Takirua v State [2025] FJCA 163; AAU0015.2021 (28 November 2025)

	Home \| Databases \| WorldLII \| Search \| Feedback Court of Appeal of Fiji

CRIMINAL APPEAL NO. AAU 0015 of 2021
[High Court at Labasa Case No. HAC 0060 of 2021]

Counsel: Ms. S. Daunivesi (by order of Court) for Appellant
Mr. R. Kumar for Respondent

[1] The appellant pleaded guilty to a charge of aggravated robbery allegedly committed in the company of two others contrary to section 311 (1) (a) of the Crimes Act.

[2] The summary of facts tendered by the prosecution in support of the charge and admitted by the appellant are as follows:

‘On the 25^th of August 2020, at around 9pm, the Accused persons with another, forcefully entered into the house of a couple, the complainants in this case, Kamlesh Chand, school teacher, and Kajal Karishma Devi, in Seaqaqa. The couple also had another adult make cousin residing with them, Shalvin Chand also a school teacher.

The Accused persons with another approached the back door of the complainant’s house and struck the house with a hard object. They then forcefully entered the house. The Accused persons with another had broken into the house knowing that the couple had a lot of cash with them and intending to steal money from couple.

Once the Accused persons entered the house, they began threatening the occupants of the house. The Accused persons then attacked the occupants of the house namely the two males, Kamlesh who was at the sitting room - he was struck on the chin 3 times, while Shalvin was being beaten on the stomach with an iron rod. Both Kamlesh and Shalvin suffered injuries from the beatings.

The Accused persons were wearing masks. They demanded for money as they continued to assault the occupants of the house, namely the 2 men – Kamlesh and Shalvin. The occupants of the house then told them to take the money inside the van, which was $400. After taking that money, the Accused persons were all still not satisfied.

Then the Accused persons also entered the bedroom of Mrs Kajal Karishma Devi and demanded more money from her. After she gave them some money, they still demanded more and threatened to rape her daughter if she did not give them more. At that point, Shalvin was brought back into the room by the Accused persons, injured. The threats to Mrs Kajal by the Accused persons continued and she was ordered to lift her dress.

The couple’s daughter was siting awake on her bed when the Accused persons threatened to rape her if they were not given more money. The whole family was in complete shock and extreme fear.

Thus, the couple then brought out another brown envelope containing $7000.00 cash. The Accused persons kept all the adults in one room, closed the door and left them there, after stealing the $7000.00. A total of about $10,000.00 was stolen that day however, $8360.75 was recovered.

The Police were later contacted and investigations conducted. The Accused persons were both arrested and there was recoveries made. The Accused persons made their free and voluntary admissions to the Police about their involvement in the robbery.

The admissions and the specific roles they played are as per their record of interviews and the injuries sustained from the above attacks are as documented in the Fiji Police Force Medical Examinations form, that are attached.

According to the medical reports of the victims, they sustained the following injuries from the assault:

Left Face – puncture wound actively bleeding, swollen and collection of blood

[3] The appellant was convicted on his own plea and sentenced on 18 May 2018 by the High Court^[1] to 12 years’ imprisonment with a non-parole period of 10 years.

[4] Though the appellant appealed against his conviction and sentence, he abandoned his conviction appeal on 11 October 2023. A judge of this court allowed him leave to appeal the sentence on 21 March 2024^[2] only on the issue of applicable tariff for the admitted offence.

[5] By the order of this court, Legal Aid Commission (LAC) assisted the appellant to pursue his sentence appeal and at the hearing of the appeal, he agreed to the counsel for LAC to make oral submission on his behalf. The appellant’s written submission too has been submitted by LAC with his agreement.

[6] There is no doubt that the admitted facts unequivocally reveal a night home invasion. The applicable tariff for an aggravated robbery of this nature is within the range of 8-16 years imprisonment^[3].

[7] However, the Supreme Court did not set a starting point in the sentencing tariff for home invasions. Wise only fixed a sentencing range.

[8] The Supreme Court fixed this issue in setting new guidelines in State v Tawake [2022] FJSC 22; CAV0025.2019 (28 April 2022) (‘Tawake’) for aggravated robberies in the form of street mugging and subsequently the Court of Appeal had the occasion to consider Tawake in the context of an aggravated robbery against a taxi driver in Tabualumi v State [2022] FJCA 41; AAU096.2016 (26 May 2022) where the following observations were made:

‘[22] The Supreme Court accordingly set new guidelines for sentencing in cases of street mugging by adopting the methodology of the Definitive Guideline on Robbery issued by the Sentencing Council in England and adapted them to suit the needs of Fiji based on level of harm suffered by the victim. The Court also stated that there is no need to identify different levels of culpability because the level of culpability is reflected in the nature of the offence depending on which of the forms of aggravated robbery the offence takes.
[23] In a significant move the Supreme Court identified starting points for three levels of harm i.e. high (serious physical or psychological harm or both to the victim), medium (harm falls between high and low) and low (no or only minimal physical or psychological harm to the victim) as opposed to the appropriate sentencing range for offences as previously used and stated that the sentencing court should use the corresponding starting point in the given table to reach a sentence within the appropriate sentencing range adding that the starting point will apply to all offenders whether they plead guilty or not and irrespective of previous convictions.
[24] The Court advised the sentencers that they should, having identified the initial starting point for sentence, must then decide where within the sentencing range the sentence should be, adjusting the starting point upwards for aggravating factors and downward for mitigating ones some of which the Court identified but admitted that they were not exhaustive.
[29] The Supreme Court in Tawake also said that is no reason why the methodology proposed and applied therein should be limited to ‘street muggings’ and hoped that in appropriate cases either party may urge the Court of Appeal for this methodology to be considered for sentencing for other offences.’
[9] The Supreme Court in Tawake stated why this kind of approach is necessary:
‘[30] This methodology is new to Fiji. In the recent past the higher courts have usually only identified the appropriate sentencing range for offences. They have only infrequently in recent times assisted judges by identifying where in the sentencing range the judge should start. That has caused difficulties identified by the Supreme Court on a number of occasions: see, for example, Seninolokula v The State [2018] FJSC 5 at paras 19 and 20 and Kumar v The State [2018] FJSC 30 at paras 55-58. If this methodology is used, that problem is avoided. Indeed, there is, in my opinion, no reason why this methodology should be limited to “street muggings”, and it may be that thought will be given in the appropriate quarters to find cases to bring to the Court of Appeal for this methodology to be considered for sentencing for other offences.
[10] By following Tawake guidance and observations, the Court of Appeal in Matairavula v State [4], replicated the sentencing methodology therein mutatis mutandis to offences of aggravated robbery against providers of services of public nature (taxi/bus drivers etc.) and produced a recalibrated sentencing table maintaining the relative differences in sentencing between the three categories (high, medium and low) while adjusting the starting points within the range of 04 to 10 years.

[11] However, for aggravated robbery in the form of home invasions, so far there is no such sentencing table. There is no reason why the approach recommended in Tawake should not be applied for home invasions as well. Recognizing that sentencing tariff for home invasions is between 08-16 years without fixed starting points and following Tawake methodology, the following table is prepared which would provide a helpful guidance to sentencing in instances of home invasions based on Wise sentencing range .

LEVEL OF HARM (CATEGORY)	ROBBERY (OFFENDER ALONE AND WITHOUT A WEAPON)	AGGRAVATED ROBBERY (OFFENDER EITHER WITH ANOTHER OR WITH A WEAPON)	AGGRAVATED ROBBERY (OFFENDER WITH ANOTHER AND WITH A WEAPON)
HIGH	Starting Point: 10 years Sentencing Range: 08 years –12 years	Starting Point: 12 years Sentencing Range: 10 years –14 years	Starting Point: 14 years Sentencing Range: 12 years –16 years
MEDIUM	Starting Point: 08 years Sentencing Range: 06 years –10 years	Starting Point: 10 years Sentencing Range: 08 years –12 years	Starting Point: 12 years Sentencing Range: 10 years –14 years
LOW	Starting Point: 06 years Sentencing Range: 04 years – 08 years	Starting Point: 08 years Sentencing Range: 06 years –10 years	Starting Point: 10 years Sentencing Range: 08 years – 12 years

[12] Wise and Tawake set out the aggravating factors for home invasion and street mugging. However, these are not exhaustive and aggravating circumstances depend on the facts of each case.

[13] If I may, I can usefully add a few other aggravating factors relevant to all robbery and aggravated robbery offences. They broaden the list while retaining Tawake and Wise logic and avoiding double counting. Some of them may have been already stated differently in Wise and Tawake:
Circumstances of the Offence

[14] Tawake also sets out some mitigating circumstances for street mugging. Similar to listing numerous aggravating factors, there is no inflexible list for mitigating factors. I may also add a few of them.

[15] All these aggravating and mitigating circumstances are applicable where appropriate to all types of robberies and aggravated robberies as well.

[16] In the matter of counting aggravating factors, courts should be mindful to avoid double counting^[6]. This is the complaint here as well. Double counting can occur in the following situations (vide Kumar):

‘[57] ...First, a common complaint is that a judge has fallen into the trap of “double-counting”, ie reflecting one or more of the aggravating features of the case more than once in the process by which the judge arrives at the ultimate sentence. If judges choose to take as their starting point somewhere in the middle of the range, that is an error which they must be vigilant not to make. They can only then use those aggravating features of the case which were not taken into account in deciding where the starting point should be.

[58] Secondly, ....... many things which make these crimes so serious have already been built into the tariff. That puts a particularly important burden on judges not to treat as aggravating factors those features of the case which will already have been reflected in the tariff itself. That would be another example of “double-counting”, which must, of course, be avoided.’

[17] Obviously, elements of the offence too should not be counted as aggravating factors. Otherwise, it would also amount to double counting.

[18] The specific criticism is that the trial judge appears to have taken 15 years as the starting point. The Supreme Court addressed a similar concern in Kumar as follows:

‘[56] If judges take as their starting point somewhere within the range, they will have factored into the exercise at least some of the aggravating features of the case. The ultimate sentence will then have reflected any other aggravating features of the case as well as the mitigating features. On the other hand, if judges take as their starting point the lower end of the range, they will not have factored into the exercise any of the aggravating factors, and they will then have to factor into the exercise all the aggravating features of the case as well as the mitigating features. Either way, you should end up with the same sentence. If you do not, you will know that something has gone wrong somewhere.

[19] In Seninolokula the Supreme Court identified a difference of judicial opinion in Fiji in taking the starting point:

‘19. There is here a difference in judicial opinion. In Koroivuki v The State [2013] FJCA 15, Goundar JA said at [26]: ‘As a matter of good practice, the starting point should be picked from the lower or middle range of the tariff.’ On the other hand, a number of trial judges choose the lower end of the range as a matter of routine. For example, in Jemesa Gaunavinaka v The State [2017] FJHC 425, Perera J said at [24] that ‘the starting point of a particular offence should be construed as the minimum period of imprisonment a particular offence should attract’.

[20] Therefore, I think if the trial judge has taken 15 years as the starting point he may have followed Koroivuki. However, the fact remains that there is no mention of a starting point in the sentencing order. What the judge has said is:

‘[5] In the present case, apart from the statutory aggravation, there are additional aggravating factors. The robbery was committed during a home invasion. The offenders threatened a couple, their daughter and a male occupant with physical violence. They wore masks to conceal their identities. They frightened the occupants by striking their house with a hard object before gaining entry. They carried an iron rod with them. They struck two male occupants with the iron rod multiple times. The occupants sustained physical injuries. One occupant was hospitalized due to the injuries from the attack. The couple’s daughter was threatened with rape. The adult female occupant was forced to lift her dress and was humiliated. The robbery was pre-planned and then executed.

[6] The court’s duty is to denounce the violent intrusion of the victims’ home in the present case and impose a sentence that has the effect of deterrence, both on the offenders and others.

[7] One offender is in his late twenties, while the other is in his early twenties. Both have relevant previous convictions. They are not entitled for any credit for their character. The mitigating factors are that the offenders have entered early guilty plea and have saved the court’s time and resources. They made full confessions to police. As a result of their cooperation, a significant amount of the stolen money was recovered and restored to the owner. For these factors I give them a total discount of 3 years. A further reduction is made to the sentence to reflect that the offenders have spent 2 ½ months in custody on remand.

[8] Both offenders are convicted and sentenced to 12 years’ imprisonment with a non-parole period of 10 years.’

[21] I think that the appellant has assumed that the judge has taken 15 years as the starting point and deducted 03 years for mitigation ending up with 12 years of imprisonment. There is no discernible double counting in that respect. However, the judge has indeed committed double counting when he took the fact that the robbery was committed during a home invasion as an aggravating factor. Obviously, robbery in the form of a home invasion is already built in to Wise sentencing tariff of 08-16 years. One cannot count that fact again as an aggravating factor.

[22] The trial judge has set out the summary of facts, aggravating factors and then mitigating factors. He has not stated a starting point. Nor has he enhanced the sentence for aggravation from any starting point. However, he has discounted the sentence for mitigation by 03 years and 2 ½ months for pre-trial custody.

[23] The trial judge has by and large not followed the two-tiered system of sentencing^[7] widely used in Fiji but adopted a ‘instinctive synthesis’ method.

[24] This court in Koyamaibole v State^[8] said on ‘instinctive synthesis’ method as follows:

‘[26] The ‘instinctive synthesis’ approach has been recognized in the sentencing process in Fiji (see Qurai v State ([2015] FJSC 15; CAV24.2014 (20 August 2015) and approved in Kumar & Vakatawa v The State AAU 33 of 2018 & AAU 117 of 2019 (24 November 2022).

[27] The Victorian Supreme Court in R v Williscroft [1975] VicRp 27; [1975] VR 292 at 300 first coined the notion of an "instinctive synthesis" approach to sentencing in 1975, a concept which has been cited and refined multiple times since [see R v Markarian (2005) 228 CLR 357; Hili v R [2010] HCA 45; (2010) 242 CLR 520; R v Morton [1986] VicRp 82; [1986] VR 863)] and which now refers to an exercise whereby "all relevant considerations are simultaneously unified, balanced and weighed by the sentencing judge" (see Sarah Krasnostein and Arie Freiberg "Pursuing Consistency in an Individualist Sentencing Framework: If You Know Where You're Going, How Do You Know When You've Got There?" (2013) 76 Law and Contemp Probs 265 at 268). As a result, the instinctive synthesis approach to sentencing has been characterised as "more art than science" (see Krasnostein and Freiberg at 269.).

[28] To this end, a judge does not need to explicitly lay out the reasons behind the sentence he or she arrives at, because all that matters is the sentence itself (see Grant Hammond "Sentencing: Intuitive Synthesis or Structured Discretion?" [2007] NZ Law Review 211 at 213). It is the intuitive weight that a sentencing judge decides to place on the circumstances of the offence and the offender after the benefit of hearing all the evidence which is important (see JUDICIAL DISCRETION IN SENTENCING: A JUSTICE SYSTEM THAT IS NO LONGER JUST? Sean J Mallett).

[25] Koyamaibole also highlighted the concerns expressed around ‘instinctive synthesis’ method:

‘[29] However, other judges and commentators have viewed this approach with a degree of consternation, noting a number of significant flaws. Kirby J of the Australian High Court felt that the approach lacked transparency and was a "retrograde step" (Wong v R [2001] HCA 64; (2001) 207 CLR 584 at [102] per Kirby J dissenting) because disclosure around how a particular sentence has been formulated and the reasons for that sentence should not be hidden by judicial reference to instinct or intuition, "which does little to provide any useful insight or engender public confidence in the task of sentencing" (see Sally Traynor and Ivan Potas "Sentencing Methodology: Two-tiered or Instinctive Synthesis?" (2002) Sentencing Trends and Issues 25 at [4.2]).

[30] Indeed, consistency itself is not of primary importance under the instinctive synthesis approach. Because judges do not need to explicitly set out the weight they give to certain factors when formulating their "intuitive" decision, it becomes virtually impossible to assess whether like offenders are routinely treated in the same way. This in turn means that "sentences can be inconsistent within a (potentially vast) margin of error yet [remain] legal" (see Krasnostein and Freiberg at 269.).

[31] A further problem around the instinctive synthesis approach is the underlying need for a clear rationale of sentencing. It is one thing to agree that judges should be left with discretion, so they may adjust the sentence to fit the particular combination of facts in the individual case. It is quite another to suggest that judges should be free to choose what rationale of sentencing to adopt in particular cases or types of case. Freedom to select from among the various rationales is a freedom to determine policy, not a freedom to respond to unusual combinations of facts (see Andrew Ashworth Sentencing and Criminal Justice (4th ed, Cambridge University Press, Cambridge, 2005) at [3.3.1]). According to Ashworth, one of the major reasons for sentencing disparity are the different penal philosophies amongst judges and magistrates (At [3.3.1]). This problem would be magnified exponentially in a situation whereby sentencing judges have unlimited discretion to impose a sentence according to their subjective intuition. Intuitions will invariably differ, and can be plagued by bias, ignorance and prejudice (see Mirko Bargaric "Sentencing: The Road to Nowhere" [1999] SydLawRw 23; (1999) 21 Syd L Rev 597 at 609.).

[32] On the other hand when a sentence is reviewed in appeal, again it is the ultimate sentence rather than each step in the reasoning process that must be considered (vide Koroicakau v The State [2006] FJSC 5; CAV0006U.2005S (4 May 2006). In determining whether the sentencing discretion has miscarried the appellate courts do not rely upon the same methodology used by the sentencing judge. The approach taken by them is to assess whether in all the circumstances of the case the sentence is one that could reasonably be imposed by a sentencing judge or, in other words, that the sentence imposed lies within the permissible range (Sharma v State [2015] FJCA 178; AAU48.2011 (3 December 2015). However, every sentence that lies within the accepted range may not necessarily fit the crime.’

[26] The appellant’s complaint of double counting is an example of the negative aspects of applying ‘instinctive synthesis’ method. Tawake methodology would avoid not only double counting associated with the starting point but also bring transparency and consistency to the whole sentencing process.

[27] In the circumstances, the best that this court could do is to follow the principled approach suggested in Sharma v State ^[9] and Koroicakau v The State ^[10]. In determining whether the sentencing discretion has miscarried the appellate courts do not rely upon the same methodology used by the sentencing judge. The approach taken by the appellate court in an appeal against sentence is to assess whether in all the circumstances of the case the sentence is one that could reasonably be imposed by a sentencing judge or, in other words, that the sentence imposed lies within the permissible range. Thus, when a sentence is reviewed on appeal, again it is the ultimate sentence rather than each step in the reasoning process that must be considered.

‘[149] The trial judge had used the instinctive synthesis method....... Nevertheless, I do not see any sentencing error in order to conclude that the sentencing discretion by the trial judge had miscarried. Nor can I say that there is a ‘striking’ or ‘startling’ or ‘disturbing’ disparity....... Neither would I say that the .....term is unreasonable or plainly unjust. I cannot also say that the ...... term ..... is so disproportionate or shocking that no reasonable court could have imposed it. It is the law that an appellate Court will not interfere with the sentence imposed by a trial Court unless it is shown to be manifestly excessive in the circumstances or wrong in principle.’

[29] Having considered all the circumstances of the case including the pre-trial remand period, in my view the sentence is one that could reasonably be imposed by a sentencing judge. The sentence of 12 years fits the gravity of the offence and is quite proportionate to the crime the appellant committed. If the tariff table above formulated were to apply, the offending would fall most likely into horizontal axis of high harm category. The appellant was with two others and they were armed. Thus, he should be considered under the last vertical axis. Therefore, the appellant’s sentence would have been 12 years -16 years. Even if he were to be considered under medium harm category still his sentence range would have been 10 years to 14 years. Therefore, either way his current sentence of 12 years is fully in compliance with Wise sentencing tariff within the application of Tawake methodology.

[30] I have read the judgment in draft and I agree with it, the reasoning and the order.

Solicitors:
Legal Aid Commission (by order of Court) for the Appellant
Office of the Director of Public Prosecution for the Respondent

^[5] A spate of robberies has a higher sentencing tariff of 10-20 years – see Livai Nawalu v The State CAV0012/2012 at paragraphs 27-29
^[6] Seninolokula v The State [2018] FJSC 5 at paras 19 and 20 and Kumar v The State [2018] FJSC 30 at paras 55-58

Court of Appeal of Fiji

Takirua v State [2025] FJCA 163; AAU0015.2021 (28 November 2025)