The rich are paying too much tax, and tax cuts made it worse. While not sounding logical, top tax payers pay a higher share of the tax than before the tax cuts. On the flip side, a logical sounding hypothetical may have no basis in reality. Consider prevalent economic theories that fail to explain why the US grew at less than half the normal pace in the last decade. If supply-side theory worked, the lowest combination of the top marginal rate and capital gains rate in eight decades would have generated strong growth. If Keynesian stimulation worked, the largest Federal spending and deficits since W.W.II would have generated strong growth. If monetary stimulus worked, the aggressive action of the Fed would have brought strong growth.
If these theories failed, maybe it's time to consider a new one that actually fits the data. Robust sustained growth requires a low average tax rate on the wealthy with a high marginal rate. The average tax rate on the wealthy has to be low enough that it is worthwhile for them to run and/or fund businesses. Their marginal rate has to be high enough that they avoid taxation with deductible or depreciable expenditures that grow the value of their business.
If Marginal Rates are Too Low
Decisions to plow money into businesses and grow the economy or to pull it out as personal income are highly sensitive to marginal tax rates. If marginal tax rates are too low, it is cheap for the wealthy to pull equity out of their businesses as personal income. Their personal income shoots up leaving less money within businesses. The wealthy become more likely to live a life of ease off existing wealth. Making real investment in productive capacity, which can be deducted, becomes relatively less attractive than making financial or speculative investments, which cannot be deducted. A dollar left in a business and likely spent on wages, capital equipment, research and development or marketing has a larger growth effect than a dollar pulled out and perhaps used to bid up the prices of such things as beach front property, stocks or gold.
When marginal tax rates are too low, growth weakens with a lag. The top marginal rate one year appears to have its biggest influence on growth two years later. The capital gains rate has its largest influence on growth five years later. Using these two lead times, the last two tax cuts to have a positive impact on growth were in 1979 and 1982. President Carter cut the capital gains rate from 39.9% to 28% in 1979. After the 5 year lag time, this influenced growth in 1984. President Reagan cut the top rate to 50% and with the two year lag time, this also benefited growth starting in 1984. GDP grew 7.2% in 1984, the strongest growth in the last 50 years.
Since those two tax cuts, every cut to the top rate or capital gains rate has been followed by weaker growth and every increase by stronger growth. In the chart above, the scales for the top rate in green and the capital gains rate in blue are proportioned to their influence on the model in red. For example, the decline in the top rate from 50% in 1986 to 28% in 1988, shown in the green line, corresponds with the drop in estimated growth from 1988 to 1990, shown in the red line. Both lines drop the same distance in the graph. The current top rate of 35% and capital gains rate of 15% are consistent with a baseline growth rate of 0.8%.
The 2003 tax cut began affecting growth in 2005 after the two-year lag time. Its full impact on growth started in 2008 after the five-year lag. Every year since 2005 has been weaker than the 3.3% average growth rate. Growth from 2008 has annualized 0.4% using quarterly data through 2012 Q2, or 0.2% using the four annual data points through 2011.
When you look past the volatility of annual growth, marginal tax rates appear to explain most of the variation in the 5-year growth rate over the last 28 years.
The last time the combination of the top tax rate and capital gains rate was lower than it is now was from 1925 to 1931 when the top rate was 25% and the capital gains rate was 12.5%. This combination of rates influenced growth from 1927 through 1933; this was the lowest growth 7-year period in history where the economy annualized shrinking 3.2%. In June 1932, President Hoover raised the top rate from 25% on income above $100,000 to 63% on income above $1 million. After the two-year lag, GDP grew 11% in 1934; it annualized 11% from 1934 through 1936. While GDP and industrial production set new highs in 1936, employment did not fully recover until the early 1940s.
While not a problem in recent decades, marginal tax rates can be too high. Prior to 1984, they were high enough to interfere with the average tax rate being low and harmed growth. Or looked at another way, the tax brackets were too low. The easy way to have low average tax rates with high marginal tax rates is to have very high tax brackets. Our current top tax bracket of $388,350 is about 8 times last year's per-capita GDP. When America had its best growth, the top bracket was $5 million or over 8,000 times per-capita GDP. The marginal tax rate was 79%.
A comparable bracket today would mean the few people taking more than $388 million in income would pay a 79% tax on the portion above that. The income below $388 million would be taxed at the lower rates. Such a top rate would give great incentive for the wealthiest to plow money into a business where it could be sheltered. If paying a dollar in wages only costs 21 cents in after tax dollars, building and growing a business is too good a deal to pass up. Especially when tax on the growth in value of the business can be perpetually deferred. The growing value of a business does not get taxed until the owner chooses to pull it out as personal income, sells the business for a capital gain or it gets taxed in the owner's estate. The low number of business startups may be due to low marginal tax rates giving insufficient incentive to avoid taxes.
Originally the top bracket of $250,000 for Clinton's 39.6% top rate was 9.3 times per-capita GDP. Today that would be comparable to $449,000. If Reagan's 50% top rate came back, I believe the growth optimizing bracket that balanced a low average rate with a high marginal rate would be about $1.5 million.
Weaker Growth Ahead
In addition to the correlation with marginal tax rates, the idea that the baseline growth rate is near or below 1% is supported by population adjusted housing starts and auto sales. With housing starts running at less than 4 per thousand people, annual growth may soon decline from the current 2% toward the 0.4% suggested in the chart below.
Auto sales in the chart below are not quite as dire, but the latest annual rate at 47 sales per 1,000 people suggest annual growth of about 1.6% for 2012. Percentage increases in auto sales and housing starts look rosy, but coming from such a low base, they are probably less meaningful than the population adjusted levels.
The correlation of the ISM with growth offers further evidence of a structural change. Up to 2005, the three month moving average of the ISM robustly correlated with growth six months later. In 2005 the relationship changed; it started over estimating growth by about 2% and the lead lengthened.
The correlation in the last seven years suggests the lead time may be as much as 18 months and that growth will slow to about 0.4% in the next year and a half. The recent ISM number of 51.5 moved the three month moving average from slightly below 50 to slightly above, but with the changed correlation, this still suggests very weak growth ahead.
Ben Bernanke recently announced the Fed downgraded estimated annual growth for 2012 to between 1.7% and 2%. This sounds innocuous, but actually implies growth will be slow or even negative in the second half of 2012. You get the annual rate by averaging the 4 quarters of GDP this year and calculating the percent change from the average of the 4 quarters last year. To hit 1.7%, GDP would have to shrink at a rate of 0.5% in the last two quarters. If GDP grows at a 1% rate in Q3 and Q4, annual GDP comes in at 2% for 2012.
At the same time, the Fed raised estimates for growth in 2013 and 2014 to 2.5% to 3%. However, this estimate probably does not correctly account for the actual influence of marginal tax rates on growth. Since 2008, virtually everyone has had a series of lowering growth forecasts. For example, in 2010 the Congressional Budget Office estimated 2011 and 2012 would grow at 4.5%. While there is general acceptance that the baseline growth rate is no longer the 3.3% long run average, many still think it is around 2.5% and that a period of below trend growth will be balanced out with a period of above trend growth. The reality may be that the baseline growth rate is less than 1% and that balancing out above trend growth of 2011 and 2012 could mean a recession comes soon.
Growth at 1% or less for the next two quarters and perhaps the next two years will wreck havoc with corporate earnings. Historically, GDP needs to grow at 2% or a little above for real annual earnings to grow. Growth below 2% usually corresponds with declining earnings. Prior to the mid 1980s, a 1% change in the growth rate of GDP corresponded with about a 3.5% change in the growth rate of S&P 500 earnings. Then earnings became more leveraged. It appears earnings growth now fluctuates about 17% for every 1% change in GDP growth. So an economic growth rate below 1% for the next couple of years implies earnings would decline at a double-digit rate for multiple years. If we actually go into a recession the decline could be worse.
If the sub 1% growth rate of the last 4 and a half years is a function of tax policy as the first chart suggests, growth for at least the next two and a half years will likely be weak or negative. The chart above suggests this weak growth would launch the next big plunge in corporate earnings. Earnings volatility has risen the last eleven years to the highest level ever. Earnings appear poised for another big plunge. Weak growth and plummeting earnings would mean stock prices (SPY) have a long way to fall.
Additional disclosure: There is no guarantee analysis of historical data and trends enable accurate forecasts. The data presented is from sources believed to be reliable, but its accuracy cannot be guaranteed. Past performance does not indicate future results. This is not a recommendation to buy or sell specific securities. This is not an offer to manage money.