Now that campaigning for the 2004 elections is under way, we’ll probably hear a lot about Social Security and why the system should be revamped to include individual accounts invested in stocks and bonds. Opponents of reform have already played their cards; they argue that it’s impossible to manage 140 million personal accounts, let alone cost-effectively. They’re wrong, and here’s why.
I recently led a group of professionals, representing many disciplines, to determine whether such a system was feasible. After careful study, we concluded that it is. The administrative architecture of our reformed Social Security structure, which I outlined in testimony before the House Budget Committee Task Force on Social Security, was reviewed by the General Accounting Office and was adopted by the President’s Commission to Strengthen Social Security.
My group started with a list of eight principles that we considered necessary for success. The system would have individual accounts with assets owned by the account holder; reasonable costs; a low employer-administrative burden; opportunities for workers of all income levels to participate; a structure so that inexperienced investors would not suffer poor returns relative to experienced investors; investment choice and a solution for participants who make no investment choice; and lastly, automatic adjustments to match technological innovations offered by the financial-services industry.
Collectively, these principles look a bit like 401(k) plans, which serve about 45 million people. We considered simply adopting the 401(k) model as the solution, but then rejected it because no nationwide infrastructure now exists that is capable of collecting and immediately investing savings from 140 million workers. The only close approximation is the payroll tax, which all employers pay for their employees. The problem with using that process is that the reconciliation of the payroll tax to the individual’s name does not happen until the following calendar year, after the employer sends W-2 forms to both its employees and the government. The system is, in a word, unworkable. An individual cannot invest in marketable securities whose prices are changing when the individual, and how much he is investing at the time of the transaction, are unknown.
The solution we settled on was to employ at least one investment option that did not require timely and detailed individual-contribution data. We called this Level 1.
At Level 1, a worker’s employer sends payroll taxes to the U.S. Treasury. The employer tells Treasury how much of the total payment is from employees who have chosen the personal retirement account option. Treasury then transfers that portion to a private-sector custodian bank, which then invests the total amount in a money-market fund priced at one dollar, a standard industry convention.
The following year, when the contribution is reconciled to the individual’s name using the W-2 form, the fund’s shares are distributed to each worker representing his contributions and interest credit. He then transfers his Level 1 assets to one of three balanced funds, each highly diversified and invested in thousands of securities. These funds are called Level 2.
One of the Level 2-balanced funds would be comprised of assets that approximate the portfolio construction of successful corporate-defined benefit plans. Such plans usually have stocks and bonds in a 60/40 percent weight, respectively. This fund is the default portfolio, where one is invested if no choice is made. The two other funds would have the same asset classes but with different weights. One fund with a higher concentration of stocks would be created for younger workers, and another would be more geared toward less-volatile bonds for those near retirement. Although workers could choose any of the three funds, the funds themselves are designed for differing ages.
After three years or so, when Level 2 has accumulated significant assets and economies of scale, each worker could elect to transfer assets from Level 2 to Level 3. This level would be more like the retail financial-services environment. Although portfolio composition would be comparable to the funds in Level 2, there would be less restrictive investment guidelines. The institutions and providers at Level 3 may want to offer additional goods and services, such as retirement-planning software, to attract assets from Level 2. Each worker could allocate his assets at will among Level 3 providers. This would ensure stiff competition as each provider strives to meet investors’ needs. Costs would most likely be greater than at Level 2, but they would be incurred only if an individual chose to shift to Level 3.
Workers could also move some or all of their Level 3 assets back to Level 2, a platform with fewer features, but also lower costs. The competition across Level 3 providers, and between Levels 2 and 3, would ensure that workers receive the greatest amount of goods and services at the lowest possible cost. While all of these financial-management decisions are taking place, new savings continue to be invested in Level 1 until the reconciliation to each individual’s name. When and if a record-keeping system is developed that tracks workers’ savings as they are made, Level 1 could be disbanded.
Workers would receive an annual statement, have Internet and phone access to their accounts, and be able to make daily choices in Level 3 and annual changes in Level 2. They would receive professional asset management, custody of their assets, and state-of the-art record keeping of their accounts.
The cost of this system through Level 2 — including asset management, custody, record keeping, 175 to 350 million phone calls a year to a customer service center, and other costs, including postage — are estimated to be about three-tenths of one percent of accumulated assets, significantly less than most mutual fund costs.
The modernization of Social Security is not just a good idea — it is a necessity. Those seeking political office in 2004 will have to respond to voters’ questions about how they would reform Social Security. Fortunately for them, much of the work has already been done.
— William G. Shipman is chairman of CarriageOaks Partners LLC and co-chairman of the Cato Institute Project on Social Security Choice.