The mid-1990s were a time of real uncertainty—and ambition—in the North American power sector. Deregulation was no longer a distant policy proposal. It was, in many states and provinces, an unfolding reality. For economists and policymakers, tracking the earliest effects of these reforms meant turning to whatever consistent datasets were available. ISIC 3510—electric power generation, transmission, and distribution—provided the main statistical backbone, but, as always, the challenge lay in wringing meaning from broad, imperfect categories.
To begin, one compiles a full list of firms under ISIC 3510 in the US and Canada for 1995. In the US, this means both long-established, vertically integrated utilities and a growing population of independent power producers (IPPs), marketers, and transmission companies. Canada’s list looks somewhat similar—though the role of provincial Crown corporations and municipal utilities is more pronounced. These registries, when paired with annual reports or state/provincial regulatory filings, yield an inventory of who was generating, transmitting, and selling electricity at the moment deregulation began to bite.
Distinguishing deregulated entities from those operating under traditional rate-of-return regulation is not always straightforward. Policy documentation helps: many states and provinces issued public lists of firms subject to new market rules. News coverage and trade publications from the era are useful for flagging the big transitions—divestitures of generation, the formation of regional transmission organizations (RTOs), or the establishment of competitive wholesale markets. Marking these firms and regions on the ISIC 3510 roster creates a first, rough map of the changing landscape.
Next comes the integration of policy change with production and price data. Government energy agencies—such as the US Energy Information Administration (EIA) and Canada’s National Energy Board (now CER)—collected data on generation by fuel type, capacity, system load, and average prices. These datasets can usually be disaggregated to state or province, and, sometimes, to individual firm or plant. The trick is aligning the timing: price and production shifts don’t always track precisely with the day new rules go into effect. There are lags—sometimes anticipation of policy changes moves the market early, sometimes inertia keeps things stable for a while.
To measure deregulation’s effects, analysts often plot pre- and post-policy metrics: total output, reserve margins, average retail and wholesale prices. Comparing trends across regions that restructured with those that stayed regulated adds nuance. For instance, Texas and Alberta—early, aggressive adopters of market-based models—can be set against provinces or states that maintained traditional regulation. It’s rarely a clean experiment; fuel price swings, weather, and technology shocks can muddy the results. Still, a composite view emerges.
Another step is to check for changes in investment. Did new generation come online faster where deregulation took hold? Were there visible shifts in the types of resources—gas turbines, renewables, merchant coal plants—being built? Capital expenditure data, licensing filings, and reports from grid operators can offer some clues. In many regions, restructuring was explicitly intended to attract new entrants and, in theory, lower prices. The actual results, though, are mixed. Some markets saw price volatility and stranded costs, others more stable competition and modest consumer benefits.
Throughout this work, documentation matters. Every source—policy records, output data, price series—should be logged. Some data is missing or patchy; some price trends are confounded by one-off events, like blackouts or market manipulations. Even these ambiguities are instructive. The late 1990s California electricity crisis, for instance, offers both a warning and a data anomaly that has to be treated with care.
The process of using ISIC 3510 to track early North American power sector restructuring is neither straightforward nor perfectly precise. It is, however, a method that allows for a meaningful, if imperfect, comparison across jurisdictions, companies, and outcomes. Patterns emerge—sometimes gradual, sometimes abrupt—of how new rules reshaped the sector’s production, pricing, and investment. The details are uneven; the story, unmistakably, is one of transition in progress.