In July we set out why three different EV shares were published for the same month and all of them were correct. The July figures are a good test of whether that discipline actually helps, and they are, because this month the release contains the problem inside a single document.

What the release says

The Federal Chamber of Automotive Industries reports that a total of 103,656 new vehicles was delivered during July 2026, surpassing the previous July record of 103,097 last year. On battery-electrics it says: Battery electric vehicle sales from all reporting sources were 23,510 during July, accounting for 21.7 per cent of the entire market.

Year to date, it adds, Australians have purchased 127,244 BEVs from all sources, representing 17.3 per cent of the total market.

21.7%

Battery-electric only, as a share of all new vehicles from all reporting sources. The same definition we used for June.

FCAI, July 2026

23.3%

The same measure for June 2026, which was the strongest month ever recorded. July is 1.6 points below it.

FCAI, June 2026

103,656

New vehicles delivered in July, a record for the month. This is the narrower count, and it is not the denominator behind the 21.7 per cent.

FCAI, July 2026

17.3%

Battery-electric share of the total market for the year to date, which is the number to use if you want the trend rather than a single month.

FCAI, year to date 2026

Why 23,510 out of 103,656 is not 21.7 per cent

Do that division and you get 22.7 per cent. The release says 21.7. Neither is a misprint, and the reason is the one this masthead keeps returning to: the two numbers have different denominators.

The BEV share is explicitly “from all reporting sources” and is measured against “the entire market”. The 103,656 is the narrower count. In June the release stated both plainly, 131,134 recorded through VFACTS against 140,058 from all sources, so the two series are a known feature rather than an anomaly. July’s release does not restate the all-sources total, but it can be recovered: 23,510 divided by 21.7 per cent implies a market of about 108,300, roughly 4,700 more vehicles than the 103,656 figure.

That is our arithmetic, and it is approximate because 21.7 is rounded, so the implied total sits somewhere near 108,100 to 108,600. The direction is not in doubt even if the last hundred is.

The gap our earlier piece asked about has narrowed

When we wrote the June story we left an open question: whether the gap between the all-sources count and the VFACTS count was growing, because that channel is non-member and direct-import supply, and a widening gap would quietly distort any share computed on VFACTS alone.

On one further month, the answer is that it narrowed. June’s gap was 8,924 vehicles, or 6.8 per cent of the VFACTS figure. July’s implied gap is about 4,700, or roughly 4.5 per cent. One month is not a trend and the July number is inferred rather than published, so we will keep asking. But the thing to take from it is that the gap moves, which is precisely why a share quoted without its denominator is not worth much.

What the industry says the constraint is

FCAI chief executive Tony Weber used the release to point at charging rather than at supply or price: “There is a risk that the rapid increase in electric vehicle uptake in 2026 will outpace the supply of public charging infrastructure. Governments and the private sector must work together to ensure that the public charging network meets growing demand particularly in regional areas and for motorists who do not have access to charging at home.” He added that accessible and dependable charging will be critical to maintaining consumer confidence.

That last clause is the one that matters for readers of this masthead. The households with the least to gain from a home charger are renters and apartment dwellers, and they are the same households the public network has to serve.

Our reading

Offered as opinion on the figures above. A 1.6 point fall after a record month is not a reversal, and July has been a smaller month than June in the vehicle market for years, so the honest description is a strong month that was not quite the previous one. The year-to-date 17.3 per cent is the steadier number, and it is the one to quote if you want to know where the market actually is rather than where it was for four weeks.

The more useful takeaway is structural. A release can be internally consistent and still produce two percentages that do not reconcile, because the publisher is measuring two different populations in the same document. If you are deciding whether the market has matured enough to buy, ask which vehicles were counted and what they were counted against, every time.