1. The Vulnerability of the United States Economy to Electricity Price Increases Presentation to the Kentucky Association of Manufacturers May 11 th, 2016 Aron Patrick Assistant Director Kentucky Energy & Environment Cabinet Department for Energy Development and Independence energy.ky.gov 502-564-7192 Download this Presentation: http://1.usa.gov/23sDw8G Download the Associated White Paper: http://1.usa.gov/1aVhOWu

2. Kentucky’s Electricity Portfolio 2

3. 40 percent of the electricity consumed in Kentucky goes to industrial and manufacturing processes. 3

4. Kentucky electricity consumption has always been primarily for manufacturing. Kentucky industrial demand has declined since 2009. 4

5. We project that industrial demand will continue to decline as industry becomes more efficient. ← Historical Forecast → 5

6. Kentucky has always been one of the most-manufacturing-intensive electricity portfolios in the United States. 6

7. Since 1990, Kentucky has maintained one of the lowest electricity prices in the United States. In 2015, Kentucky industrial electricity prices were 4 th lowest in the United States and the lowest east of the Mississippi. In 2015, Kentucky electricity prices decreased in both real and nominal terms due to low coal prices. 7

8. Kentucky electricity prices have remained low and stable due primarily to low coal prices. 8

9. United States Electricity Prices <6.85 ¢ 6.85¢ ↔ 7.9¢ 7.9¢ ↔ 8.7¢ 8.7¢ ↔ 9.5¢ 9.5¢ ↔ 10.25¢ 10.3¢ ↔ 11.0¢ 11.0¢ ↔ 12.0¢ 12.0¢ ↔ 14.0¢ 14.0¢ ↔ 25.0¢ 25.0¢ ↔ 55.0¢ Cents per kWh in 2012 Animated versions of these maps are online at : https://www.youtube.com/watch?v=QExe0zRFzfE 9

10. In 2015, 87 percent of Kentucky’s electricity was coal-fired and 7 percent gas fired and 3 percent hydro. 10

11. 11

12. 12

13. ← Historical Forecast → Renewables Natural Gas 13

14. Kentucky power plants reduced sulfur dioxide emissions by 36% in 2015. Kentucky has reduced electric-power emissions by 76% since the year 2000 and emissions in 2016 are on track to set another all-time-record low. 14

15. In 2015, Kentucky power plants decreased carbon dioxide emissions by 11% to 86 million tons. 15

16. Electricity & Manufacturing 16

17. Kentucky Electricity Portfolio & Employment Other Employment that Depends on Electricity ≥1,952,624 Full Time Jobs Electricity Generation ≤15,000 Direct Jobs Employment Depends Upon Electricity – Direct employment for electricity generation, such as power plant operators and coal miners, is a relatively small portion of total employment in Kentucky compared to the millions of jobs that depend upon the reliable and inexpensive electricity they produce. Electricity- dependent jobs can be modeled with electricity price elasticity of employment coefficients. Manufacturing ≥234,983 Full Time Jobs 17

18. 18 Intensity Rank Manufacturing Sector Kilowatt-hours per $ of Product Electricity Cost per $ of Product Percentage of Production Costs United States Workers 1Primary Aluminum Production 12.9$0.57357.3%8,652 2Iron & Steel Mills 1.65$0.0868.6%100,719 5Glass Manufacturing 0.81$0.0555.5%82,661 24Motor Vehicle Parts 0.29$0.0212.1%101,715 52Motor Vehicle Assembly 0.12$0.0080.8%166,937 Complete List of Manufacturing Sectors on Page of 4 of the White Paper The Vulnerability of Kentucky's Manufacturing Economy to Increasing Electricity Prices: http://1.usa.gov/1HGuHDN http://1.usa.gov/1HGuHDN Embodied Electricity Increasing the costs of the embodied energy of a product—the energy required to convert the raw materials into a usable product—has multiplier effects throughout the economy. Whenever the production cost of primary material inputs increases, for whatever reason, the production cost of higher value products further down the commodity chain will increase as well.

19. 19 Electricity-intensity varies substantially by industry. At 3-Digit NAICS code level Paper, Metals, Textile, Minerals, Wood, and Chemical are the most intensive manufacturing sectors.

20. 20 The variability in intensity between sectors becomes more apparent, at 5-digit NAICS code level granularity.

21. 21

22. 22

23. 23

24. 24 The United States economy will continue to grow in spite of changing electricity costs and most sectors of the economy exhibit relatively inelastic responsiveness to electricity prices; however, a 10% increase in the real price of electricity nationally, the United States will decrease annual Gross Domestic Product (GDP) by $142 billion.

25. 25 A 10% increase in the real price of electricity nationally will decrease annual employment by 1,010,740 full time jobs across the United States.

26. Regional Disparity 26

28. 28 Kentucky’s economy is one of the most-electricity-intensive in the United States.

29. Primary Metal Manufacturing 29

30. 30 Century Aluminum in Sebree, Kentucky in 2015 30

31. 31 Century Aluminum in Sebree, Kentucky in 2015

32. North American Stainless in Carroll County, Kentucky March, 2015 32

33. 33

34. 34 Primary metal manufacturers, who use 0.56 kWh per dollar of product, and are the most responsive to electricity price increases in terms of value added, we estimate a reduction of $5.5 billion and 14,000 employees from 2012 levels in response to a 10% increase in electricity prices, ceteris paribus. Fully 41% of primary metal employment and 43% of value added is in Indiana, Ohio, Pennsylvania, Michigan, Illinois, and Wisconsin.

35. Textile Manufacturing 35

36. 36

37. 37 Textile manufacturing uses 0.45 kWh per dollar of product and is the most responsive to electricity price changes in terms of employment. We estimate a 10% increase in electricity prices would induce a reduction of 16,000 textile jobs, and a reduction of $580 million in value added from textiles, ceteris paribus. The sector clusters in the interior Piedmont region of Georgia, North Carolina, South Carolina, and Virginia, where 54% of the sectoral employment and 46% of sectoral GDP is located.

38. Chemical Manufacturing 38

39. 39

40. 40 Chemical manufacturers use 0.292 kWh per dollar of product. We estimate that a 10% increase in electricity prices would reduce value added by $14.2 billion annually.

41. Paper Manufacturing 41

42. 42

43. 43 Paper manufacturers use 0.646 kWh per dollar of product. We estimate that a 10% increase in electricity prices would reduce employment by 20,561 full time jobs across the United States and reduce annual GDP by $4.1 billion.

44. Conclusions 44

45. Conclusions Manufacturing is Kentucky’s largest source of Gross Domestic Product (GDP) at $36.2 billion and a leading source of employment at 234,983 in 2014. Manufacturing employs 12.2 million people across the United States with 1.5 million working in electricity-intensive sectors in 2014. Low and stable electricity prices are important factors of economic growth for manufacturing-intensive economies, particularly for electricity-intensive production. We estimate that a hypothetical 10% increase in the real price of electricity nationally, for whatever reason, would be associated with a decrease of one million jobs and $142 billion in Gross Domestic Product (GDP). Impacts will vary substantially by industry, with electricity-intensive manufacturing experiencing the most-significant losses of productive capacity. While most economic sectors are relatively unaffected by changes in electricity prices, the United States Environmental Protection Agency (EPA) should incorporate economy-wide modeling of electricity price elasticity effects that vary by sector when evaluating the costs of proposed environmental regulations. 45

46. Questions? Download this Presentation: http://1.usa.gov/23sDw8G Download the Associated White Paper: http://1.usa.gov/1aVhOWu Download the Excel VBA Interactive Version of this Model at: http://1.usa.gov/1G31BuN This presentation is based on a 2015 whitepaper by Aron Patrick, Adam Blandford, and Leonard K. Peters Kentucky Energy & Environment Cabinet Department for Energy Development and Independence energy.ky.gov 502-564-7192 46

47. Statistical Appendix 47

48. Electricity Intensity of United States Manufacturing, 2010 48

49. Analytical Methodology A multiple regression of panel data model using fixed effects was estimated for employment and value added by economic sector for a total of 24 models. Fixed effects were used to control for the numerous factors inherently affecting sector-specific employment as well as electricity prices from state to state that have not been accounted for in the independent variables included in this study. The multiple regression of panel data model with fixed effects can be generally given by, Where i and t index states and years, such that y it is the dependent variable of interest, employment by industry, in state i in year t, β 0 is the constant y intercept across all states, X is a k by 1 vector of explanatory variables, β j X jit is the product of the observation for each independent variable j through k for state i in year t and the coefficient of X, k is the total number of included independent variables, α i is the time-invariant fixed effect for state i, and ε it are the residuals, and where ε it ~ N(0, σ 2 ), or are approximately normally distributed with a mean of zero. 49

50. Analytical Methodology A multiple regression of panel data model using fixed effects was estimated for employment and value added by economic sector for a total of 24 models. Fixed effects were used to control for the numerous factors inherently affecting sector-specific employment as well as electricity prices from state to state that have not been accounted for in the independent variables included in this study. The multiple regression of panel data model with fixed effects can be generally given by, Where i and t index states and years, such that y it is the dependent variable of interest, employment by industry, in state i in year t, β 0 is the constant y intercept across all states, X is a k by 1 vector of explanatory variables, β j X jit is the product of the observation for each independent variable j through k for state i in year t and the coefficient of X, k is the total number of included independent variables, α i is the time-invariant fixed effect for state i, and ε it are the residuals, and where ε it ~ N(0, σ 2 ), or are approximately normally distributed with a mean of zero. 50 Predicted Employment by Industry in State (i) and Time (t) Constant Intercept Across States Stochastic Error Term (~ N(0, σ 2 )) Fixed Effect for State (i) Sum of the product of the observation for each independent variable j through k for state i in year t and the coefficient of Xj

51. Logged Variables ManufacturingServices MetalsPaperWoodChemicalTextilesMineralsNon-IntenseConstructTransportOtherGovernmentUtilities V.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmpV.AddEmp Electricity Price β1β1 -0.845-0.347-0.769-0.540-0.635-0.569-0.385-0.072-0.359-0.680-0.318-0.106-0.079-0.126-0.328-0.230-0.193-0.273-0.098-0.065-0.014-0.0070.5380.356 SE 0.0890.0640.0910.0640.0590.0410.0620.0510.1240.1140.0390.0310.0450.0340.0250.0230.0220.0200.0150.0080.0100.0080.0280.022 P>|t| 0.000 0.1540.0040.000 0.0010.0790.000 0.1590.3300.000 Personal Income β2β2 0.427-0.2540.078-1.1600.042-0.0421.2840.408-0.870-1.2461.8380.772-0.058-0.7752.5821.5770.7620.0811.2690.2200.6380.0590.093-0.004 SE 0.2190.1570.2230.1570.1460.1000.1520.1240.3050.2800.0960.0760.1100.0840.0600.0570.0540.0500.0370.0190.0240.0190.0690.053 P>|t| 0.0520.1060.7270.0000.7740.6770.0000.0010.0040.000 0.6010.000 0.1070.000 0.0010.1780.933 Education β3β3 1.5890.8392.5861.4593.3922.4490.2380.646-1.657-0.4130.1990.3050.5041.077-0.0370.1010.2780.6070.3140.618-0.1740.1470.7560.411 SE 0.1950.1400.1980.1400.1300.0890.1360.1110.2710.2500.0850.0680.0980.0750.0540.0510.0480.0450.0330.0170.0220.0170.0610.047 P>|t| 0.000 0.0790.000 0.0980.0190.000 0.4960.0490.000 Population β4β4 0.8630.7632.1362.1900.7040.3010.5090.6961.4172.0031.5111.4031.6871.4441.3871.2901.1591.2351.1180.9400.6500.6650.9890.820 SE 0.1150.0820.1170.0830.0770.0530.0800.0650.1600.1470.0500.0400.0580.0440.0320.0300.0290.0260.0190.0100.0130.0100.0360.028 P>|t| 0.000 0.000.000 Petroleum Price β5β5 0.399-0.101-0.240-0.223-0.226-0.1920.104-0.059-0.368-0.1140.033-0.045-0.112-0.1910.105-0.0030.058-0.0980.061-0.068-0.019-0.041-0.248-0.129 SE 0.0500.0360.0510.0360.0330.0230.0350.0280.0690.0640.0220.0170.0250.0190.0140.0130.0120.0110.0080.0040.0060.0040.0160.012 P>|t| 0.0000.0050.000 0.0030.0370.0000.0740.1240.0090.000 0.8250.000 0.0010.000 Year β6β6 -144.2-81.3-161.2-85.3-193.8-134.9-6.3-48.987.413.0-96.0-78.1-28.6-60.9-68.8-46.7-29.4-24.2-14.5-3.722.60.3-7.8-12.0 SE 10.767.6910.927.707.164.917.476.1014.9513.744.703.725.414.122.962.812.672.471.810.931.200.913.382.61 P>|t| 0.00 0.400.00 0.340.00 0.740.020.00 Constant ( β 0 ) β0β0 108061511926331459102434364-665-1056945722014584843302061749024-1784986 SE 79.656.980.957.053.036.355.345.1110.6101.734.827.540.030.521.920.819.818.213.46.98.96.725.019.3 P>|t| 0.00 0.540.00 0.300.00 0.930.050.00 N21922193 21922193 R 2 Within0.1450.2620.1910.3050.3940.4410.5040.0720.1280.1340.3650.4510.4350.4220.9100.7610.7510.7870.9510.9790.9610.9150.7110.549 R 2 Between0.5780.6530.5830.6450.1210.0190.6420.6920.5200.5210.8360.8390.8970.8690.9570.9510.8530.9460.9830.9860.8830.9280.9040.905 R 2 Overall0.5400.6210.5500.6260.1400.0340.5560.6690.4720.4810.8120.8240.8660.8520.9320.9300.8450.9340.9760.9860.8810.9260.8870.891 Complete Model of Employment & Value Added by Sector