Shift in what is relevant to the CPI

The argument

Many things that were once a part of the Consumer Price Index are no longer relevant due to modern technologies. If you consider a smart phone, it can now do many things that represented a big part of the CPI in the past. For example, taking a photograph, processing it and printing it, is now essentially free. Likewise, while calling was extremely expensive twenty to thirty years ago, Skype now provides unlimited video calling for free. GPS navigation systems provide a hands-free, real-time alternative to maps. Economists call this “consumer surplus”. The consumer benefits and spends less money. This inability for the CPI to pick up on these gains to consumers means that inflation may be overstated.

Premises

Counter-arguments

The premise that these gains go uncaptured describes a practice statistical agencies do not follow. The consumer price index is not a fixed basket carried forward from the past: its contents and weights are revised regularly, with obsolete items withdrawn and new ones introduced — film processing and standalone satellite-navigation devices have been removed from national baskets, smartphones and streaming subscriptions added. Goods that ceased to be relevant do not sit in the index inflating the measured rate, because they are taken out. Quality improvement is handled too, through hedonic adjustment: where a product's specification improves, the price change is decomposed so that the improvement is not counted as inflation. This is applied most aggressively in exactly the categories the argument names — computing, telecommunications, consumer electronics — and the standing criticism of the technique is that it is applied too readily, understating measured inflation rather than overstating it. Whether the adjustment is adequate is a genuine dispute; that it does not happen is not. There is also a conceptual problem with the argument's central move. Consumer surplus is by definition value the consumer does not pay for. A price index measures the changing cost of a basket, and an index that credited unpaid-for benefit would no longer be measuring prices — it would be measuring welfare, which is a different instrument for a different purpose. Excluding surplus is a design decision, not an oversight. Finally, the selection runs one way. Quality change moves in both directions, and the categories that dominate household budgets — housing, healthcare, insurance, childcare, education — have risen faster than the headline index. Choosing the goods that became cheaper and inferring overstatement is the exact mirror of the sibling position's method, and neither selection settles the question.

Rejecting the premises

[Rejecting P1] The basket is not fixed: contents and weights are revised regularly, with obsolete items withdrawn and new ones introduced, so goods that ceased to be relevant are not left in the index as the premise assumes. [Rejecting P2] Quality improvement is already handled through hedonic adjustment, applied most aggressively in the very categories named, and the standing criticism is that it understates rather than overstates measured inflation. Consumer surplus is also by definition value not paid for, so it is not a price and does not belong in a price index.