History versus the Tea Party: Why Most of the Tea Party's Economic Claims are Simply Not True

Debt v. GDP:  Are High Levels of Government Debt Harmful to Economic Growth?

It has become such a Tea Party axiom that high levels of government debt crush economic growth that you think the two must be strongly inversely correlated, that periods of high government debt are followed by periods of lower economic growth and vice versa.

Although there is a relationship, it is not nearly this simple.  

First, of course, the two variables are not independent.  As economic growth declines, government debt increases as tax revenues of a slower national income stream dry up while more people require expensive government programs from unemployment benefits to Medicaid.  Conversely, when economic growth picks up, so does government revenue, while demand for government services drops as fewer people require public assistance.  

Table One:  Debt Versus GDP and Next Year's GDP Growth 1977-2010:

Budget Year:	Pres:	Debt:	GDP:	dGDP+%
1977	Carter	0.699	1.974	12.4%
1978	Carter	0.772	2.218	12.8%
1979	Carter	0.827	2.502	8.9%
1980	Carter	0.908	2.725	12.3%
1981	Reagan	0.994	3.059	5.5%
1982	Reagan	1.137	3.226	6.7%
1983	Reagan	1.371	3.443	11.7%
1984	Reagan	1.564	3.847	7.9%
1985	Reagan	1.817	4.149	6.2%
1986	Reagan	2.120	4.407	5.6%
1987	Reagan	2.345	4.654	7.7%
1988	Reagan	2.601	5.012	7.8%
1989	Bush	2.867	5.402	6.2%
1990	Bush	3.206	5.737	3.4%
1991	Bush	3.598	5.934	5.2%
1992	Bush	4.001	6.241	5.4%
1993	Clinton	4.351	6.578	5.9%
1994	Clinton	4.643	6.964	5.2%
1995	Clinton	4.920	7.325	5.1%
1996	Clinton	5.181	7.697	6.4%
1997	Clinton	5.369	8.187	5.4%
1998	Clinton	5.478	8.626	5.8%
1999	Clinton	5.605	9.127	6.4%
2000	Clinton	5.628	9.708	3.6%
2001	Bush	5.769	10.060	3.2%
2002	Bush	6.198	10.378	4.1%
2003	Bush	6.760	10.804	6.5%
2004	Bush	7.354	11.504	6.4%
2005	Bush	7.905	12.235	6.3%
2006	Bush	8.451	13.010	4.9%
2007	Bush	8.951	13.642	4.3%
2008	Bush	9.654	14.222	5.7%
2009	Obama	10.413	15.027	5.1%
2010	Obama	13.954	15.792	3.0%

Note that both debt and GDP have increased every year in this particular time series, but it's the one most often used by the Tea Party to make their case.

Table 1b.  Debt as a Percentage of GDP Versus Subsequent 12-Month Change in GDP:

Budget Year:	Pres:	Debt as % of GDP:	dGDP+%
1977	Carter	35.4%	12.4%
1978	Carter	34.8%	12.8%
1979	Carter	33.0%	8.9%
1980	Carter	33.3%	12.3%
1981	Reagan	32.5%	5.5%
1982	Reagan	35.2%	6.7%
1983	Reagan	39.8%	11.7%
1984	Reagan	40.7%	7.9%
1985	Reagan	43.8%	6.2%
1986	Reagan	48.1%	5.6%
1987	Reagan	50.4%	7.7%
1988	Reagan	51.9%	7.8%
1989	Bush	53.1%	6.2%
1990	Bush	55.9%	3.4%
1991	Bush	60.6%	5.2%
1992	Bush	64.1%	5.4%
1993	Clinton	66.1%	5.9%
1994	Clinton	66.7%	5.2%
1995	Clinton	67.2%	5.1%
1996	Clinton	67.3%	6.4%
1997	Clinton	65.6%	5.4%
1998	Clinton	63.5%	5.8%
1999	Clinton	61.4%	6.4%
2000	Clinton	58.0%	3.6%
2001	Bush	57.3%	3.2%
2002	Bush	59.7%	4.1%
2003	Bush	62.6%	6.5%
2004	Bush	63.9%	6.4%
2005	Bush	64.6%	6.3%
2006	Bush	65.0%	4.9%
2007	Bush	65.6%	4.3%
2008	Bush	67.9%	5.7%
2009	Obama	69.3%	5.1%
2010	Obama	88.4%	3.0%

If debt were inversely correlated with GDP growth, you would expect to see this from a simple correlation of the two variables, as indeed you do:
The correlation coefficient between the two is -.612 indicating a modest inverse relationship but the R2, a measure of how well the data points fit a line, is a weak .375 (normal for noisy economic time series, however).

When debt ranged from $700 billion to $930 trillion, the average growth in GDP was 11.6% the next year*.  As debt increased, average growth in GDP dropped, with the exception of the 75th-90th percentile of debt (6.09-8.80 trillion) which was followed by a higher after GDP growth rate (5.6%) than the stratum below (5.1% for debt ranging from $4.50 to $6.09 trillion).  

When adjusted for size of the economy, the debt relationship was a bit stronger, with a -.697 correlation coefficient and a .485 R2 (1.0 would indicate a perfect fit with a straight line).

Table Two:  Starting Debt level versus subsequent 12-month change in GDP, 1977-2009:

Sliced data:			Debt
		n:	From:	To:	GDP growth:
	Min-10%	4	0.70	0.93	11.6%
	10-25%	5	0.93	1.89	7.6%
	25-50%	8	1.89	4.50	6.4%
	50-75%	8	4.50	6.09	5.1%
	75-90%	5	6.09	8.80	5.6%
	90-Max	4	8.80	13.95	4.5%
		34
In fourths:
		n:	From:	To:	GDP growth:
	Min-25%	9	0.70	1.89	9.4%
	25-50%	8	1.89	4.50	5.9%
	50-75%	8	4.50	6.09	5.1%
	75-Max	9	6.09	13.95	5.1%
		34
In thirds:
		n:	From:	To:	GDP Growth:
	Bottom	11	0.70	2.57	8.9%
	Middle	11	2.57	5.58	5.6%
	Top	12	5.58	13.95	4.9%
		34

The economy did seem to grow fastest when the debt was in the lowest quartile, but the strength of the correlation weakened as the debt increased.   In other words, average growth almost halved from the lowest quartile to the second highest quartile of debt (from 9.4% to 5.1%) but this decrease did not continue into the highest quartile which was roughly equal to the penultimate quartile and only marginally lower than the second quartile (5.1% v. 5.9%). 

Debt versus GDP year-over-year change following each debt observation 1977-2010 divided into strata by starting debt with subsequent average GDP growth.  While very low debt levels were followed by strong growth, the inverse trend flattened out with only slightly below average growth for above average debt.
 * the reason these GDP growth numbers are much higher than the headline GDP growth numbers reported is because they are nominal, using raw GDP, rather than inflation-adjusted (real) using GDP adjusted for inflation.

When debt is expressed as a percentage of GDP, the relationship becomes a bit clearer - lower levels of debt are followed by higher levels of GDP growth.  However, the highest levels of debt, those in the top decile, are followed by faster growth than those in the 75th-90th percentiles.  

Change in debt versus GDP Growth

Fair enough.  The absolute level of debt does seem to have some inverse correlation with subsequent economic growth, but what about changes in the debt level (or the deficit)?   Do markets respond better when the government is running a surplus than a deficit, and worse when the deficits are very large?  

Table Three:  Changes in Debt Year-Over-Year Versus Next 12 Months Change in GDP, 1977-2010:

Budget Year:	Pres:	dDebt%	dGDP+%
1978	Carter	10.4%	12.8%
1979	Carter	7.1%	8.9%
1980	Carter	9.8%	12.3%
1981	Reagan	9.5%	5.5%
1982	Reagan	14.4%	6.7%
1983	Reagan	20.6%	11.7%
1984	Reagan	14.1%	7.9%
1985	Reagan	16.2%	6.2%
1986	Reagan	16.7%	5.6%
1987	Reagan	10.6%	7.7%
1988	Reagan	10.9%	7.8%
1989	Bush	10.2%	6.2%
1990	Bush	11.8%	3.4%
1991	Bush	12.2%	5.2%
1992	Bush	11.2%	5.4%
1993	Clinton	8.7%	5.9%
1994	Clinton	6.7%	5.2%
1995	Clinton	6.0%	5.1%
1996	Clinton	5.3%	6.4%
1997	Clinton	3.6%	5.4%
1998	Clinton	2.0%	5.8%
1999	Clinton	2.3%	6.4%
2000	Clinton	0.4%	3.6%
2001	Bush	2.5%	3.2%
2002	Bush	7.4%	4.1%
2003	Bush	9.1%	6.5%
2004	Bush	8.8%	6.4%
2005	Bush	7.5%	6.3%
2006	Bush	6.9%	4.9%
2007	Bush	5.9%	4.3%
2008	Bush	7.9%	5.7%
2009	Obama	7.9%	5.1%
2010	Obama	34.0%	3.0%

The answer here seems to be a resounding Not Really.

First, the correlation coefficient is a positive .12, meaning that increases in debt are correlated, but very weakly, with growth in GDP.   The higher the percentage increase in debt, the higher the growth you can expect next year.  However, the R2 is .015, telling us the fit to a straight line is very poor.    At the very best, this tells us the Tea Party claim that increases in debt tend to be followed by poor economic growth is not true.  The fairest reading of these data is that there is no real correlation at all.  

It's quite clear that this relationship is neither linear nor inverse.   The economy performed weakly following periods of very low increases in debt, strongest when debt was increasing in the second quartile (50th-75th percentile) and actually performed second-strongest of all when debt was increasing maximally on a year-over-year basis (in the top decile).

If the Tea Party thesis were correct, then you would see tall bars on the left and shorter bars on the right.  You don't, so the Tea Party is wrong on this point. 

Now, this is not to argue that we should increase debt to goose the economy; there are many reasons not to increase debt and to balance your budget.  It just turns out that over the past 40 years, avoiding an economic slowdown has not been one of them. 

Does High Government Spending Lead to Slower Economic Growth? 

This is the next Tea Party thesis, that unless we shrink government spending, economic growth will be slow.  Since government spending relative to the economy has swung from a low of 18.4% under Clinton to a high of 23.5% under Reagan, it seems we might be able to answer this question. 

Table Four:  Government Spending as a % of GDP versus next 12 months change in GDP.

Budget Year:	Pres:	Expenses as a % of GDP:	dGDP+%
1977	Carter	20.7%	12.4%
1978	Carter	20.7%	12.8%
1979	Carter	20.1%	8.9%
1980	Carter	21.7%	12.3%
1981	Reagan	22.2%	5.5%
1982	Reagan	23.1%	6.7%
1983	Reagan	23.5%	11.7%
1984	Reagan	22.1%	7.9%
1985	Reagan	22.8%	6.2%
1986	Reagan	22.5%	5.6%
1987	Reagan	21.6%	7.7%
1988	Reagan	21.2%	7.8%
1989	Bush	21.2%	6.2%
1990	Bush	21.8%	3.4%
1991	Bush	22.3%	5.2%
1992	Bush	22.1%	5.4%
1993	Clinton	21.4%	5.9%
1994	Clinton	21.0%	5.2%
1995	Clinton	20.7%	5.1%
1996	Clinton	20.3%	6.4%
1997	Clinton	19.6%	5.4%
1998	Clinton	19.2%	5.8%
1999	Clinton	18.6%	6.4%
2000	Clinton	18.4%	3.6%
2001	Bush	18.5%	3.2%
2002	Bush	19.4%	4.1%
2003	Bush	20.0%	6.5%
2004	Bush	19.9%	6.4%
2005	Bush	20.2%	6.3%
2006	Bush	20.4%	4.9%
2007	Bush	20.0%	4.3%
2008	Bush	20.9%	5.7%
2009	Obama	20.7%	5.1%
2010	Obama	22.5%	3.0%

Once again, the Tea Party has got this relationship wrong.  Now it maybe that values extremely above or below the relatively narrow range of government spending we have seen in this time period may produce different results, or a different mix of government spending (more on education, research and development, less on defense, let's say) might yield different results, but if one simply looks at the aggregate size of government spending relative to the economy, government spending is positively correlated with changes in GDP:  the more government spending, the higher the economic growth the next year.   Now the relatively weak correlation coefficient of 0.25 and paltry .06 R2 means we should not base policy on this correlation, but if the Tea Party were correct, we would have a strong negative correlation coefficient and a much larger R2.

The linear equation from these data would be: 

  GDP Growth = .47 x government spending (as a % of GDP) + 3%

Again, given the very low R2, these data points do not fit a line, so it's dangerous even making such an equation, but if we had to, that is the best fit line we would make. 

Note that the highest level of government spending as a percentage of GDP led to the highest subsequent growth in GDP.  In this respect, the Keynesians are right and the Tea Party is wrong:  government spending, all things being equal, does seem to stimulate subsequent economic growth.

The sliced data tables follow:

Table Five:  Stratified Data showing various average GDP Growth Rates  for ranges of  government spending as a % of GDP:

Sliced data:			Government         Spending as a % of GDP:
		n:	From:	To:	dGDP:
	Min-10%	4	18.4%	19.2%	4.7%
	10-25%	5	19.2%	20.0%	5.3%
	25-50%	8	20.0%	20.8%	6.4%
	50-75%	8	20.8%	22.1%	6.8%
	75-90%	5	22.1%	22.5%	5.9%
	90-Max	4	22.5%	23.5%	6.9%
		34
In fourths:
		n:	From:	To:	dGDP:
	Min-25%	9	18.4%	20.0%	5.1%
	25-50%	8	20.0%	20.8%	7.7%
	50-75%	8	20.8%	22.1%	6.8%
	75-Max	9	22.1%	23.5%	6.4%
		34
In thirds:
		n:	From:	To:	dGDP:
	Bottom	11	18.4%	20.3%	5.5%
	Middle	11	20.3%	21.5%	7.0%
	Top	12	21.5%	23.5%	6.7%
		34

Do Tax Cuts Stimulate Economic Growth?

There are many ways to answer this question, but because of complexities in our tax code, the easiest way to address this question is to measure the individual income taxes collected relative to the size of the economy and see if the economy grows stronger when less is collected.   

Let's start though with all taxes collected, including individual income, corporate, estate, and excise taxes:

Table Six:  Taxes (Revenue) Collected as a Percentage of GDP versus subsequent GDP Growth:

Budget Year:	Pres:	Revenue as % of GDP:	dGDP+%
1977	Carter	18.0%	12.4%
1978	Carter	18.0%	12.8%
1979	Carter	18.5%	8.9%
1980	Carter	19.0%	12.3%
1981	Reagan	19.6%	5.5%
1982	Reagan	19.2%	6.7%
1983	Reagan	17.4%	11.7%
1984	Reagan	17.3%	7.9%
1985	Reagan	17.7%	6.2%
1986	Reagan	17.5%	5.6%
1987	Reagan	18.4%	7.7%
1988	Reagan	18.1%	7.8%
1989	Bush	18.3%	6.2%
1990	Bush	18.0%	3.4%
1991	Bush	17.8%	5.2%
1992	Bush	17.5%	5.4%
1993	Clinton	17.5%	5.9%
1994	Clinton	18.1%	5.2%
1995	Clinton	18.5%	5.1%
1996	Clinton	18.9%	6.4%
1997	Clinton	19.3%	5.4%
1998	Clinton	20.0%	5.8%
1999	Clinton	20.0%	6.4%
2000	Clinton	20.9%	3.6%
2001	Bush	19.8%	3.2%
2002	Bush	17.9%	4.1%
2003	Bush	16.5%	6.5%
2004	Bush	16.3%	6.4%
2005	Bush	17.6%	6.3%
2006	Bush	18.5%	4.9%
2007	Bush	18.8%	4.3%
2008	Bush	17.7%	5.7%
2009	Obama	14.0%	5.1%
2010	Obama	13.7%	3.0%

These numbers ranged from a high of 20.9% under Clinton's final year in office to a low of 13.7% in 2010 under President Obama.   How have they correlated with next year's growth in GDP?

Answer:  not at all.  The correlation coefficient is statistically zero (.04) indicating a small but positive correlation (higher tax collections leading to greater subsequent growth rates).  But the teeny .002 R2 indicates these numbers are all over the place, linear-speaking.  

Table Seven:  Revenues collected as a percentage of GDP versus next 12 Months Change in GDP.

Sliced data:
		n:	From:	To:	y:
	Min-10%	4	13.7%	16.7%	5.2%
	10-25%	5	16.7%	17.6%	7.3%
	25-50%	8	17.6%	18.0%	6.4%
	50-75%	8	18.0%	18.9%	6.2%
	75-90%	5	18.9%	19.7%	7.2%
	90-Max	3	19.7%	20.9%	5.1%
		33
In fourths:
		n:	From:	To:	y:
	Min-25%	9	13.7%	17.6%	6.4%
	25-50%	8	17.6%	18.0%	7.0%
	50-75%	8	18.0%	18.9%	6.2%
	75-Max	9	18.9%	20.9%	6.1%
		34
In thirds:
		n:	From:	To:	y:
	Bottom	11	13.7%	17.7%	6.4%
	Middle	11	17.7%	18.5%	6.9%
	Top	12	18.5%	20.9%	6.1%
		34

A picture is worth a thousand sound bites; if you can see a pattern in here, you're a better person than I am:

Extremely low or high tax collections are associated with weaker growth but beyond that, nothing much can be said.  The economy grows with all different rates of taxation, although admittedly the band of collection has been narrow.  Whenever total tax collection stays near 20%, taxes tend to get cut.

Do individual income tax collections inhibit economic growth?
Finally, I addressed what could be a quibble with the idea of looking at total government revenue; suppose the economy were very sensitive to changes in individual income tax collections?  If so, doing the same analysis as we did with the other variables should lead to an inverse correlation (as collections drop, the GDP grows faster).  As it turns out, more bad news for one of the Tea Party's pet theories:  the "burden" of individual income taxes holds back economic growth and lowering those taxes (which if done correctly should be reflected in lower government collection of income taxes) should lead to higher subsequent GDP growth.  It doesn't.

Table Eight:  Individual Income Tax Collections as a Percentage of GDP versus Subsequent Growth in GDP:

Budget Year:	Pres:	Individual Income taxes as % of GDP:	dGDP+%
1977	Carter	8.0%	12.4%
1978	Carter	8.2%	12.8%
1979	Carter	8.7%	8.9%
1980	Carter	9.0%	12.3%
1981	Reagan	9.3%	5.5%
1982	Reagan	9.2%	6.7%
1983	Reagan	8.4%	11.7%
1984	Reagan	7.8%	7.9%
1985	Reagan	8.1%	6.2%
1986	Reagan	7.9%	5.6%
1987	Reagan	8.4%	7.7%
1988	Reagan	8.0%	7.8%
1989	Bush	8.3%	6.2%
1990	Bush	8.1%	3.4%
1991	Bush	7.9%	5.2%
1992	Bush	7.6%	5.4%
1993	Clinton	7.7%	5.9%
1994	Clinton	7.8%	5.2%
1995	Clinton	8.1%	5.1%
1996	Clinton	8.5%	6.4%
1997	Clinton	9.0%	5.4%
1998	Clinton	9.6%	5.8%
1999	Clinton	9.6%	6.4%
2000	Clinton	10.3%	3.6%
2001	Bush	9.9%	3.2%
2002	Bush	8.3%	4.1%
2003	Bush	7.3%	6.5%
2004	Bush	7.0%	6.4%
2005	Bush	7.6%	6.3%
2006	Bush	8.0%	4.9%
2007	Bush	8.5%	4.3%
2008	Bush	8.1%	5.7%
2009	Obama	6.1%	5.1%
2010	Obama	5.7%	3.0%

The correlation between the two is very weak (.05) but positive, indicating there really is no relationship.  R2 is a paltry .002.    The stratified data show no clear overall trend, except for a counter-Tea Party one:  from the minimum to the top decile, as you increase individual income tax collection, you get a higher subsequent GDP growth until the very highest rates of individual income tax collections 9.5-10.3% of the economy, which had the lowest growth rates of all.  So perhaps the Tea Party is correct that there is a sweet spot beyond which taxation leads to slower rates of economic growth, but there was no tax rate in this range that led to negative growth and it is unclear what to make of this single slice, or whether higher rates of tax collection would have resumed the positive trend in economic growth.  

Table Nine:  Stratified data showing different ranges of individual income tax collection as a percentage of GDP versus subsequent 12-month change in GDP, 1977-2010:

Sliced data:
		n:	From:	To:	y:
	Min-10%	4	5.7%	7.4%	5.2%
	10-25%	5	7.4%	7.8%	6.1%
	25-50%	8	7.8%	8.1%	6.4%
	50-75%	8	8.1%	8.7%	7.1%
	75-90%	5	8.7%	9.5%	7.7%
	90-Max	3	9.5%	10.3%	5.1%
		33
In fourths:
		n:	From:	To:	y:
	Min-25%	9	5.7%	7.8%	5.7%
	25-50%	8	7.8%	8.1%	6.6%
	50-75%	8	8.1%	8.7%	7.1%
	75-Max	9	8.7%	10.3%	6.4%
		34
In thirds:
		n:	From:	To:	y:
	Bottom	11	5.7%	8.0%	5.7%
	Middle	11	8.0%	8.4%	7.3%
	Top	12	8.4%	10.3%	6.3%
		34

Individual income tax collections arranged by percentile bins versus average GDP growth in each bin.  Note that GDP growth INCREASES as individual income tax collections as a percentage of GDP increase until beyond the 90th percentile, when subsequent economic growth slows to about the same rate as at the lowest individual income tax collection rates.  

Summary
Although the Tea Party is correct that very high levels of debt are correlated, although not always, with slower rates of subsequent GDP growth, beyond the highest extremes, this relationship is weak and quickly flattens out.  Periods of very low debt have been correlated with stronger economic growth than periods of extremely high debt.

However, no other Tea Party talking points are supported by vigorous historical analysis.  

Changes in debt from year to year, the size of government spending, government tax collection, or individual income tax collection are either not correlated at all with subsequent growth in GDP or are slightly positively correlated.  There are many reasons to decrease the size of the government and its level of tax collection, but stimulating economic growth does not seem to be one of them, if the last 40 years are any guide.  And unfortunately for the Tea Party, these data are the only guide we have.