The Claim
“Invented new non-standard metrics to measure NBN performance, which make Australia appear to rank higher than otherwise.”
Original Sources Provided
✅ FACTUAL VERIFICATION
The claim contains an element of truth but requires significant clarification about what actually occurred [1]. The government did commission research using different metrics to assess NBN performance, but the narrative of intentional deception is overstated.
In October 2019, NBN Co commissioned AlphaBeta Advisors to conduct research on broadband speed comparisons [2]. This research moved away from relying solely on speed-test rankings to assess Australia's broadband performance. The AlphaBeta report, titled "Speed Check: Calibrating Australia's broadband speeds," used government-validated subscription speed data rather than crowdsourced speed-test samples [3].
However, one year prior to this, the Bureau of Communications, Arts and Regional Research (BCARR) had engaged PricewaterhouseCoopers to develop what it described as more appropriate comparative metrics for Australia [1]. This work involved changing the comparison set of countries—removing Singapore and South Korea as comparators and instead using nations deemed more comparable to Australia's geography and development level [1]. With this revised metric, Australia ranked 1st when measuring connections capable of 25Mbps (the NBN's minimum speed requirement), whereas it ranked 8th out of 17 comparable nations on raw percentage of households with fixed broadband access [1].
Missing Context
The claim omits several important contextual points:
1. Speed-test methodology criticisms are legitimate
The AlphaBeta report highlighted genuine methodological problems with popular speed-test rankings. Speed tests often rely on unrepresentative samples—for example, comparing Australia's 3.5 million test samples with Macao's 831 tests (a 95-fold difference in sample size) [2]. These tests also ignore population segments without broadband access entirely, creating a biased picture [2].
2. Different metrics measure different things
Australia's lower ranking on raw speed tests versus its higher ranking on "capability to meet minimum standards" reflects a fundamental difference in policy priorities. The NBN rollout prioritized universal access (equity) over maximum speeds. This is not necessarily deceptive—it's a different policy choice [2][3]. Measuring success by "percentage of population with 25Mbps+" is legitimate if that was the stated goal.
3. International practice
Changing metrics or comparison sets is not unique to Australia. The BCARR's argument—that small city-states like Singapore and South Korea are not valid comparators for a large, geographically dispersed country—is reasoned [1]. Canada, a larger country with similar geographic challenges, was chosen as a more appropriate comparator [1].
4. The broader picture on speeds
Australia's actual average broadband speeds did improve substantially: from 16Mbps in 2014 to 37Mbps in 2019, largely due to the NBN rollout [2]. This is a factual improvement independent of metric choice.
Source Credibility Assessment
ZDNet Article (Primary Source)
ZDNet is a mainstream technology publication owned by Ziff Davis, a respectable tech media company [1]. However, the article is opinion-written by Chris Duckett with a clear editorial stance. The headline states the government "has chosen to deny reality," which is opinion, not neutral reporting [1].
The article is credible as factual reporting (the metrics change did occur) but biased in interpretation (frames these as purely self-serving rather than legitimate methodological debates). ZDNet's framing is more critical than some other sources.
AlphaBeta Report (Referenced in Support)
AlphaBeta is a legitimate strategy and economics advisory firm that was commissioned by NBN Co, so there is some conflict of interest—they were paid by the organization whose performance they were assessing [2]. However, the report uses government-validated data from the ABS (Australian Bureau of Statistics) and OECD sources, not NBN-proprietary data [2].
BCARR (Bureau of Communications, Arts and Regional Research)
BCARR is a government research bureau that conducted this work as part of its mandate to provide policy analysis [1]. However, as a government bureau during a Coalition government administration, it had institutional alignment with the government's interests.
Neutral reference: ACCC (Australian Competition and Consumer Commission)
The ACCC runs the independent "Measuring Broadband Australia" program, which provides real-world performance data separate from both NBN Co and the government's metric-creation efforts [4]. This is the independent measuring body.
Labor Comparison
Did Labor do something similar?
Labor did not inherit or create similar alternative metrics systems for broadband measurement. However, the context is important: Labor developed the original NBN policy (before Coalition took office) with goals centered on fiber-to-the-premises (FTTP) delivering maximum speeds [5]. When the Coalition changed to a multi-technology approach (including slower FTTN), the metrics question became more fraught.
Key difference: Labor's NBN plan emphasized speed as the primary metric (hence why speed-tests were the natural measure). The Coalition's multi-technology approach optimized for cost and speed "adequacy" rather than maximum speeds, which required different metrics to show success [5].
The metric change was therefore not a Labor precedent but rather a response to the Coalition's different policy approach requiring different success measures.
Search finding: No evidence of Labor government creating alternative broadband measurement metrics or gaming speed rankings before their 2007-2013 period. During Labor's governance, standard speed-test metrics and ACCC measurement methods were relied upon [5].
Balanced Perspective
The Criticism Is Partially Valid
Critics are correct that the government strategically chose metrics that made its policy look better. Removing Singapore and South Korea as comparators does change Australia's ranking from much lower to competitive [1]. This is selective presentation of data.
The ZDNet framing of "deny reality and impose a view of its own" is somewhat justified—the government was choosing favorable comparisons rather than the most straightforward "raw speed test" approach [1].
However, the government's logic has merit
Speed-test methodology is genuinely problematic - The AlphaBeta analysis of why speed-tests are flawed (unrepresentative samples, missing access data, volatility) is methodologically sound and acknowledged by researchers [2].
Different metrics measure different policy outcomes - If the goal was universal access with minimum adequate speeds (25Mbps), measuring "percentage at 25Mbps+" is a legitimate success metric, not a deception [3]. If the goal was maximum speeds, speed-tests would be appropriate [3].
Comparator selection matters - The argument that Singapore and South Korea are not valid comparators for Australia is reasonable (city-states vs. large geographically dispersed country), even if conveniently made [1].
The independent ACCC measurements exist - To verify whether Australia's actual performance was being misrepresented, the ACCC independently measured broadband performance. Their reports provide a check on government claims [4].
The core issue: was this deceptive or just different priorities?
Deceptive framing: Claiming Australia ranked #1 globally because it met the 25Mbps standard, without mentioning it ranked 8th on raw broadband access percentage [1]. This is cherry-picking.
Legitimate methodological choice: Using government-validated subscription data instead of crowdsourced speed-tests reflects a different (and arguably more rigorous) methodology [2].
The truth is somewhere in between: the government made legitimate methodological improvements but also strategically selected comparison frameworks and metrics that presented the most favorable picture.
PARTIALLY TRUE
6.0
out of 10
The government did develop alternative metrics that presented NBN performance more favorably—this is factually true [1][2]. However, calling this "inventing non-standard metrics to deny reality" overstates the case. The AlphaBeta research used legitimate, government-validated data and identified real flaws in speed-test methodology [2]. The metrics change reflected different policy priorities (universal access vs. maximum speeds) rather than pure deception [3]. That said, the selective choice of country comparators and emphasis on metrics favoring Australian performance was indeed strategic and somewhat misleading [1]. The ACCC's independent measurements provide a more neutral assessment [4].
Final Score
6.0
OUT OF 10
PARTIALLY TRUE
The government did develop alternative metrics that presented NBN performance more favorably—this is factually true [1][2]. However, calling this "inventing non-standard metrics to deny reality" overstates the case. The AlphaBeta research used legitimate, government-validated data and identified real flaws in speed-test methodology [2]. The metrics change reflected different policy priorities (universal access vs. maximum speeds) rather than pure deception [3]. That said, the selective choice of country comparators and emphasis on metrics favoring Australian performance was indeed strategic and somewhat misleading [1]. The ACCC's independent measurements provide a more neutral assessment [4].
📚 SOURCES & CITATIONS (5)
-
1
zdnet.com
Department of Communications has chosen to deny reality and impose a view of its own.
ZDNET -
2
nbnco.com.au
Nbnco Com
-
3PDF
AlphaBeta speed check report
Nbnco Com • PDF Document -
4
accc.gov.au
The ACCC runs the Measuring Broadband Australia program. It provides information on the real world performance of broadband plans. We publish updates on this page every 3 months.
Australian Competition and Consumer Commission -
5
spectrum.ieee.org
The newly elected government will inherit a floundering AUD $51 billion broadband network that’s providing slower service to fewer properties than planned
IEEE Spectrum
Rating Scale Methodology
1-3: FALSE
Factually incorrect or malicious fabrication.
4-6: PARTIAL
Some truth but context is missing or skewed.
7-9: MOSTLY TRUE
Minor technicalities or phrasing issues.
10: ACCURATE
Perfectly verified and contextually fair.
Methodology: Ratings are determined through cross-referencing official government records, independent fact-checking organizations, and primary source documents.