Power Generation, Transmission, and Use

Markets, Regulation, and Oversight

Impacts of Power Generation and Transmission

Looking Ahead


CEIR Report Map


Maryland Power Plants and the Environment (CEIR-18)

Appendix C – Determinants of Electricity Demand Growth in Maryland


This appendix provides an overview of the basic theoretical foundations upon which forecasts of electricity consumption and peak demand rest, and an analysis of the trends of the key economic and non-economic determinants of the demand for electricity. The Maryland data presented herein were obtained from the Maryland Department of Planning, the Bureau of Economic Analysis of the U.S. Department of Commerce, and the Bureau of Labor Statistics of the U.S. Department of Labor. Economic variables include income, price of electricity, and employment; non-economic variables include population (which is itself influenced by income and employment) and weather. Historical information is required for estimation purposes, while projected data are necessary to forecast the demand for power using the statistical relationships between these variables and electricity consumption determined during the estimation process based on the historical data.

This appendix is composed of five sections. The following section presents a brief discussion of the theoretical foundations used for modeling the demand for electricity econometrically. This section sets the stage for the rest of Appendix C, which examines economic and demographic trends for Maryland by region. For purposes of presentation, the State has been divided into six regions, as shown in Table C-1. The section covering the theoretical foundations is followed by a section discussing trends in per capita income, which, in turn, is followed by a section discussing trends in employment. Trends in population and the number of households follow the employment section. The final section of Appendix C presents a brief summary.

Table C-1 Principal Regions in Maryland

Region Counties Predominant Electric Distribution Utility
Baltimore Anne Arundel
Baltimore City
Baltimore Gas and Electric Company
Washington Suburban Montgomery
Prince George’s
Potomac Electric Power Company
Southern Maryland Calvert
St. Mary’s
Southern Maryland Electric Cooperative
Western Maryland Allegany
Potomac Edison Company
Upper Eastern Shore Caroline
Queen Anne’s
Delmarva Power and Choptank Electric
Lower Eastern Shore Dorchester
Delmarva Power and Choptank Electric

Theoretical Foundations for Econometrically Modeling Electricity Demand

”Econometric” forecast studies use the economic theory of demand as the organizing principle to model the demand for electricity. The total demand for any good or service, including electricity, is simply the sum of the demands of the individual consumers in the market. The portion of market demand for residential use of electricity is driven by factors to which individual residential consumers are sensitive. Similarly, for the commercial and industrial sectors of the market demand for electricity, the factors affecting demand are those to which producers are sensitive.

The residential demand for electricity is assumed to result from the exercise of choice by which the consumer maximizes his or her welfare subject to a budget constraint. Consumer demand for electricity is taken to be a function of its price, consumer income, weather, and the price of related commodities (i.e., substitutes and complements such as natural gas for home heating). It is important to note that electricity, in and of itself, conveys no benefits to the consumer. Rather, the consumer benefits from the services of the stock of appliances that require electricity. These services include space conditioning, refrigeration, cooking, clothes washing and drying, and numerous other services and functions. Consequently, the demand for electricity can be appropriately viewed as a derived demand; that is, it results from the demand for the services provided by electricity-consuming appliances.

For commercial and industrial customers, electricity is a factor of production, i.e., an input. For the profit-maximizing producer, demand for a commodity (including electricity) is driven by its price, the price of related inputs, and the level of output. Producer demand for electricity is also driven by other factors, including weather.

Both the residential and non-residential demand for electric power are discussed above in terms of the individual consumer or producer. The market demand for electric power, for example, in Maryland or within regions in Maryland, is also dependent on the number of consumers (households) and the level of goods and services produced in the region. Because no satisfactory time series of output data is available at a suitably disaggregated level, we use employment as a proxy for output. Commercial and industrial electric sales are projected per employee, which is then multiplied by the number of forecasted employees to project total commercial and industrial demand for electricity. 

The growth in electricity use has historically been linked to the level of economic growth. The rate of growth of electricity use nationwide exceeded the rate of increase in gross domestic product (GDP) in the 1950’s by 5 percent. As shown in Figure C-1, the differential between the growth in real GDP and the growth in electric use has declined steadily from 1950 until the 1990’s when growth in electric use fell below GDP growth. Similar to the recession in the early 1980’s, the differential between GDP growth and growth in electric use during the Great Recession of the late 2000’s is minimal. The U.S. Energy Information Administration (EIA) reports in its 2015 Annual Energy Outlook (AEO) that average electric use is projected to grow less than 1 percent per year from 2016 through 2040, compared to average real GDP growth of 2.5 percent over the same period. Over the next three decades, the EIA projects that electricity use will continue to grow, but the rate of growth will slow over time. The EIA does not expect growth in electricity use to equal or exceed real GDP growth for any sustained period of time because efficiency standards for lighting and other appliances will continue to put downward pressure on the growth in electricity consumption.

Figure C-1 U.S. Electricity Use and Economic Growth, 1950-2040

U.S. electricity use and economic growth, 1950-2040

Source: U.S. Bureau of Economic Analysis; U.S. Energy Information Agency’s Annual Energy Outlook for 2014 and 2015.

According to the Edison Foundation’s Innovation Electricity Efficiency Institute (IEE), the major factors that are expected to affect growth in electricity use through mid-century are:

The IEE projects that improvements in building energy codes, adoption of appliance/equipment energy standards and expansion of ratepayer-funded energy efficiency programs could result in declining electricity use through 2020 after which time economic growth and the potential growth in use of electric vehicles could result in modest electric growth through 2035. This effect is illustrated in Figure C-2.

Figure C-2 Projected U.S. Electric Energy Use 2010 – 2035

Projected U.S. Electric Energy Use 2010-2035

Source: Innovation Electricity Efficiency, an Institute of The Edison Foundation. “Factors Affecting Electricity Consumption in the U.S. (2010 – 2035).

Per Capita Income Trends

Income is an important determinant of the residential demand for electricity, and changes in income will affect the quantity of electricity purchased. Changes in income affect electric power consumption in two ways. First, a change in income will induce a change in the intensity of use of the existing stock of electricity-consuming appliances; for example, consumers will re-evaluate the intensity of use of a more constrained budget if there is a decline in income. This can be manifested in higher air-conditioning settings or use of lower-wattage lamps for electricity requirements. Second, an income change will induce changes in the stock of electricity-consuming appliances as it impacts consumers purchasing energy efficient devices. As income changes, therefore, the demand for electricity will rise or fall. Previous PPRP forecast studies have demonstrated a positive and, typically, statistically significant relationship between income and the residential demand for electricity.

Real (i.e., inflation adjusted) per capita income can be used as an explanatory variable for residential per-customer electricity consumption. Real per capita income figures are reported in Table C-2 for the Maryland regions defined in Table C-1. Table C-2 summarizes historical and projected data as well as average annual growth rates for the period 2000 through 2025. As shown by the historical data, the rate of income growth has remained constant or has slowed for all regions in Maryland. For the State as a whole, growth in real per capita income declined to 0.73 percent per year between 2005 and 2010, compared to an average annual growth rate of 2.23 percent between 2000 and 2005. All regions of the State, with the exception of Southern Maryland (owing to its proximity to Washington, D.C. and federal government employment opportunities, which drive up wages and the in-migration of relatively high-income households), saw considerable decreases in the rate at which income grew during the 2005-2010 time period relative to 2000-2005. The Upper Eastern Shore region saw a decline in inflation-adjusted income between 2005 and 2010. This slowing was a product of the severe economic downturn and associated job losses affecting numerous Marylanders who lost their incomes, and economic conditions placed downward pressure on wages as the competition for available jobs became more intense.

From 2010 to 2015, the rate of real per capita income growth increased relative to the 2005-2010 period. A forecast by the Maryland Department of Planning for 2015-2020 shows that as the nation (and Maryland) emerges from the recession and the economy once again begins to grow, income will follow the economy’s upward trajectory. Income growth is projected to once again slow (but is not negative) between 2015 and 2020 as the economy returns to steady-state rates of growth lower than those expected during the rebound period that follows the recession.

Table C-2 Historical and Projected Per Capita Income for Maryland, 2000-2025

Region Per Capita Income (2009 $) Annualized Growth
2000 2005 2010 2015 2020 2025 '00-'05 '05-'10 '10-'15 '15-'20 '20-'25
Maryland $42,501 $47,467 $49,221 $52,000 $56,854 $60,112 2.23% 0.73% 1.10% 1.80% 1.12%
Baltimore $41,240 $46,709 $48,850 $52,498 $57,965 $61,589 2.52% 0.90% 1.45% 2.00% 1.22%
Washington Suburban $48,357 $53,167 $54,395 $56,155 $60,675 $63,808 1.91% 0.46% 0.64% 1.56% 1.01%
Southern Maryland $37,765 $41,536 $44,827 $46,626 $51,162 $54,298 1.92% 1.54% 0.79% 1.87% 1.20%
Western Maryland $28,638 $32,391 $34,428 $36,452 $40,332 $42,947 2.49% 1.23% 1.15% 2.04% 1.26%
Upper Eastern Shore $37,822 $42,076 $42,110 $46,155 $50,940 $54,017 2.15% 0.02% 1.85% 1.99% 1.18%
Lower Eastern Shore $30,646 $34,698 $35,873 $37,824 $41,320 $43,592 2.51% 0.67% 1.06% 1.78% 1.08%

Source: Prepared by the Maryland Department of Planning, Planning Data Services, January 2015. Historical data, 1970-2010, from the U.S. Bureau of Economic Analysis.

Employment Trends

Non-residential demand from commercial and industrial electricity consumers is largely driven by their economic output (e.g., customers served, quantities manufactured, etc.). Higher output implies some additional use of electricity. Output data at the county level are not available on a consistent basis, hence, a proxy for output needs to be used. Non-farm employment has typically been relied upon for this purpose. By virtue of the necessity to have adequate numbers of employees to achieve a desired level of output, it is a sound alternative and it is not subject to data consistency problems. Employment data at the regional level are reported in Table C-3.

Table C-3     Historical and Projected Employment for Maryland, 200–2025

Region Total Jobs (thousands) Annualized Growth
2000 2005 2010 2015 2020 2025 '00-'05 '05-'10 '10-'15 '15-'20 '20-'25
Maryland 3,065 3,309 3,345 3,552 3,752 3,881 1.54% 0.22% 1.21% 1.10% 0.68%
Baltimore 1,514 1,609 1,627 1,754 1,846 1,900 1.21% 0.22% 1.52% 1.04% 0.58%
Washington Suburban 1,088 1,183 1,197 1,252 1,324 1,372 1.68% 0.24% 0.91% 1.12% 0.71%
Southern Maryland 124 147 156 162 174 184 3.43% 1.15% 0.84% 1.39% 1.16%
Western Maryland 130 137 136 143 149 156 1.08% -0.20% 0.96% 0.87% 0.87%
Upper Eastern Shore 99 114 115 123 133 140 2.90% 0.26% 1.36% 1.51% 1.03%
Lower Eastern Shore 110 119 114 118 126 130 1.70% -0.85% 0.62% 1.26% 0.63%

Source:  Historical data from the U.S. Bureau of Economic Analysis, Tables CA25 and CA25N. 
Projections from 2015 to 2040 prepared by the Maryland Department of Planning, Planning Data Services, January 2015.

As shown in Table C-3, while every region of the State has seen consistently positive employment growth over the past two decades, the Lower Eastern Shore and Western Maryland were the hardest hit by the recession. Growth between 2010 and 2020 is projected to be most rapid in the Southern Maryland and Upper Eastern Shore regions and slowest in Western Maryland and the Lower Eastern Shore. The City of Baltimore emerged from a recent trend of employment growth lower than the State average (2000-2005) to have a rate of employment slightly higher than the State as a whole from 2010-2015. Overall employment trends for the State tend to track those in the Baltimore and Washington, D.C. suburban regions as these areas contain the largest number of jobs. Both the Baltimore and Washington, D.C. suburban regions, and subsequently the State of Maryland in aggregate, are projected to see similar growth rates through 2025.

The economic downturn in the late 2000’s continued to greatly affect employment, as well as energy consumption, and considerably slowed the employment growth rates between 2005 and 2010. Maryland’s unemployment rate rose from 3.4 percent in 2007 to 7.8 percent in 2010. However, Maryland has still fared better than the United States as a whole. The nationwide unemployment rate in 2010 was 9.6 percent. As with real per capita income, the anticipated growth rebound out of the recession has considerably increased the forecast of job creation through 2025 relative to the recent, much less robust growth between 2005 and 2010. Now out of the recession, the national unemployment rate was down to 5.3% in 2015; Maryland’s unemployment rate was 5.2% the same year.

Recent forecasts of economic indicators (income and employment) have tended to be overly optimistic as the United States begins to emerge from the recent recession, as evidenced by the actual levels of growth in real GDP that the U.S. has experienced in the past few years. Should GDP forecasts continue to underperform, then Maryland PSC 10-Year Plan forecasts will, by virtue of relying on overly optimistic expectations for economic indicators, predict growth in electricity consumption that does not appear as quickly as expected, other factors equal.

Population Trends

Population is an important causal variable because population trends determine (in large part) the number of residential customers. Both the number of households and household size play a role in influencing electricity demand. The number of households affects the number of residential customers purchasing electricity, and changes in average household size can affect usage per customer. Larger numbers of customers mean higher demand, and smaller household sizes (for a given total population) will typically result in higher demand. While smaller households use less electricity in absolute terms, the relationship between size and usage does not scale linearly, as household electricity uses (such as heating and lighting) decline at rates lower than the decline in number of household members. Population growth and the rate of household formation are closely related, and both affect the residential use of electricity. However, household size has seen a slow but steady decline (in Maryland and the United States as a whole) as cultural and societal norms change over time. Deferred marriage and the decision to limit or forgo child-rearing have steadily lowered the size of the average household. Accordingly, increases in population lead to increases in the number of households (and hence residential customers), although these rates of change need not coincide due to changes in the size of households. Population and household data are reported in Tables C-4 and C-5.

Population data at regional and State levels are reported in Table C-4. The table summarizes historical and projected data, as well as average annual rates of growth for the period 2000-2025. The rates of growth in population have been positive since 2000 for every region of Maryland. Between 2000 and 2010, population growth in Maryland was on average 0.87 percent per year. The growth in population for the State is projected to slow through 2025. While following these trends generally, Southern Maryland and the Upper Eastern Shore have seen much more rapid population growth than that in the rest of the State. The rates of growth in population are uneven across the State. Historically, the largest growth rates were reported for Southern Maryland and the smallest rates for Western Maryland. Baltimore’s growth rates are expected to be the lowest during the 2015-2025 period.

Table C-4 Historical and Projected Population for Maryland, 2000-2025

Region Total Population (thousands) Annual Rate of Growth
2005 2010 2015 2020 2025 '05-'10 '10-'15 '15-'20 '20-'25
Maryland 5,296 5,774 6,010 6,225 6,430 0.87% 0.81% 0.70% 0.65%
Baltimore 2,512 2,663 2,746 2,828 2,886 0.58% 0.62% 0.59% 0.41%
Washington Suburban 1,870 2,069 2,182 2,247 2,326 1.01% 1.07% 0.59% 0.69%
Southern Maryland 281 340 363 395 426 1.93% 1.27% 1.73% 1.53%
Western Maryland 237 252 256 266 277 0.65% 0.26% 0.78% 0.81%
Upper Eastern Shore 209 240 247 261 277 1.38% 0.61% 1.04% 1.24%
Lower Eastern Shore 187 209 216 228 238 1.15% 0.63% 1.07% 0.89%

Source:  Projections for the Baltimore region based on Round 8A from the Baltimore Metropolitan Council of Government's Cooperative Forecasting Committee.  Projections for the Washington suburban region based on Round 8.3 of the Metropolitan Washington Council of Governments Cooperative  Forecasting Committee.  Aggregated data prepared by the Maryland Department of Planning, July 2014.

Household data for the State are shown in Table C-5.  The table shows a summary of historical and projected data, as well as average annual rates of growth for the period 2000-2025. Household growth rates differ from population growths due to population demographics and differences in household size. Because of this, household growth captures certain variables, such as the establishment of new households by young adults or the movement of childless couples into the region, which a raw population statistic fails to convey. On average, areas with high household sizes will see higher increases in electricity demand from household growth. Inspecting the rate of change in household size can convey the type of households being added. For example, Southern Maryland is expected to see the highest growth rates in both population and housing in the State. However, it will also see the most rapid decline in household size, suggesting that the households being added may be smaller, and subsequently elicit different changes in electricity demand.

Since 2000, household size in each of the six Maryland regions has been declining or flat, and the decline is forecast to continue through 2025. For the State, average household size was level at 2.61 people during the period 2000-2015. Household size is expected to decline to 2.54 people by 2025.

Table C-5 Historical and Projected Number of Households and Average Size of Households in Maryland, 2000-2025

Region Number of Households (thousands) Average Annual Rate of Growth
2005 2010 2015 2020 2025 '05-'10 '10-'15 '15-'20 '20-'25
Maryland 1,981 2,156 2,248 2,360 2,470 0.85% 0.83% 0.98% 0.91%
Baltimore 959 1,021 1,057 1,102 1,141 0.63% 0.70% 0.84% 0.69%
Washington Suburban 681 746 783 819 858 0.91% 0.97% 0.92% 0.94%
Southern Maryland 98 120 129 143 157 2.05% 1.45% 2.10% 1.88%
Western Maryland 91 97 99 104 109 0.68% 0.42% 0.93% 0.97%
Upper Eastern Shore 80 91 96 102 110 1.39% 0.91% 1.33% 1.50%
Lower Eastern Shore 73 82 85 90 95 1.14% 0.74% 1.22% 1.11%
  Household Size Average Annual Rate of Growth
Maryland 2.61 2.61 2.61 2.57 2.54 0.00% 0.00% -0.31% -0.23%
Baltimore 2.55 2.54 2.53 2.5 2.46 -0.08% -0.08% -0.24% -0.32%
Washington Suburban 2.7 2.73 2.74 2.7 2.66 0.22% 0.07% -0.29% -0.30%
Southern Maryland 2.83 2.8 2.78 2.73 2.68 -0.21% -0.14% -0.36% -0.37%
Western Maryland 2.44 2.43 2.41 2.39 2.37 -0.08% -0.17% -0.17% -0.17%
Upper Eastern Shore 2.58 2.58 2.54 2.5 2.47 0.00% -0.31% -0.32% -0.24%
Lower Eastern Shore 2.43 2.42 2.4 2.38 2.35 -0.08% -0.17% -0.17% -0.25%

Source: Historical data from the U.S. Census.  Forecasts prepared by the Maryland Department of Planning, July 2014.


This appendix provides a review of the theoretical and demographic foundations used for modeling the demand for electricity econometrically. In doing so, emphasis is placed on some of the key determinants of the demand for electric power. The determinants of demand are classified into residential and non-residential, as well as into economic and non-economic for purposes of exposition. Per capita income is an explanatory economic variable that influences the residential demand for electricity; population, the number of households, and average household size are non-economic explanatory variables affecting residential electricity consumption. This appendix also shows trends in employment, which affect the non-residential demand for electricity. Selected data on these determinants of demand are reported and trend analyses presented. The broad conclusion to emerge from these trends is that electricity demand should continue to grow in Maryland.