On 26 February 2026, I submitted my 2025 German tax return electronically. I knew from the beginning that it was not the simplest possible return, and I fully expected that some of the deductions might attract questions. In particular, I had claimed expenses related to my home office and legal fees connected with an employment matter, both of which could reasonably require clarification or supporting evidence. I had no objection to that. A tax authority is entitled to verify claims, ask for documentation and challenge a deduction where the legal treatment is uncertain. What I did not expect was that the process would stretch for almost seven months, involve repeated explanations and submissions of information that had already been provided, and eventually reach the rather extraordinary point where I was asked again for my bank details even though those details had been included in the electronic return from the beginning and were associated with an account I had already used for German tax matters for years. The refund was finally credited to my bank account on 11 September 2026, bringing the matter to a close, but the path between submission and payment became much more than a simple tax assessment. It became an administrative project in its own right.
The first substantial area of discussion concerned my home office. I had a dedicated room that I used exclusively for work, and because I was regularly working from home it formed a genuine part of my employment setup. I understood that this was precisely the type of claim the Finanzamt might want to examine carefully, particularly because home-office deductions depend on actual use, the employment circumstances and the way household costs are allocated. I therefore supplied the requested information and supporting material, including a floor plan and photographs. There were also questions around household costs such as electricity and internet, and I answered those as well. Taken individually, none of these requests seemed unreasonable. What gradually became frustrating was the sense that the same underlying facts were not simply being verified once and then treated as established, but were instead returning repeatedly in slightly different forms, creating uncertainty about whether earlier submissions had been received, understood, associated with the file or accepted.
This distinction became increasingly important as the months passed. There is a significant difference between being asked to substantiate a claim and having to manage a process in which previously supplied information appears not to close the issue. From a taxpayer's perspective, each new request creates a new layer of ambiguity. Was the earlier evidence insufficient? Had it not been received? Had it been received but not linked correctly to the case? Was a new format required? Had another person taken over the file and asked again because the earlier response was not visible? The problem is not only the additional effort involved in answering. It is the absence of clear process state. Instead of responding to a clearly defined open item, the taxpayer begins trying to work out what the administration already knows, what it has accepted and what is actually blocking the case. The home-office discussion came to represent that wider problem for me. I was willing to provide evidence, but I increasingly felt that the process was not cumulative in the way one would expect from a modern administrative system.
A second strand concerned legal fees of about €3,000. The underlying matter related to the previous tax year, while the invoice was paid during 2025, which created a question about the correct year in which the expense should be recognised. The Finanzamt referred to the German cash-flow principle, under which expenses are generally attributed to the year in which they are actually paid. Again, I could understand why this required examination. I had already informed the tax authority about the matter earlier, and I was certainly not trying to claim the same expense twice. The issue was simply one of correct allocation between tax years. Nevertheless, this became another parallel line of correspondence, requiring explanation and clarification while the final assessment remained pending. By this stage, the return had effectively become a multi-threaded case: one part concerned the home office, another the legal costs, and the overall process continued to depend on additional replies and supporting information.
The clearest example of the process problem came in August. On 19 August 2026, almost six months after I had filed the return, I received another letter asking me to provide my bank details. This was particularly difficult to understand because the bank details had already been included in the electronic return submitted on 26 February. They were not new, they had not recently changed, and the same account had already been used for German tax matters. Nevertheless, I supplied the information again. I updated the details through ELSTER, replied in writing and included evidence from the original filing showing that the IBAN had already been present. At that point, the issue was no longer whether I had omitted something required for payment. The issue was that the administration already had the information and was nevertheless asking for it again. If I had failed to provide the bank details originally, the situation would have been straightforward: I would have made an error, the Finanzamt would have identified it, and the responsibility would have been mine. But that was not what happened. The information had been present from the beginning.
By then, the delay had also acquired a much more personal significance. I was unemployed, so this was not simply an accounting issue or money that I could comfortably leave unresolved for an indefinite period. The refund had become important to me in practical terms. I had already planned to send another follow-up letter on 13 September, two weeks after the previous correspondence, making the point that almost seven months had passed, that all requested information had been supplied again and that I needed a clear indication of when the payment would actually be made. That letter ultimately became unnecessary because the refund was credited on the morning of 11 September, two days before I intended to send it. I was relieved, obviously, but the payment did not erase the months that had come before it. Instead, it gave me the opportunity to look at the process as a completed case rather than as an ongoing frustration.
What became clear to me is that the real issue was never scrutiny itself. I do not object to a tax authority reviewing a home-office deduction, questioning the treatment of legal expenses or asking for evidence where the facts are not immediately obvious. That is part of its function. The problem is the quality and continuity of the process through which that scrutiny is carried out. A well-designed administrative process should accumulate knowledge. Once information is provided, it should remain part of the case in a way that reduces future uncertainty. Once an issue is resolved, the taxpayer should not feel that the same question can reappear later without explanation. If data already exists in an electronic filing, the system should ideally be able to reuse it instead of asking the citizen to provide it again. The distinction matters because repeated requests do not simply consume time; they also create uncertainty about whether the process itself has memory.
That is also where the broader question of digital government enters the story. Germany has invested heavily in electronic systems and ELSTER is a central part of the tax process, but digitalisation cannot be measured only by whether a form can be submitted online. A genuinely digital administrative process should reduce duplication behind the scenes as well as paper at the front end. Information entered once should be available where it is needed. Documents submitted electronically should remain attached to the correct case. Previous responses should be visible to whoever handles the next stage. A taxpayer should not have to become the integration layer between different parts of the same administration, repeatedly supplying the same information because one part of the process cannot reliably see what another part already holds. An electronic portal can make submission more convenient while the underlying workflow remains fragmented, and in my experience this case exposed precisely that distinction.
I also do not want to present this as though every difficulty was necessarily caused by the Finanzamt or as though my own return was completely routine. It was not. The legal-fee question involved a legitimate issue of timing. The home-office deduction required detailed factual support. My employment circumstances and later move away from Germany made the overall context less straightforward than a basic salary-only return. Public administrations also operate under constraints that taxpayers do not see: workload, staffing, internal procedures, legal requirements and different responsibilities between departments or individual case handlers. It is entirely possible that some requests that appeared redundant from my side had internal explanations. But recognising those constraints does not make the duplication disappear. If information has to be supplied twice, it has still been supplied twice. If a bank account already present in the original filing has to be provided again before payment, that duplication remains real regardless of the internal reason for it.
This experience also became part of a broader personal conclusion about living in Germany. I would not claim that one delayed tax refund caused me to leave the country, and I would not suggest that my experience represents every resident or every public authority. That would be an unfair generalisation. But administrative friction accumulates. Over time, repeated encounters with processes that require more correspondence, follow-up and persistence than expected begin to affect how one evaluates the practical experience of living somewhere. A country can have strong institutions, detailed laws, sophisticated infrastructure and competent public servants while still producing an exhausting experience for individuals when administrative processes are repetitive or fragmented. For me, this tax case became another example of that accumulated friction. Nothing catastrophic happened. The money ultimately arrived. But almost seven months of uncertainty, repeated submissions and unnecessary process management still represented a substantial amount of mental overhead.
If I were redesigning this kind of process from the taxpayer's perspective, I would not start by demanding faster decisions in every case. Complex claims deserve time. What I would prioritise is continuity and transparency. When information is requested and supplied, the taxpayer should be able to see that the requirement has been satisfied. If the same information is later needed again for a different reason, that reason should be explicit. If the administration already has bank details or supporting documents, the system should recognise that before generating another request. Ideally, there should be a visible case history showing what has been requested, what has been received, what remains unresolved and what is currently preventing completion. Better visibility of process state would reduce uncertainty and eliminate much of the frustration without requiring the administration to compromise the substance of its review.
The matter is now closed. The return was filed on 26 February 2026 and the refund arrived on 11 September 2026. Between those dates came almost seven months of questions, explanations, evidence, legal clarification, ELSTER messages, repeated bank information and follow-up correspondence. The final outcome was successful, and I am genuinely glad that the money is now in my account, particularly because, being unemployed, it matters to me at this stage of my life. But I do not think the correct lesson is simply that the system worked because it eventually reached the right endpoint. A process should also be judged by how much unnecessary effort it imposes on the people required to move through it.
For me, that is the real conclusion. I submitted the return electronically, provided the requested evidence, clarified the legal questions, answered follow-up requests and eventually even re-supplied bank details that had been present from the beginning. Almost seven months later, the refund arrived. The issue was never a lack of information. It was the difficulty of ensuring that information already provided could be used efficiently and consistently throughout the process. That is why, when I think back on this case, I do not primarily see a story about a tax refund. I see a story about the difference between an administration that possesses information and an administration that can actually use it well.
Where relevant, factual references were checked against the sources cited in the article. AI assistance does not replace professional historical or genealogical research, and the final selection, interpretation, opinions and conclusions presented here remain my own.
Comments
Post a Comment