Evaluating the ACT’s 20 Year Land Value Taxation transition after 8 years
The ACT’s 20 year transition is paying dividends, but not necessarily of the sort we expected. The motivation for this tax transition was to improve the predictability, efficiency and equity of the ACT’s taxation system. Eight years into the twenty year transition it appears that the reforms are achieving all three objectives.
Whilst we had expected a more stable revenue stream, we were pleasantly surprised how much more predictable revenues have become.
We calculated the mean absolute percentage error (MAPE) between ACT Treasury forecast revenue and actual revenue. The orange bars reflect the volatility of conveyances (stamp duties). The blue bars show that rates are generally more predictable. The grey bars represent total taxation, which has become easier to forecast since 2012. For the years 2006-07 to 2011-12 Treasury revenue forecast error averaged 7.9%. In the years since the reforms began, forecast error has fallen to an average of 2.6%
The ACT government now has greater confidence to plan for the future with much more stable expected revenue. The ACT has produced two budget surpluses in a row, with another four planned. Contrast this to NSW where in the 2018-2019 budget they wrote-down $7.2 billion forecast stamp duty revenue only to revise it upwards by $4 billion later in the same year.
We found that the ACT economy is growing strongly, with the reductions in deadweight costs adding to the Gross State Product. The graph below shows positive growth per capita suggesting that productivity gains are driving GSP growth rather than increased migration. Increased productivity is something other states will need post-COVID as we can no longer rely on population growth to underpin our standard of living.
The ACT government has been criticised by both former bureaucrats and the Opposition for failing to deliver on its promise of revenue neutrality. We looked into it from a number of angles. We found that a similar trend would have occurred had the tax reform not taken place, though the proportional contribution of stamp duties and rates would have been reversed, as the following graphs demonstrate.
With land prices and house sale prices rising substantially during the reform period, revenues have risen quite sharply, particularly from general rates. However, stamp duties would also have risen sharply, driven by the same trends, had the rate cuts not been implemented.
We calculated what rates would have been without the reforms, showing they still would have risen, along with the rising property market. So too would have stamp duties, a point often overlooked.
The next big question surrounded how the shift to general rates in the residential sector affected the market. Increased rates means increasing the holding charges for land. We could expect reduced transaction costs to result in more appropriate use of the housing stock, measured by the number of needed or spare bedrooms.
While there has been little change in the reform period so far, we would expect this to be a slow transition taking place over decades, as only decisions at the margins will be affected and housing turnover is slow.
Prices, rents & turnover
In theory, the price impact of increased land taxation is likely to be negative because, to some extent, prospective buyers factor land tax obligations into their purchase price.
That’s the theory, where higher land rates replace stamp duties, leading to lower land prices, alongside a slight increase in the housing component. But what we saw was a raft of outcomes that had us scratching our heads – for a moment or two.
Lead author Warwick Smith saw neighboring Queanbeyan (NSW) as a perfect comparator to the ACT. Just kilometers away and part of the same housing market, Queanbeyan had no increase in land rates and the median price of $600,000 falls under NSW’s threshold for land taxes. We can infer that many investors in Queanbeyan would pay zero land tax – but still stump up significant stamp duty.
Comparing ACT and Queanbeyan it appears that the removal of stamp duty has had more of an effect than the increase in land rates.
While it seems unlikely that the ACT tax transition is entirely responsible for this sharp increase in ACT house prices relative to Queanbeyan, it does indicate that, on balance, the reduction in stamp duties has lifted prices more than they have been depressed by increased rates
As we were pondering our findings, ACT Treasury released three highly modelled and detailed reports by the University of Canberra’s National Centre for Social and Economic Modelling (NATSEM) and ANU’s Tax and Transfer Policy Institute’s (TTPI), Victoria Uni’s Centre of Policy Studies (CoPs) and the Treasury itself. NATSEM’s (2020) analysis helps to make some sense of the Queanbeyan comparison:
“…obtaining a deposit is a far greater obstacle to purchasing property than mortgage repayments (and by extension rates). By contrast, the price impact of reduced stamp duties is likely to be positive. Additional up-front cash, saved from being paid in stamp duty, may be leveraged as part of a larger deposit, meaning prices could rise at greater than one dollar per dollar of stamp duty reductions.”
The ACT’s planning strategy complicates comparisons with Quenbeyan. The majority of new housing development in the ACT is attached dwellings and units due to the strategic preference for densification. Queanbeyan is mostly detached dwellings.
It is notable that unit prices after the transition have remained stable relative to Queanbeyan which is due to ACT unit prices remaining stable.
Illustrating the complexity of the real estate market, we found a strongly positive relationship between rental prices and dwelling supply. Additional supply is usually expected to push rental prices downwards, after an initial period of stickiness. The graph below shows an alternate reality.
The TTPI analysis (NATSEM and TTPI, 2020) indicates that the reforms may have increased rental supply by 200 properties per a month.
However, the TTPI also found the reforms have resulted in downward pressures on rents. Interestingly, these effects were progressively distributed. Rents in the lower four quintiles were estimated to have fallen 2-16% (depending on specification, data, and dwelling type). However rents increased in the top quintile.
The TTPI suggested that existing homeowners may have bought an additional property as their new home, and converted their previous home into rental stock while waiting for further stamp duty cuts before selling.
Such a strategy would allow existing property owners to capitalise on the strong rental market while moving into more suitable housing. This explanation would also fit with differences observed in turnover and price trends between units and houses.
This could also explain why rents have been increasing with increasing dwelling completions. If homebuyers are driving increasing dwelling demand, and this is concentrated at the higher end of the market, it is possible that rents have increased due to higher end rental stock being in shorter supply despite ongoing completions.
With the ACT’s density agenda (similar to Melbourne’s 70% infill/ 30% sprawl) amplified by tax mix shift, a scarcity of detached homes could be driving rents higher. This may also explain why unit prices are more evenly related to Queanbeyan, but house prices aren’t.
This was provocative as market incentives were playing out before us, challenging assumptions. Homeowners were upsizing and converting their old property into a rental. The upsizing may have resulted in higher end rental stock being crowded out (it is unclear if those new rentals have been picked up in the mortgage finance statistics as investor owned properties).
Certainly, these results should serve to allay any fears that the tax transition is a threat to property prices in the ACT! The steep rise in rates revenue has not deterred buyers (nor resulted in a voter backlash).
It also appears to have contributed to tipping the balance towards more homeowners over investors.
Since the reforms began, owner-occupier finance has dramatically outpaced investment finance. The trend also starts prior to APRA’s macroprudential changes to investor financing, and prior to elimination of stamp duty for first home buyers.
Prices have also increased, which means we cannot know if owner-occupiers are buying more property proportionately, in absolute numbers, or are using the stamp duty savings to pay more for property (buying more property value).
Rising prices and stable investor finance would also suggest investors are buying less property, all else equal. In 2012, homebuyers made up about 60% of the purchases by value. In 2020 they now make up about 78%.
However, recent CoreLogic analysis shows that the investor share of lending has been falling in all states, with NSW and ACT leading the pack. So we must be careful not to over attribute causation to the tax shift.
Still, it is great news for potential home buyers who are less often competing with those Mercedes driving silver-foxes at auction.
However, buyer beware: perhaps after years of pent-up frustration at life locked out of the Aussie home owning dream, some buyers have bid on prices based on current cash flows, rather than incorporating future land rates liabilities into their calculations. This homebuyer accounting error may pose a political problem for future governments.
It will take time to see if this is what is occurring but when we look at how investors are behaving in the commercial sector, it suggests the tax transition is contributing to the moderation of commercial prices.
Commercial prices tapering
Rates for commercial land moved from 0.73% in 2011 to a progressive rate of between 3 and 5.15% in 2019 (partly offset by the abolition of land tax on commercial land).
These are significantly higher charges than for residential where properties above $600,000, the top threshold, pay some 0.57%. The ACT’s median is $779,000.
With the higher annual rates in commercial, prospective bidders have to do their sums.
Commercial real estate investment decision making is more likely to be influenced by rational calculation of future cash flows than residential purchasing, particularly by owner-occupiers.
In addition, commercial investors are more driven by rental return and less by capital gains when compared to residential investors. These factors combine to explain the different impact of the reforms on commercial land versus residential land.
Prosper’s Total Resource Rents of Australia report demonstrates how taxes on the productive economy could be shifted onto land and natural monopolies. If a similar commercial rate (5.15%) was set in residential, land price tapering could be expected to occur.
We accept that such a high charge isn’t part of the ACT’s tax shift agenda, but what they are doing is a positive step in the right direction. In time, we expect small business creation will benefit from lower rents.
The graph below wasn’t referenced in the report.
It shows average ACT residential land price growth, only including blocks that existed in 2008-09. This excludes the impact of newly released cheaper blocks.
Land value growth appears to have slowed during the reform period but many factors influence land prices other than taxation, including the supply of residential land and increased amenity. The graph does however provide encouraging signs for affordability.
Higher rates worth screaming about?
The Canberra Liberals campaigned heavily on the cost of living pressures wrought by increasing general rates.
We found average rate payers have paid a cumulative $3,648 extra over 8 years, but still enjoyed land value increases of $63,681.
That’s an extra $456 per year for a nation leading reform. We didn’t cover it, but the ACT’s pensioner assistance scheme has not been over-subscribed, suggesting that financial pain may not be as great as the media headlines have led us to believe.
The election results suggest that Canberrans support the aims of their pioneering tax shift.
And so they should, because while they may be subtle the reform is delivering positive outcomes. We are glad that the reforms are proceeding as planned, and look forward to the third interval. It will take further analysis to get to the bottom of the varied market responses we are witnessing.
More on ACT Land tax transition
Download the full research report The Second Interval: Evaluating the ACT’s 20 Year Land Value Taxation transition after 8 years by Warwick Smith and Jesse Hermans.
Read Warwick’s op-ed ACT’s tax reform agenda hangs in the balance this election published in The Canberra Times: