice of weights which might approximate this goal would require a geometric average of the two types of indices (the Laspeyres and the Paasche) noted above for any two periods. This calculation would generate what is called a Fisher ideal index and is the system currently used to generate the PDGDP index, the export price index, PDX, and the PDCE index shown in the table.

What Is the Nature of the Change?

Bias in the CPI as a measure of a true cost-of-living index has been a contentious issue for many years—certainly back to 1959. It is a vitally important issue to many retirees because the CPI is used on an annual basis to adjust much of federal entitlement spending including social security. It is also used on a regular basis to adjust income tax brackets, certain union wage contracts, as well as many public and private pension plans. The major recent controversy about bias in the CPI as a measure of the COLI stems from the Boskin Commission report to the Senate Finance Committee in December 1996 (titled Toward a More Accurate Measure of the Cost of Living). As might be expected from the discussion above, the main arguments deal with measurement (choices of items, choices of weights), and application (choices of which index for which purpose).

Measurement. Calculation of the CPI by the BLS presents an awesome task if we consider that their responsibility requires monthly samples of 71,000 item prices from 22,000 outlets for 44 geographic areas. Separately, they gather monthly information from 5,000 renters and 10,000 homeowners for the housing components of the CPI which is the largest component of the overall index (see table 10-3 for the 1982–1984 based weights currently used). The commission found that there were four sources of bias—all on balance in the direction of upward bias (see table 10-4). Upper-level substitution bias refers to the weighting issue discussed earlier as between, for example, chicken and beef or between starches and vegetables. Switching from the Laspeyres-type fixed weights toward a more ideal set of composite weights is estimated to account for an upward bias of .15 percentage points. The lower-level substitution bias, for example, between beef steak and beef hamburger is said to cause a bias of .25 percentage points, even though the degree of switching between such items is not well understood. New outlet bias is another form of substitution bias which refers to the fact that suburban shopping mall superstores (such as Wal-Mart Stores) have replaced downtown outlets (such as Macy’s stores). The main criticism is that this type of switch is only caught with a long-time delay. New outlet bias is estimated to be worth .10 percentage points. All these separate substitution biases have been studied by the BLS itself over many years, and these estimates fall within the range of

BLS estimates although they are not strictly additive. (For example, buying at a suburban discount supermarket may lead to a choice of more beef, less chicken, even though overall our society is choosing more chicken, less beef).

1. Michael J. Boskin et. al. Toward a More Accurate Measure of the Cost of Living: Final Report to the Senate Finance Committee from the Advisory Commission to Study the Consumer Price Index, December 4, 1996, Washington, DC.

The largest recommended adjustment for upward bias targets the issues of new product introduction (such as VCRs) into the CPI and quality change (improvement) in the itemized goods and services already in the item list of the CPI (longevity for automobiles). This purported failure to adequately measure new products and quality change is said to cause an upward bias of .60 percentage points—the largest single source of bias. Adding all of these sources together, as if they are independent, leads to the overall estimate of bias as being of the order of 1.10 percentage points per annum with a plausible range of 0.80 to 1.60. This is a large estimate of bias considering that the CPI itself has averaged 3.6 percent over the last 10 years.

The Commission’s critique of the CPI as a strongly upward-biased estimator of the COLI has itself been criticized on the following grounds:²

• Quality and new-goods bias—The Commission’s estimates continue to lack comprehensiveness and may not apply with equal force to the average consumer; most examples have an anecdotal quality and are not the outcome of rigorous economic analysis.
• Substitution bias—The Commission disagrees with the Bureau of Labor Statistics (BLS) as to the size of some of the components of this bias and how to correct for each of them; the Commission treats these component biases as additive, a view that BLS and other researchers dispute.
• With respect to other issues of quality of life such as crime, the deteriorating environment, and longer commuting time that must be considered for a cost-of-living index, the Commission provides evidence that is scantier than that on quality change and new goods.
• The Commission fails to consider the need for a CPI/COLI for older people due to differences in expenditure weights and patterns and items chosen.
• Research recommendations of the Commission are narrowly devoted to correcting for upward bias rather than being comprehensive toward improvement.

2. Joel Popkin and Company. CPI Commission’s Findings Are Not Convincing: Final Report to the Senate Finance Committee from the Advisory Commission to Study the Consumer Price Index. Prepared for American Association for Retired Persons, January 6, 1997, Washington, DC, pp. 2, 13, 15.

How and When Does This Change Affect Clients?

If the CPI bias problem is to be solved, the solution must arise from the BLS itself. Several corrective measures are already in place for the next major revision in 1998 when the CPI’s new reference base years will be 1993–1995. Although it is unlikely that all of the recommendations of the Boskin report will be incorporated, the anticipated effect of certain corrections will likely be to slow CPI/COLI growth compared to historical inflation rates. Immediately, this means that annual social security payments and tax brackets will be adjusted less than prior projections. In addition, the inflation component used for annuity and pension planning will be less, thus requiring less wealth and income to be set aside for a given real standard-of-living projection. Interest rate and yield assumptions may also have to be adjusted downward because the financial markets might require and expect less nominal return in order to achieve a certain real return after inflation.

Does CPI Correction Have Other Effects Upon Other Indices? The issue of bias in measurement is more far-reaching than the CPI. The other indices shown in Table 10-2 will be affected if new product and quality change adjustments are made by the BLS, since many of the bias-adjusted components of the CPI would affect components of those indices as well. In particular, the PDCE deflator, the PDX deflator and, hence, the overall PDGDP deflator (which uses the former as components) would all be affected. This is good news in the sense that integrated systems of price measurement imply that improvements anywhere become generalized. The PPI will also be improved in part, insofar as the components for consumer goods in the finished goods index are improved. Furthermore, improvements in the techniques of adjustment for quality change would improve all indices to some extent.

Are There Any Other Issues or Opportunities regarding Other Indices aside from the CPI? During the Nixon Administration the federal government chose the CPI as the most appropriate (popular and available) indicator of the cost of living, in a period when inflationary pressures were rising and fairness in treatment of social security (CSI) retirees was a big issue. From 1935 to the early 1970s, intermittent inflation-related adjustments to SSI payouts had been made as Congress and the President saw fit. The 1972 legislation codified the process of indexation, removed discretion, and improved the process of management of entitlements. (Unfortunately, poorly written legislation was enacted, and age 65 retirees from 1972 to 1976 got a double dose of inflation protection. This was corrected using a notch adjustment for retirees after 1976.)

At the time, the dispute about bias in the CPI as a measure of the cost of living was moot because it was so widely used as a cost-of-living adjustment (COLA) mechanism in labor contracts and (some) defined-benefit pension plan adjustments in private and public pension plan management. In truth, the alleged bias was recognized as a possibility much earlier in congressional hearings of the early 60s. The BLS was adamant then (as it is now), that the CPI was not and was never intended to be a measure of the cost of living. Rather, it was intended to be an indicator of inflationary pressures on urban clerical and blue-collar workers (now named CPI-W and covering about 32 percent of the population as of today). Those hearings led to some new procedures and, most important, to broadening the sample size of household expenditures to include all urban consumers (called CPI-U and covering about 80 percent of the population). The essential methods for the two indices remain the same and any biases in one will be reflected in the other.

Recently, as we have noted, the CPI is widely criticized as an inadequate indicator of the cost of living, even though it or its subcomponents are ever more widely used for unintended purposes such as adjustment of tax brackets and rental-contract indexation, as well as the uses mentioned earlier. Significantly, medicare Supplement B inflation adjustments use a subset of the CPI dealing with medical services inflation. The Boskin Commission claims that this particular subset has been upward biased by approximately 3 percent per annum prior to 1995. Are there other indices that might be used for some of these purposes?

Turning back to Table 10-2, we have already noted that the consumer expenditures deflator (PDCE) is a chain-weighted ideal index, composed of almost identically the same components and items as the CPI. It clearly could be used in preference to the CPI as a proxy for the cost of living, since it has more current expenditure patterns built into its very construction. For this reason, it might have been chosen initially as the index for SSI and income tax bracket adjustments if it were more widely known. The Congress and the President could make this change today through a simple legislative adjustment—it would remove much of the product substitution and new product bias in the CPI. Bias due to quality change and outlet substitution would still remain, however.

The components of the PPI are frequently used in long-term purchase contracts by the federal government (defense and procurement items), and by private firms (for example, energy items purchased by utilities, aluminum companies, and pipelines; chemical item feedstocks purchased by chemical companies; pulp purchased by paper companies; and such). The problems of quality change are much easier to handle in commodity-type price indices. Construction contracts are also often written with clauses identifying price adjustments using PPI item prices (steel reinforcing bars, cement, asphalt, and such).

What Should Be Done?

Financial services professionals should be aware of what indices are available when they are called upon to consider inflation adjustments in estate planning cases involving business valuation. Forecasts of these indices are widely available from specialist forecasting firms and are much preferred to casual use of the CPI as the only choice of inflation indicator for such purposes. Similarly for pension planning, the consumption expenditures deflator is probably to be preferred to the CPI for most households due to its more current weighting.

Finally in the field of portfolio management, the CPI is frequently misused as the sole indicator of inflation when calculating inflation-adjusted returns to assets over various time periods. Generally, financial markets will discount inflation expectations into the required rates of return of different financial instruments. For example, when inflation expectations fall, interest rates (discount rates) fall raising the present values of stocks and bonds. But which indicator(s) of inflation is (are) appropriate? In the case of returns to individual investors, who are all considered ultimately to be households or consumers, a consumption-based cost-of-living index may be appropriate. But are all investors in the market ultimately households? That is, are all financial intermediaries such as banks, mutual funds, pension plans, and insurance companies ultimately acting for (agents of) consumers? Or are they acting as well for some other economic entities whose requirements for inflation protection are entirely different, such as state governments, charitable trusts, college endowment funds, property-casualty insurance funds, nonfinancial corporations, and all the other myriad "ultimate" asset holders in the economy?

To question does raise the point that a more broad-based index of inflation, such as the GDP deflator (PDGDP in table 10-2), may in fact be more appropriate as the gauge of financial markets’ inflation expectations than even a bias-corrected CPI. For financial planning purposes for individual clients, cost-of-living inflation protection is probably most appropriate. Financial markets as a whole may be using a different gauge, however, and this difference means that discounts for inflation are likely to be less than what they would be using even a bias-corrected CPI. Therefore required rates of return built into asset values in the market may be less than the typical client’s requirements.

Hedging, using PDCE projections, therefore, would be a preferred alternative for client purposes. This would accomplish most of the COLI result without overstating the differences between client inflation risk protection versus the markets’ built-in inflation risk protection.

Where Can I Find Out More?

• Bureau of Labor Statistics, Monthly Labor Review, U.S. Government Printing Office, Washington, DC, monthly.
• Also see footnotes.

Proposals for social security privatization in whole or part, both in the U.S. and abroad, have been in the news for some time. The larger question is demographic as well as financial: Will the total savings-spending balance between workers and retirees in the mid-21st century be capable of maintaining appropriate living standards for both groups, given that the growth of the U.S. economy and other developed economies is modest, and that the ratio of people over 60 years of age to people in the normal working ages of 16 to 59 is growing rapidly in all countries? This discussion outlines the main issues in the debate and shows that considerable further research must be done soon.

What Is the Current Situation?

Government-sponsored social security schemes in the United States go back to 1935 when the federal government passed the Social Security Act. We know it today as the Old Age, Survivors and Disability Insurance Program (OASDI). The disability portion will not be discussed here, even though the OASI and DI funds are actually held in trust together. We will use the term SSI to refer to all the federal government’s social insurance programs.

The meaning of the trust funds themselves requires explanation. Since 1983 and the Greenspan Commission’s recommended adjustments (discussed later), the trust funds hold surplus accumulations (contributions from employers and employees less benefits paid), plus nominal interest earned on the accumulations, plus some portion of federal income taxes paid on social security benefits by high-income recipients. The basic system is still designed as pay-as-you-go (PAYG) with the current surpluses designed originally in 1983 to cover the expected bubble of baby-boomer retirements beginning about 2010, growing through 2030 (when the last of the baby-boomers retires), and ending in 2050 (assuming an average mortality at age 85). There has never been a reserve fund to cover all obligated and contingent liabilities as is found in all private pension schemes. Thus, by its nature, SSI has always had unfunded liabilities. The issues now arise because some of the underlying assumptions of the Greenspan Commission have not been borne out, and it is now 14 years later. Thus for a 75-year fix, as before, the bubble must be addressed again. Under current (intermediate) projections, social security benefits exceed all sources of revenue by 2019 (13 years earlier than originally projected). Then the trust funds begin to decline, and by 2029 the trust funds disappear (about 30 years earlier than expected) and annual obligations exceed annual receipts. Social security taxes then finance only about 77 percent of obligated benefits, so SSI becomes technically bankrupt.

The Greenspan Commission thought they had made a 75-year fix to the system by delaying cost-of-living adjustments, raising the overall SSI tax rate by 2.4 percent, pushing up payroll taxes on the self-employed, raising the retirement age, and taxing up to 50 percent of high-income recipients’ benefits. The trust funds’ accumulations were the result and they did buy some time, but other aspects of the projections were incorrect. From 1983 to 1996 economic growth was slower, unemployment was higher, money wages barely kept up with inflation, and male participation in the workforce slowed more than expected, raising the retiree-to-worker ratios even faster. These changes to economic assumptions were approximately offset by changes to demographic assumptions, however. The main sources of net forecast error were much larger disability payments (as the government liberalized the definition of disability) and "one shot changes in the methodology used in the projections."³ It is estimated now, on a basis of present discounted values, that the difference between SSI revenues and expenditures averages 2.2 percent of payroll over 75 years. Therefore, one fix to the system would require an increase in the payroll tax by about 18 percent (2.2 percent on top of 12.4 percent currently paid in equal portions by employees and employers). An alternative fix would require a similar percentage decrease in benefits for all current and future beneficiaries for 75 years.

3. Economic Report of the President, p. 103.

What Is the Nature of the Change?

The usual downside risks to this intermediate scenario are legion. Aside from possible forecast-type errors about growth, unemployment, inflation, and worker-retiree ratios as occurred before, new issues with regard to medicare-funding requirements and the overall savings-consumption balance in the economy have become more important. In addition, employees have started to take defensive actions which may also inhibit some of the easy solutions of the past.

The medicare trust funds are in more of a crisis than social security and are projected to be depleted much earlier, perhaps by 2001. A partial solution to that situation is likely to require another payroll tax increase. Moreover, the medicare deficits will also be compounded even more than the social security funds by the baby boomers’ arrival after 2010. Therefore, avoiding even a small payroll tax increase for social security is more critical than usual.

Another solution would be merited because of decreasing payback to workers who continue to be in the system—and hence decreasing political support for the system. After the year 2010, it is estimated that no single (unmarried) workers will earn as much as a 2 percent real rate of return on their contributions (including employer contributions) to the system. A 2 percent real return is considered a reasonable gauge of the break-even real return for an alternative low-risk investment. Whereas, social security paybacks have shown fantastic real returns (well above a 5 percent, virtually riskless real return) for nearly all cohorts and all income groups, married and single, since inception to about 1980, that has changed dramatically. After the year 2000, more and more two-earner couples will be effectively getting paybacks of less than 2 percent real. Only the declining number of single-earner couples will be continuing to earn returns in the 5 to 10 percent range (due primarily to the survivorship provisions of SSI). What has happened to cause this? The baby-boomer demographics and the slowing economy have finally removed the chain-letter effect of the system. As a result, political support for the current system is waning among younger cohorts, and that is a devastating blow to one of the most well-supported and successful social insurance schemes ever devised in the United States.

Privatization of at least a portion of the SSI program becomes more and more attractive to future recipients and current politicians. President Clinton’s quadrennial Social Security Advisory Council has agreed on the basic facts outlined above. They differ significantly on the nature of the fix, however.

Plan A—the Maintenance of Benefits Plan—supported by 6 out of 13, would basically keep the present system, tunes, tightens and taxes again along the lines of the Greenspan Commission (adding 1.6 percent increase in payroll tax after 2145), and permits the steady increasing allocation of up to 40 percent of the trust funds into common equity investments by 2014.

Plan B—the Individual Accounts Plan—supported by two members, would keep payroll tax as is, tunes along the same lines as in Plan A, but most importantly requires a mandatory individual account, like an IRA to be created using a payroll contribution of an additional 1.6 percent (equally drawn from employers and employees). Such contributions while mandatory are not considered a tax, since they would not be included in federal budget calculations. The assets in the accounts would be administered by the SSI system with portfolio allocation decisions made by individuals from among a small number of stock and bond index funds. Payouts would only be available as an annuity after retirement.

Plan C—the Personal Security Accounts Plan—proposed by five members, is the clearest break with the current program. Five percent of the 12.4 percent payroll tax would be extracted and put into mandatory IRAs for all workers. The funds thus set aside would be controlled by individuals outside of SSI. At retirement, the funds could be taken out as a lump sum or as an annuity. Another 2.4 percent would be used to fund existing SDI requirements, with the remaining 5 percent devoted to a minimum level defined benefit, independent of earnings (equivalent to $410 per month in 1996 dollars, which is about 57 percent of the average current retiree benefit). In spite of the reduction of the defined benefit, coverage of the remaining unfunded liability for current retirees is still required. For this, Plan C envisions new federal borrowing with a payback using the proceeds of an additional 1.5 percent payroll tax over the next 72 years.

How Does This Change Affect Clients?

The main motivation behind some form of SSI privatization is the hope that average returns can be raised above the expected low real returns under the current SSI program. Some of these hopes are based on a false premise, however. It is not possible to raise total returns to capital (savings) in the economy as a whole in the future, unless the economy-wide growth rate is raised—and this can only happen if capital investment today or other growth-enhancing initiatives (for example, more education, research and development, etc.) are taking place. Nothing in the plans outlined above is targeted to increase growth rates, however, The proponents of Plans B and C hope that net savings rates will rise due to a more efficient private allocation and use of at least a portion of the payroll tax receipts. But this begs the question about private versus public uses of receipts being more or less efficient. The broader point is that higher returns earned by a portion of SSI receipts may be at the expense of exactly offsetting lower returns elsewhere for other forms of investment, unless the overall economy-wide growth rate is increased. Therefore the total return to savers as they retire may not be raised at all—just the sources will be redistributed. The SSI program will gain but other pension sources will lose.

Another point is raised by the whole notion that the SSI program should have a favored place in the totality of American economic security mechanisms. Why not let the payout fall to 77 percent of current obligations after 2019? After all, this level of benefit would still be considerably higher in real terms than current benefits—it simply wouldn’t be as high in relative terms compared to payrolls at that time. Why not simply change employees’ expectations now rather than change the program itself over the next 75 years. It has more than done its work of raising many of our senior citizens out of the poverty trap of the 1950s and 60s. Why should preservation of SSI payouts be sacred? Why not encourage other sources of income for seniors?

For example, Table 10-5 shows the distribution of income receipts by all seniors over 65 in 1994.⁴ It is clear that there are other major sources of income for seniors that could be enhanced through wise public policy. In particular, with increasing longevity, employment income could be encouraged rather than discouraged as it is now for the over-65 cohorts. Such earnings constituted 26 percent of total income of older persons in 1962 as compared with the 18 percent in 1994. These figures are in spite of the depression era and labor-union inspired penalties for working beyond age 65, in terms of loss of SSI benefits if certain wage income limits are exceeded before age 70. Imagine how much larger this income might be if such penalties were removed. It would help seniors who want to work and help correct the demographic load by raising the proportion of workers to (full-time) retirees. Partial pensions for partial retirement have been proposed in several European countries in order to help mitigate the problem of a shrinking workforce. Prolongation of working life should be more of an option for more people. Aside from the perversity of the penalty noted above, the delayed retirement credit (DRE), which is 4.5 percent for retirees born in 1929 and 1930 and is designed to compensate workers for not receiving benefits when they defer retirement, should be made actuarially fair now. This could be done at an 8 percent rate rather than waiting until 2009 for those born in 1943 and later to receive an actuarially fair return. Encouraging work after first retirement should be a primary goal of public policy.

4. Yung-Ping Chen, "The Role of the Fourth Pillar in the Redesign of Social Security," Studies on the Four Pillars, The Geneva Papers on Risk and Insurance, October 1996, number 81, Geneva, pp. 469–477.

In fact, note that the grey labor market in the U.S. is growing anyway, with many seniors not reporting their labor income in order to avoid the penalties noted above. Other defensive actions that seniors are taking include annuitizing much of their estate (through such gimmicks as taking reverse mortgages), thereby consuming the estate, rather than leaving it as a family asset (bequeathing it). The overall result of this annuitization of resources is that savings rates among seniors are lower now than ever before in our history.⁵ This helps explain a large part of the decline in household savings rates in U.S. society compared to other countries. This trend, in turn, has had deleterious effects upon our national saving and investment rate and our long-term growth prospects.

5. Jagadeesh Gokhale, Lawrence J. Kotlikoff, and John Sabelhaus, "Understanding the Postwar Decline in U.S. Saving: A Cohort Analysis," Brookings Papers on Economic Activity, no. 1 (1996), Washington, DC: Brookings Institution, pp. 315–407.

Some of the arguments for resolving SSI’s unfunded obligations are based on intergenerational equity consideration. The first baby boomers have recently turned 50. In about 15 years they will reach normal retirement age with the effects noted earlier. They should now be bearing as much of the burden on society as is feasible for their own retirement costs. To the extent that they save more(and invest wisely) now, they make it easier on themselves and the rest of society when they retire. Herein lies the dispute between those who think the federal government is part of the problem, not part of the solution. To the extent that the SSI program and its trust funds are federalized and are counted in the federal budget, it is claimed that any potential capital accumulation will forever be held hostage to the wasteful whims of politicians. True savings (and investment) of the current surpluses wrung out of the baby boomers will never bear fruit for their retirement years. Hence privatization is necessary even if the baby-boomer problem is exaggerated! That is the argument anyway.

What Should Be Done?

The decline in bequeathable resources and the rise of annuitized resources as a percent of all resources of the elderly was dramatic from 1960–61 to 1987–90 as shown in table 10-6.⁶ A large part of the shift was due to the relative growth of SSI benefits, which are treated as an annuity in this calculus. Now, SSI is at risk. Whether it is saved another time or not, the best advice to clients is to harbor more resources in other forms, because society as a whole will be ever more stretched to meet the needs of its nonworking, aging (but healthy) seniors with the toil of its younger generations. Therefore the calculus of expected and required retirement income should now be planned to include larger proportions of asset income, occupational pensions, and employment income in order to reach the required replacement ratios of normal working age income. SSI should henceforth be viewed as a risky asset. A calculus assuming SSI benefits to be about 75 percent of today’s benefits in real terms would be a constructive and conservative assumption for retirement and estate planning purposes.

6. Table adapted from J. Gokhale et al., op. cit., pp. 361, 362, Tables 11 and 12.

Where Can I Find Out More?

Aside from the sources in the footnotes:

• Report of the 1994–1996 Advisory Council on Social Security Volume I: Findings and Recommendations, Washington, DC, January 1997.
• Economic Report of the President, chapter 3, Washington, DC, February 1997.
• Joseph Quinn. Entitlements and the Federal Budget: Securing the Future, Washington, DC: National Academy on Aging, May 1996.

Chapter 9

Top

Chapter 11