A tax shift for marriage; indexing effect; R&D bonus

From a tax standpoint, it should generally pay to be married next year.

It used to be that many couples who were legally married, filing a joint federal tax return, paid more in taxes to Uncle Sam than a couple living together without being married and filing separate returns.

Well, Congress moved to reduce this so-called ''marriage penalty'' in the major tax law of 1981.

Moreover, notes Daniel Feenberg, a research associate of the National Bureau of Economic Research (NBER), the reduction in this ''virtue tax'' is almost overshadowed by an extra tax bonus for married couples filing joint returns. In fact, many of these couples will be better off from a tax standpoint than those couples living together without legal binding and filing separately.

This will be true next year when the ''secondary earner's deduction,'' permitting a couple filing a joint return to exempt 10 percent of the secondary worker's earnings (but not more than $3,000) from taxation, takes effect.

This is one of the interesting economic tidbits flooding out of the National Bureau, which was headed until recently by President Reagan's new chairman of the Council of Economic Advisers, Martin Feldstein. Here's a look at three such research papers.

The Feenberg paper notes that before the 1981 tax law changes, some 18.7 million couples paid an average ''marriage penalty'' of $481 in 1979. Another 20 .3 million married couples, however, enjoyed a tax break amounting on average to from their joint tax returns.

With the new secondary earner's deduction in place, only 2 percent of couples will be ''abused'' financially by the tax law. The marriage bonus, however, is increased by $1.7 billion, to $14.7 billion in tax savings overall.

So, Mr. Feenberg notes, Congress did not leave the tax system any more neutral regarding marriage vs. living together. It just shifted the balance toward tying the knot legally. Further, the new system somewhat reduces the progressivity of the system, since it was relatively well-off couples with both husband and wife earning incomes who tended to be hit hardest by the marriage penalty.

There has been considerable fuss over the last few years about the indexing of social security benefits for the retired. It was argued that because most older people live in their own paid-for houses and do not move often, they were not hit as hard by rising mortgage interest rates as younger folk. Since social security benefits are indexed to the consumer price index (CPI), and that index reflects higher mortgage interest rates, the retired were doing especially well in these high-interest years, it was argued.

Two NBER researchers, Michael J. Boskin and Michael D. Hurd, took a look at this thesis and found it true. But they add that in both 1979 and '80 the CPI overstated inflation for everyone who didn't move - old or young - by 2 or 3 percent.

Removing this housing factor distortion, Mr. Boskin and Mr. Hurd made up their own inflation index for the elderly which takes special account of their typical living expenses. Then they compared it with the CPI for the years 1961- 81, corrected for housing. They concluded that the rise in the cost of living was about the same for all age groups.

The use of the CPI for indexing social security payments resulted in about $5 .7 billion worth of cumulative overpayments to the retired from mid-1978 through mid-1981, they reckon.

Starting next year, the Bureau of Labor Statistics will use a rental equivalent, rather than mortgage interest rates and house prices, in calculating the consumer price index. And in 1985, that new index will be used in reckoning social security payment increases.

A corporate investment in research and development pays off handsomely in improved productivity. That's the conclusion of a recent study by NBER associates Kim B. Clark and Zvi Griliches.

Some earlier research using broad, aggregate industrial data indicated that the productivity payoff from R&D had declined somewhat in the economically turbulent 1970s from the '50s and '60s. This, it was thought, might be a partial explanation for the decline in productivity growth in the United States in the last decade.

The two authors, both Harvard University economists, have looked at more detailed data on some 924 individual, narrowly defined ''business units,'' drawn from the Profit Impact of Market Strategies project of the Strategic Planning Institute, Cambridge, Mass. The ''units'' are in many cases divisions or other portions of corporations, dealing with a well-defined set of products and customers, listed in the Fortune 500, and almost all of them are found in the Fortune 1,000.

With the use of this more detailed data, they found that the productivity return on R&D actually did not drop in the 1970s as previously calculated from aggregated data. It averaged 18 to 20 percent, about the same as before.

The productivity referred to is ''total factor productivity,'' which considers inputs of capital and purchased materials as well as of man-hours of labor.

The authors' econometric study also indicates that the productivity return on R&D was higher in businesses where significant technological change had taken place within the previous eight years - about 24 percent.

Further, those businesses that have accumulated a larger stock of proprietary processes, shown by ownership of patents and ''trade secrets,'' do have higher productivity than other business units with a lesser accumulation of such know-how. But the productivity return on new investment in R&D for both these groups of businesses shows no discernible difference in the data.