Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. trade deficit economic variable:
n 64
R2 0.98
Adjusted R2 0.98
SE 33.95
Term Coefficient 95% CI SE t statistic DF p
Intercept -227.5 -377.7 to -77.2 74.80 -3.04 50 0.0037
Population 0.9905 0.1843 to 1.7967 0.40138 2.47 50 0.0171
Unemployment 12.56 1.38 to 23.75 5.569 2.26 50 0.0285
Inflation 2.745 -0.866 to 6.355 1.7976 1.53 50 0.1331
GDP 0.1939 0.0017 to 0.3861 0.09570 2.03 50 0.0481
Debt 0.1539 0.0876 to 0.2201 0.03299 4.66 50 <0.0001
Tax Receipts 0.1222 -0.2344 to 0.4788 0.17755 0.69 50 0.4945
Gov Spending -0.1355 -0.5305 to 0.2596 0.19669 -0.69 50 0.4941
Budget 0.4444 0.0062 to 0.8826 0.21816 2.04 50 0.0469
Consumer Spending -1.338 -1.595 to -1.082 0.1278 -10.47 50 <0.0001
State Deficit -0.8165 -1.8917 to 0.2587 0.53530 -1.53 50 0.1335
State Social Payment 4.607 2.484 to 6.730 1.0572 4.36 50 <0.0001
Gov Social Benefits -1.488 -2.259 to -0.717 0.3838 -3.88 50 0.0003
Personal Income 0.6272 0.3585 to 0.8959 0.13378 4.69 50 <0.0001
The economic parameters used to model the U.S. trade deficit over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, federal government spending, the federal government budget levels, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the U.S. trade deficit (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Trade Deficit in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Trade Deficit in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Trade Deficit in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Trade Deficit in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. trade deficit?
Increased U.S. population, unemployment, inflation, GDP, federal debt, federal budget levels, state social payments, and personal income have negative effects on the trade deficit. Consumer spending and government social benefit payments have the biggest impact to lower our trade deficit. As people and the government spend less money on exports – the trade deficit will decline.
My Book: Is America Dying? (Amazon.com, Barnes and Noble)
Tuesday, June 28, 2011
Thursday, June 23, 2011
Federal Government Budget Model
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. federal government budget economic variable:
n 64
R2 1.00
Adjusted R2 1.00
SE 21.1
Term Coefficient 95% CI SE t statistic DF p
Intercept -4.498 -106.370 to 97.375 50.7193 -0.09 50 0.9297
Population 0.06869 -0.46284 to 0.60022 0.264634 0.26 50 0.7963
Unemployment -1.457 -8.760 to 5.845 3.6357 -0.40 50 0.6903
Inflation -0.1558 -2.4562 to 2.1447 1.14532 -0.14 50 0.8924
GDP -0.02328 -0.14766 to 0.10109 0.061921 -0.38 50 0.7085
Debt -0.001651 -0.051092 to 0.047789 0.0246149 -0.07 50 0.9468
Tax Receipts -0.4659 -0.6457 to -0.2862 0.08948 -5.21 50 <0.0001
Gov Spending 0.6586 0.4970 to 0.8203 0.08049 8.18 50 <0.0001
Trade Deficit 0.1724 0.0024 to 0.3425 0.08465 2.04 50 0.0469
Consumer Spending 0.2863 0.0124 to 0.5603 0.13639 2.10 50 0.0408
State Deficit 1.382 0.821 to 1.944 0.2796 4.94 50 <0.0001
State Social Payment -2.45 -3.84 to -1.06 0.692 -3.54 50 0.0009
Gov Social Benefits 0.5422 0.0167 to 1.0677 0.26165 2.07 50 0.0434
Personal Income -0.02417 -0.22488 to 0.17654 0.099928 -0.24 50 0.8099
The economic parameters used to model Government Budget levels over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, federal government spending, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Government Budget (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Government Budget in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Government Budget in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Government Budget in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Government Budget in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. federal government budget?
Not surprisingly, government spending, state deficits, government social benefits, the trade deficit, and consumer spending have the biggest impact on the budget size. The only thing that will significantly lower the federal budget size is if states increase their share of social payments. If tax receipts go up, the government can also control the federal budget from going further into debt.
n 64
R2 1.00
Adjusted R2 1.00
SE 21.1
Term Coefficient 95% CI SE t statistic DF p
Intercept -4.498 -106.370 to 97.375 50.7193 -0.09 50 0.9297
Population 0.06869 -0.46284 to 0.60022 0.264634 0.26 50 0.7963
Unemployment -1.457 -8.760 to 5.845 3.6357 -0.40 50 0.6903
Inflation -0.1558 -2.4562 to 2.1447 1.14532 -0.14 50 0.8924
GDP -0.02328 -0.14766 to 0.10109 0.061921 -0.38 50 0.7085
Debt -0.001651 -0.051092 to 0.047789 0.0246149 -0.07 50 0.9468
Tax Receipts -0.4659 -0.6457 to -0.2862 0.08948 -5.21 50 <0.0001
Gov Spending 0.6586 0.4970 to 0.8203 0.08049 8.18 50 <0.0001
Trade Deficit 0.1724 0.0024 to 0.3425 0.08465 2.04 50 0.0469
Consumer Spending 0.2863 0.0124 to 0.5603 0.13639 2.10 50 0.0408
State Deficit 1.382 0.821 to 1.944 0.2796 4.94 50 <0.0001
State Social Payment -2.45 -3.84 to -1.06 0.692 -3.54 50 0.0009
Gov Social Benefits 0.5422 0.0167 to 1.0677 0.26165 2.07 50 0.0434
Personal Income -0.02417 -0.22488 to 0.17654 0.099928 -0.24 50 0.8099
The economic parameters used to model Government Budget levels over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, federal government spending, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Government Budget (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Government Budget in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Government Budget in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Government Budget in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Government Budget in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. federal government budget?
Not surprisingly, government spending, state deficits, government social benefits, the trade deficit, and consumer spending have the biggest impact on the budget size. The only thing that will significantly lower the federal budget size is if states increase their share of social payments. If tax receipts go up, the government can also control the federal budget from going further into debt.
GDP Model
Here is another set of models I ran on economic indicators which support my earlier claims from previous posts.
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the Gross Domestic Product (GDP) variable:
n 64
R2 1.00
Adjusted R2 1.00
SE 48.2
Term Coefficient 95% CI SE t statistic DF p
Intercept 27.19 -205.03 to 259.42 115.619 0.24 50 0.8150
Population 0.3053 -0.9047 to 1.5152 0.60240 0.51 50 0.6146
Unemployment -9.103 -25.583 to 7.377 8.2047 -1.11 50 0.2725
Inflation -0.04267 -5.29017 to 5.20484 2.612572 -0.02 50 0.9870
Debt -0.1244 -0.2315 to -0.0173 0.05331 -2.33 50 0.0237
Tax Receipts -0.1041 -0.6123 to 0.4041 0.25301 -0.41 50 0.6825
Gov Spending 0.7386 0.2151 to 1.2620 0.26060 2.83 50 0.0066
Budget -0.1211 -0.7680 to 0.5258 0.32208 -0.38 50 0.7085
Trade Deficit 0.3913 0.0034 to 0.7793 0.19314 2.03 50 0.0481
Consumer Spending 0.9912 0.4034 to 1.5790 0.29265 3.39 50 0.0014
State Deficit 1.486 -0.019 to 2.990 0.7491 1.98 50 0.0528
State Social Payment 0.9698 -2.5630 to 4.5025 1.75886 0.55 50 0.5838
Gov Social Benefits -0.897 -2.120 to 0.326 0.6088 -1.47 50 0.1469
Personal Income 0.3691 -0.0767 to 0.8150 0.22198 1.66 50 0.1026
The economic parameters used to model GDP over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the federal government debt, federal government tax receipts, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of GDP (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (GDP in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (GDP in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (GDP in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (GDP in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on GDP?
Consumer spending has the biggest positive impact on GDP and economic growth. Government spending can increase GDP as well, but increased tax receipts have a negative effect. Unemployment, deficit spending (Debt), government expenditures for social benefits, and increased federal budgets all have a negative effect on GDP. An increased trade deficit also increases GDP – it means consumers and both the federal and state governments are spending more money to import products including oil (our biggest import). These results are not surprising and support conservative fiscal claims for decades.
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the Gross Domestic Product (GDP) variable:
n 64
R2 1.00
Adjusted R2 1.00
SE 48.2
Term Coefficient 95% CI SE t statistic DF p
Intercept 27.19 -205.03 to 259.42 115.619 0.24 50 0.8150
Population 0.3053 -0.9047 to 1.5152 0.60240 0.51 50 0.6146
Unemployment -9.103 -25.583 to 7.377 8.2047 -1.11 50 0.2725
Inflation -0.04267 -5.29017 to 5.20484 2.612572 -0.02 50 0.9870
Debt -0.1244 -0.2315 to -0.0173 0.05331 -2.33 50 0.0237
Tax Receipts -0.1041 -0.6123 to 0.4041 0.25301 -0.41 50 0.6825
Gov Spending 0.7386 0.2151 to 1.2620 0.26060 2.83 50 0.0066
Budget -0.1211 -0.7680 to 0.5258 0.32208 -0.38 50 0.7085
Trade Deficit 0.3913 0.0034 to 0.7793 0.19314 2.03 50 0.0481
Consumer Spending 0.9912 0.4034 to 1.5790 0.29265 3.39 50 0.0014
State Deficit 1.486 -0.019 to 2.990 0.7491 1.98 50 0.0528
State Social Payment 0.9698 -2.5630 to 4.5025 1.75886 0.55 50 0.5838
Gov Social Benefits -0.897 -2.120 to 0.326 0.6088 -1.47 50 0.1469
Personal Income 0.3691 -0.0767 to 0.8150 0.22198 1.66 50 0.1026
The economic parameters used to model GDP over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the federal government debt, federal government tax receipts, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of GDP (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (GDP in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (GDP in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (GDP in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (GDP in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on GDP?
Consumer spending has the biggest positive impact on GDP and economic growth. Government spending can increase GDP as well, but increased tax receipts have a negative effect. Unemployment, deficit spending (Debt), government expenditures for social benefits, and increased federal budgets all have a negative effect on GDP. An increased trade deficit also increases GDP – it means consumers and both the federal and state governments are spending more money to import products including oil (our biggest import). These results are not surprising and support conservative fiscal claims for decades.
Federal Debt Model
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. federal government Debt economic parameter:
n 64
R2 1.00
Adjusted R2 1.00
SE 121.5
Term Coefficient 95% CI SE t statistic DF p
Intercept 1376 941 to 1812 216.9 6.35 50 <0.0001
Population -3.83 -6.69 to -0.97 1.422 -2.69 50 0.0096
Unemployment -100.6 -131.4 to -69.8 15.33 -6.56 50 <0.0001
Inflation -26.64 -37.48 to -15.81 5.396 -4.94 50 <0.0001
GDP -0.7892 -1.4687 to -0.1097 0.33831 -2.33 50 0.0237
Tax Receipts 0.6127 -0.6578 to 1.8831 0.63251 0.97 50 0.3374
Gov Spending 1.05 -0.34 to 2.44 0.691 1.52 50 0.1353
Budget -0.0545 -1.6863 to 1.5773 0.81244 -0.07 50 0.9468
Trade Deficit 1.971 1.122 to 2.819 0.4225 4.66 50 <0.0001
Consumer Spending 1.742 0.176 to 3.307 0.7794 2.23 50 0.0300
State Deficit 2.745 -1.113 to 6.603 1.9209 1.43 50 0.1593
State Social Payment -0.6712 -9.5954 to 8.2530 4.44307 -0.15 50 0.8805
Gov Social Benefits 5.672 2.969 to 8.374 1.3455 4.22 50 0.0001
Personal Income -0.9377 -2.0603 to 0.1849 0.55891 -1.68 50 0.0996
The economic parameters used to model Debt over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government tax receipts, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Debt (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Debt in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Debt in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Debt in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Debt in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. federal government Debt?
The trade deficit has the biggest negative impact on our debt (importing oil) as does both federal and state government spending (in particular spending on government social benefits). Increased inflation and GDP can drive the federal government debt lower. The odd statistic is that increased unemployment tends to lower debt and increased tax receipts raise the debt. Actually, this is not that odd since the federal government tends to spend money than it receives regardless as to whether or not they raise more tax revenue. Thus, the federal debt continues to go up even when consumer spending is high
n 64
R2 1.00
Adjusted R2 1.00
SE 121.5
Term Coefficient 95% CI SE t statistic DF p
Intercept 1376 941 to 1812 216.9 6.35 50 <0.0001
Population -3.83 -6.69 to -0.97 1.422 -2.69 50 0.0096
Unemployment -100.6 -131.4 to -69.8 15.33 -6.56 50 <0.0001
Inflation -26.64 -37.48 to -15.81 5.396 -4.94 50 <0.0001
GDP -0.7892 -1.4687 to -0.1097 0.33831 -2.33 50 0.0237
Tax Receipts 0.6127 -0.6578 to 1.8831 0.63251 0.97 50 0.3374
Gov Spending 1.05 -0.34 to 2.44 0.691 1.52 50 0.1353
Budget -0.0545 -1.6863 to 1.5773 0.81244 -0.07 50 0.9468
Trade Deficit 1.971 1.122 to 2.819 0.4225 4.66 50 <0.0001
Consumer Spending 1.742 0.176 to 3.307 0.7794 2.23 50 0.0300
State Deficit 2.745 -1.113 to 6.603 1.9209 1.43 50 0.1593
State Social Payment -0.6712 -9.5954 to 8.2530 4.44307 -0.15 50 0.8805
Gov Social Benefits 5.672 2.969 to 8.374 1.3455 4.22 50 0.0001
Personal Income -0.9377 -2.0603 to 0.1849 0.55891 -1.68 50 0.0996
The economic parameters used to model Debt over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government tax receipts, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Debt (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Debt in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Debt in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Debt in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Debt in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on the U.S. federal government Debt?
The trade deficit has the biggest negative impact on our debt (importing oil) as does both federal and state government spending (in particular spending on government social benefits). Increased inflation and GDP can drive the federal government debt lower. The odd statistic is that increased unemployment tends to lower debt and increased tax receipts raise the debt. Actually, this is not that odd since the federal government tends to spend money than it receives regardless as to whether or not they raise more tax revenue. Thus, the federal debt continues to go up even when consumer spending is high
Federal Tax Receipts Model
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. federal government Tax Receipts economic variable:
n 64
R2 1.00
Adjusted R2 1.00
SE 26.91
Term Coefficient 95% CI SE t statistic DF p
Intercept -16.14 -145.72 to 113.43 64.513 -0.25 50 0.8034
Population -0.03525 -0.71208 to 0.64159 0.336976 -0.10 50 0.9171
Unemployment -1.122 -10.425 to 8.181 4.6316 -0.24 50 0.8095
Inflation 1.511 -1.386 to 4.407 1.4421 1.05 50 0.2998
GDP -0.03242 -0.19065 to 0.12582 0.078782 -0.41 50 0.6825
Debt 0.03006 -0.03228 to 0.09240 0.031038 0.97 50 0.3374
Gov Spending 0.5752 0.3063 to 0.8441 0.13389 4.30 50 <0.0001
Budget -0.7547 -1.0457 to -0.4636 0.14492 -5.21 50 <0.0001
Trade Deficit 0.07679 -0.14733 to 0.30092 0.111584 0.69 50 0.4945
Consumer Spending 0.008488 -0.355196 to 0.372171 0.1810668 0.05 50 0.9628
State Social Payment 0.2803 -1.6954 to 2.2560 0.98365 0.28 50 0.7768
Gov Social Benefits -0.776 -1.437 to -0.115 0.3292 -2.36 50 0.0224
Personal Income 0.442 0.219 to 0.665 0.1108 3.99 50 0.0002
State Deficit 2.557 2.075 to 3.039 0.2402 10.65 50 <0.0001
The economic parameters used to model Tax Receipts over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of Tax Receipts (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Tax Receipts in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Tax Receipts in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Tax Receipts in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Tax Receipts in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on U.S. federal government Tax Receipts?
As personal income, federal government spending, and state government debt increase so does the amount of tax receipts received by the federal government. This, once again, proves the federal government will spend more if it receives more revenue. The federal deficit continues to increase even as the government receives more tax revenue. Interestingly, the only thing that will help reduce federal government tax revenues is to reduce the budget, and in particular, government social benefit spending. Our current government is following the opposite strategy.
n 64
R2 1.00
Adjusted R2 1.00
SE 26.91
Term Coefficient 95% CI SE t statistic DF p
Intercept -16.14 -145.72 to 113.43 64.513 -0.25 50 0.8034
Population -0.03525 -0.71208 to 0.64159 0.336976 -0.10 50 0.9171
Unemployment -1.122 -10.425 to 8.181 4.6316 -0.24 50 0.8095
Inflation 1.511 -1.386 to 4.407 1.4421 1.05 50 0.2998
GDP -0.03242 -0.19065 to 0.12582 0.078782 -0.41 50 0.6825
Debt 0.03006 -0.03228 to 0.09240 0.031038 0.97 50 0.3374
Gov Spending 0.5752 0.3063 to 0.8441 0.13389 4.30 50 <0.0001
Budget -0.7547 -1.0457 to -0.4636 0.14492 -5.21 50 <0.0001
Trade Deficit 0.07679 -0.14733 to 0.30092 0.111584 0.69 50 0.4945
Consumer Spending 0.008488 -0.355196 to 0.372171 0.1810668 0.05 50 0.9628
State Social Payment 0.2803 -1.6954 to 2.2560 0.98365 0.28 50 0.7768
Gov Social Benefits -0.776 -1.437 to -0.115 0.3292 -2.36 50 0.0224
Personal Income 0.442 0.219 to 0.665 0.1108 3.99 50 0.0002
State Deficit 2.557 2.075 to 3.039 0.2402 10.65 50 <0.0001
The economic parameters used to model Tax Receipts over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government spending, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of Tax Receipts (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Tax Receipts in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Tax Receipts in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Tax Receipts in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Tax Receipts in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on U.S. federal government Tax Receipts?
As personal income, federal government spending, and state government debt increase so does the amount of tax receipts received by the federal government. This, once again, proves the federal government will spend more if it receives more revenue. The federal deficit continues to increase even as the government receives more tax revenue. Interestingly, the only thing that will help reduce federal government tax revenues is to reduce the budget, and in particular, government social benefit spending. Our current government is following the opposite strategy.
Federal Government Spending Model
Below are the results of running a linear regression model on various economic data (obtained from the Bureau of Economic Analysis [BEA] government site) from 1947 to the present solving for the U.S. federal government spending economic variable:
n 64
R2 1.00
Adjusted R2 1.00
SE 24.29
Term Coefficient 95% CI SE t statistic DF p
Intercept -61.46 -177.20 to 54.27 57.620 -1.07 50 0.2912
Population 0.05713 -0.55371 to 0.66796 0.304115 0.19 50 0.8518
Unemployment 6.904 -1.266 to 15.075 4.0679 1.70 50 0.0959
Inflation 0.537 -2.102 to 3.176 1.3138 0.41 50 0.6845
GDP 0.1874 0.0546 to 0.3202 0.06612 2.83 50 0.0066
Debt 0.04197 -0.01357 to 0.09751 0.027650 1.52 50 0.1353
Tax Receipts 0.4687 0.2496 to 0.6878 0.10910 4.30 50 <0.0001
Trade Deficit -0.06938 -0.27169 to 0.13294 0.100727 -0.69 50 0.4941
Consumer Spending -0.11 -0.44 to 0.22 0.163 -0.68 50 0.5020
State Deficit -1.749 -2.360 to -1.139 0.3039 -5.76 50 <0.0001
State Social Payment -0.1523 -1.9367 to 1.6321 0.88841 -0.17 50 0.8646
Gov Social Benefits 0.1649 -0.4625 to 0.7923 0.31236 0.53 50 0.5999
Personal Income -0.1914 -0.4156 to 0.0328 0.11163 -1.71 50 0.0927
Budget 0.8692 0.6559 to 1.0826 0.10623 8.18 50 <0.0001
The economic parameters used to model Government Spending over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Government Spending (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Government Spending in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Government Spending in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Government Spending in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Government Spending in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on U.S. federal government spending?
Tax Receipts! The more money the federal government receives the more money it squanders. Other factors such as increased GDP, federal debt, the federal budget, and unemployment play a big role in increased government spending. The only thing that significantly lowers federal government spending is if state governments’ are running higher deficits (spending more) and if personal income drops.
n 64
R2 1.00
Adjusted R2 1.00
SE 24.29
Term Coefficient 95% CI SE t statistic DF p
Intercept -61.46 -177.20 to 54.27 57.620 -1.07 50 0.2912
Population 0.05713 -0.55371 to 0.66796 0.304115 0.19 50 0.8518
Unemployment 6.904 -1.266 to 15.075 4.0679 1.70 50 0.0959
Inflation 0.537 -2.102 to 3.176 1.3138 0.41 50 0.6845
GDP 0.1874 0.0546 to 0.3202 0.06612 2.83 50 0.0066
Debt 0.04197 -0.01357 to 0.09751 0.027650 1.52 50 0.1353
Tax Receipts 0.4687 0.2496 to 0.6878 0.10910 4.30 50 <0.0001
Trade Deficit -0.06938 -0.27169 to 0.13294 0.100727 -0.69 50 0.4941
Consumer Spending -0.11 -0.44 to 0.22 0.163 -0.68 50 0.5020
State Deficit -1.749 -2.360 to -1.139 0.3039 -5.76 50 <0.0001
State Social Payment -0.1523 -1.9367 to 1.6321 0.88841 -0.17 50 0.8646
Gov Social Benefits 0.1649 -0.4625 to 0.7923 0.31236 0.53 50 0.5999
Personal Income -0.1914 -0.4156 to 0.0328 0.11163 -1.71 50 0.0927
Budget 0.8692 0.6559 to 1.0826 0.10623 8.18 50 <0.0001
The economic parameters used to model Government Spending over the past 64 (n) years are: the U.S. population, the unemployment rate, the inflation rate, the U.S. Gross Domestic Product (GDP), federal government debt, federal government tax receipts, the federal government budget size, the trade deficit, consumer spending, state government deficits, state government spending on social benefits, federal government spending on social benefits, and personal income. The intercept value in the above table is not a parameter – it is the value of the Government Spending (in billions of dollars) if all other parameters equal zero. These economic parameters are denoted in the above table.
The R² statistic illustrates how closely the linear regression model resembles a straight line (the ideal condition). If R² equals one then the model is 100% linear and the parameters correlate 100%. On the other hand, if R² is equal to zero then there is no correlation and the data in the linear regression model is completely random. T statistics reveal which of the economic parameters has the best correlation to the parameter being tested (Government Spending in this case). The higher the absolute value of the t statistic, the better the correlation the corresponding economic parameter has to the tested variable (Government Spending in this case). If a coefficient value of an economic parameter is positive then it trends in the same direction of the tested variable (Government Spending in this case). If a coefficient value is negative then the corresponding variable trends in the opposite direction of the tested variable (Government Spending in this case). It is time to do some math to prove higher taxes and government spending cripple economies. What economic parameters have the biggest effect on U.S. federal government spending?
Tax Receipts! The more money the federal government receives the more money it squanders. Other factors such as increased GDP, federal debt, the federal budget, and unemployment play a big role in increased government spending. The only thing that significantly lowers federal government spending is if state governments’ are running higher deficits (spending more) and if personal income drops.
What Caused the Financial/Housing Collapse?
Let’s do the math. Over 97% of U.S. mortgage debt has been handed out by commercial banks, savings institutions, credit unions, government sponsored programs, agency and government sponsored programs, asset backed programs, and financial institutions. We can model mortgages handed out through these programs/institutions as well as those mortgage backed securities (MBS - both private and government that were handed out by the before mentioned mortgage institutions and programs), that are believed to have caused the financial/housing collapse, against the results of the Dow Jones, S&P500, NYSE, NASDAQ, foreclosure rates, home ownership rate, home vacancy rate, and the mortgage payment delinquency rate. The data used in the model came from the Census Bureau, Market sites, and SIFMA statistical research company. Here are the results (I will draw conclusions at the end of each section. I will cover the market results in this post and the economic indicator results in Part II):
Dow Jones
R2 0.87
Adjusted R2 0.75
SE 1699.374
Term Coefficient 95% CI SE t statistic DF p
Intercept -9154 -23045 to 4737 6234.4 -1.47 10 0.1728
Commercial Banking 7.901 -39.748 to 55.550 21.3851 0.37 10 0.7195
Savings Institutions 8.797 -13.315 to 30.908 9.9237 0.89 10 0.3962
Credit Unions -100.2 -301.3 to 101.0 90.28 -1.11 10 0.2932
Government sponsored -2.618 -30.477 to 25.241 12.5034 -0.21 10 0.8383
Agency and Government Sponsored 9.434 -4.832 to 23.700 6.4027 1.47 10 0.1714
Asset Backed -3.34 -20.37 to 13.69 7.644 -0.44 10 0.6714
Finance Corporation -24.58 -71.01 to 21.85 20.838 -1.18 10 0.2655
Non Agency MBS (B) 11.34 -5.62 to 28.30 7.613 1.49 10 0.1673
Total U.S. Gov MBS (B) -3 -9 to 3 2.7 -1.10 10 0.2956
S&P 500
R2 0.79
Adjusted R2 0.60
SE 243.213
Term Coefficient 95% CI SE t statistic DF p
Intercept -837.7 -2825.8 to 1150.4 892.26 -0.94 10 0.3699
Commercial Banking 0.7569 -6.0626 to 7.5764 3.06061 0.25 10 0.8097
Savings Institutions 0.8265 -2.3381 to 3.9911 1.42027 0.58 10 0.5735
Credit Unions -7.911 -36.701 to 20.879 12.9210 -0.61 10 0.5540
Government sponsored -0.5208 -4.5080 to 3.4664 1.78947 -0.29 10 0.7770
Agency and Government Sponsored 1.087 -0.955 to 3.128 0.9164 1.19 10 0.2631
Asset Backed -0.68 -3.12 to 1.76 1.094 -0.62 10 0.5482
Finance Corporation -3.472 -10.117 to 3.173 2.9823 -1.16 10 0.2713
Non Agency MBS (B) 1.876 -0.551 to 4.304 1.0896 1.72 10 0.1158
Total U.S. Gov MBS (B) -0.4876 -1.3542 to 0.3791 0.38897 -1.25 10 0.2385
NYSE
R2 0.89
Adjusted R2 0.80
SE 1041.118
Term Coefficient 95% CI SE t statistic DF p
Intercept -3636 -12146 to 4875 3819.5 -0.95 10 0.3636
Commercial Banking 4.888 -24.304 to 34.080 13.1015 0.37 10 0.7169
Savings Institutions 3.411 -10.136 to 16.957 6.0797 0.56 10 0.5872
Credit Unions -37.9 -161.1 to 85.3 55.31 -0.69 10 0.5088
Government sponsored -1.798 -18.866 to 15.270 7.6602 -0.23 10 0.8191
Agency and Government Sponsored 4.823 -3.917 to 13.563 3.9226 1.23 10 0.2470
Asset Backed -2.759 -13.193 to 7.675 4.6830 -0.59 10 0.5688
Finance Corporation -18.36 -46.80 to 10.09 12.766 -1.44 10 0.1810
Non Agency MBS (B) 9.863 -0.529 to 20.255 4.6640 2.11 10 0.0606
Total U.S. Gov MBS (B) -2.265 -5.975 to 1.445 1.6650 -1.36 10 0.2036
NASDAQ
R2 0.59
Adjusted R2 0.22
SE 792.941
Term Coefficient 95% CI SE t statistic DF p
Intercept -1534 -8015 to 4948 2909.0 -0.53 10 0.6096
Commercial Banking 0.6868 -21.5465 to 22.9202 9.97844 0.07 10 0.9465
Savings Institutions 0.9137 -9.4037 to 11.2310 4.63046 0.20 10 0.8475
Credit Unions -7.779 -101.642 to 86.084 42.1261 -0.18 10 0.8572
Government sponsored -1.026 -14.025 to 11.973 5.8342 -0.18 10 0.8639
Agency and Government Sponsored 2.488 -4.169 to 9.145 2.9876 0.83 10 0.4244
Asset Backed -1.98 -9.93 to 5.97 3.567 -0.56 10 0.5910
Finance Corporation -8.576 -30.240 to 13.088 9.7230 -0.88 10 0.3985
Non Agency MBS (B) 4.971 -2.944 to 12.886 3.5522 1.40 10 0.1920
Total U.S. Gov MBS (B) -1.314 -4.139 to 1.512 1.2681 -1.04 10 0.3246
The mortgage model results when compared against the markets are mixed. Although the four models are not super strong in predicting outcomes (NYSE best and NASDAQ is the worst – by R² data), the models show that there is an inverse relationship between Non Agency MBS and Total U.S. Gov. MSB. Non Agency MBS are those toxic Mortgage Backed Securities (MBS) assets issued by private sector companies and Total U.S. Gov. MBS are those toxic mortgage backed securities issued by government entities Fannie Mae, Ginnie Mae, and Freddie Mac. Non Agency MBS work to drive up market values whereas; Total U.S. Gov. MBS work to drive market values lower. It is not all bad news for the government since agency and government sponsored mortgages drive up market values, but government sponsored mortgages drive down the market. I am not sure of the difference between these two, but I suspect they include VA, FHA, and GSE mortgage loans. Unfortunately, I am not sure how this data is divided between the two programs. Commercial banks and savings institution mortgages tend to drive the market up; whereas credit union and finance corporation mortgages drive the market value down. However, it is important to note that the correlation is strongest between Non Agency MBS and Total U.S. Gov. MBS with the markets than the other variables (based on higher t statistic and lower p results).
Here are the model results for the economic indicators when modeled against commercial banks, saving institutions, credit unions, government sponsored, agency and government sponsored, asset backed, finance corporations, non-agency MBS, and total U.S. government MBS mortgages.
Foreclosure Rate
R2 0.99
Adjusted R2 0.98
SE 0.087
Term Coefficient 95% CI SE t statistic DF p
Intercept 0.8404 -0.4088 to 2.0896 0.55222 1.52 9 0.1624
Commercial Banking -0.001989 -0.004591 to 0.000613 0.0011501 -1.73 9 0.1178
Savings Institutions 0.0006746 -0.0014666 to 0.0028159 0.00094655 0.71 9 0.4941
Credit Unions 0.009413 -0.008600 to 0.027427 0.0079630 1.18 9 0.2674
Government sponsored -0.0006434 -0.0022507 to 0.0009639 0.00071051 -0.91 9 0.3888
Agency and Government Sponsored 0.0003073 -0.0006934 to 0.0013079 0.00044233 0.69 9 0.5048
Asset Backed 0.001188 -0.000392 to 0.002768 0.0006984 1.70 9 0.1231
Finance Corporation -0.0007006 -0.0031269 to 0.0017257 0.00107256 -0.65 9 0.5299
Non Agency MBS (B) -0.001926 -0.003387 to -0.000465 0.0006458 -2.98 9 0.0154
Total U.S. Gov MBS (B) 0.0004141 0.0000832 to 0.0007451 0.00014629 2.83 9 0.0197
Home Ownership Rate
R2 0.96
Adjusted R2 0.93
SE 0.5043
Term Coefficient 95% CI SE t statistic DF p
Intercept 57.73 53.61 to 61.85 1.850 31.20 10 <0.0001
Commercial Banking 0.000874 -0.013267 to 0.015015 0.0063465 0.14 10 0.8932
Savings Institutions 0.005253 -0.001309 to 0.011815 0.0029451 1.78 10 0.1048
Credit Unions -0.04506 -0.10476 to 0.01464 0.026793 -1.68 10 0.1235
Government sponsored 0.0001896 -0.0080783 to 0.0084576 0.00371069 0.05 10 0.9602
Agency and Government Sponsored 0.004247 0.000013 to 0.008480 0.0019002 2.23 10 0.0494
Asset Backed -0.001434 -0.006488 to 0.003621 0.0022685 -0.63 10 0.5416
Finance Corporation 0.004906 -0.008873 to 0.018685 0.0061841 0.79 10 0.4460
Non Agency MBS (B) 0.001022 -0.004012 to 0.006056 0.0022593 0.45 10 0.6606
Total U.S. Gov MBS (B) -0.001057 -0.002855 to 0.000740 0.0008066 -1.31 10 0.2192
Home Vacancy Rate
R2 0.98
Adjusted R2 0.96
SE 0.0846
Term Coefficient 95% CI SE t statistic DF p
Intercept 1.161 0.470 to 1.853 0.3103 3.74 10 0.0038
Commercial Banking 0.001398 -0.000973 to 0.003770 0.0010643 1.31 10 0.2183
Savings Institutions 0.00136 0.00026 to 0.00246 0.000494 2.75 10 0.0204
Credit Unions 0.001868 -0.008143 to 0.011879 0.0044930 0.42 10 0.6863
Government sponsored -0.001912 -0.003298 to -0.000525 0.0006223 -3.07 10 0.0118
Agency and Government Sponsored -0.0006303 -0.0013403 to 0.0000797 0.00031864 -1.98 10 0.0761
Asset Backed 0.0006792 -0.0001684 to 0.0015268 0.00038041 1.79 10 0.1045
Finance Corporation -0.002279 -0.004590 to 0.000032 0.0010370 -2.20 10 0.0527
Non Agency MBS (B) -0.0009889 -0.0018330 to -0.0001447 0.00037887 -2.61 10 0.0260
Total U.S. Gov MBS (B) 0.0005488 0.0002475 to 0.0008502 0.00013525 4.06 10 0.0023
Mortgage Delinquency Rate
R2 0.93
Adjusted R2 0.87
SE 0.239
Term Coefficient 95% CI SE t statistic DF p
Intercept 3.886 0.459 to 7.313 1.5147 2.57 9 0.0304
Commercial Banking -0.003408 -0.010544 to 0.003729 0.0031548 -1.08 9 0.3082
Savings Institutions 0.003294 -0.002580 to 0.009167 0.0025964 1.27 9 0.2364
Credit Unions 0.007254 -0.042158 to 0.056665 0.0218426 0.33 9 0.7474
Government sponsored -0.002348 -0.006757 to 0.002061 0.0019489 -1.20 9 0.2590
Agency and Government Sponsored 6.1360E-05 -2.6834E-03 to 2.8061E-03 1.2133E-003 0.05 9 0.9608
Asset Backed 0.002569 -0.001765 to 0.006903 0.0019158 1.34 9 0.2128
Finance Corporation 0.002907 -0.003748 to 0.009563 0.0029420 0.99 9 0.3489
Non Agency MBS (B) -0.004429 -0.008436 to -0.000421 0.0017715 -2.50 9 0.0339
Total U.S. Gov MBS (B) 0.001141 0.000233 to 0.002049 0.0004013 2.84 9 0.0193
These are strong predictions models with high R² values and they are consistent with the results in Part I of this blog. The strongest correlation exists between Non Agency MBS and Total U.S. Gov. MBS with the economic indicators. While Non Agency MBS works to drive down foreclosure rates, home vacancy rates, delinquency rates, and increase home ownership rates; Total U.S. Gov. MBS works to do the opposite. This would lead me to believe that toxic mortgage assets from the government were the prime culprit in the meltdown and not so much those toxic assets from the private sector. In these economic indicator models savings institutions and credit union mortgages had a positive effect on economic variables; whereas commercial banking and finance corporations were mostly negative. The results for government sponsored and agency and government sponsored mortgages were mixed (in some cases improved and in other cases decreased results) on the economic indicators.
Subscribe to:
Posts (Atom)